The Collected Papers of Lord Rutherford of Nelson, Volume One: New Zealand—Cambridge—Montreal [Reprint ed.] 9781138013650

Citation preview

ROUTLEDGE LIBRARY EDITIONS: 20TH CENTURY SCIENCE

Volume 15

THE COLLECTED PAPERS OF LORD RUTHERFORD OF NELSON

This page intentionally left blank

THE COLLECTED PAPERS OF LORD RUTHERFORD OF NELSON Volume One: New Zealand—Cambridge—Montreal

ERNEST RUTHERFORD Edited by JAMES CHADWICK

First published in 1962 This edition first published in 2014 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN and by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 1962 George Allen & Unwin Ltd All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library ISBN: 978-0-415-73519-3 (Set) eISBN: 978-1-315-77941-6 (Set) ISBN: 978-1-138-01365-0 (Volume 15) eISBN: 978-1-315-77925-6 (Volume 15) Publisher’s Note The publisher has gone to great lengths to ensure the quality of this book but points out that some imperfections from the original may be apparent. Disclaimer The publisher has made every effort to trace copyright holders and would welcome correspondence from those they have been unable to trace.

THE COLLECTED PAPERS

OF

LORD RUTHERFORD OF NELSON O.M., F.R.S.

PU BLISHED UNDER THE SCIENTIFIC DIRECTION OF

SIR JAMES CHADWICK.

VOLUME

F.R.S.

ONE

NEW ZEALAND-CAMBRIDGE-MONTREAL

LONDON

GEORGE ALLEN AND UNWIN LTD

FIRST PUBLISHED IN

1962

This book is copyright under the Berne Convention. Apart from any fair dealing for the purposes of private study, research, criticism or review, as permitted under the Copyright Ac!, 1956, no portion may be reproduced by any process without written permission. Enquiries should be addressed 10 the publisher

This volume © George Allen & Unwin Ud, 1962

PRINTED IN GREAT BRITAIN

in 10 pt. Times Roman type BY UNWIN BROTHERS LTD WOKING AND LONDON

FOREWORD A PROPOSAL to publish the seientifie work of the late Lord Rutherford was diseussed shortly after the war, but the aeute shortage of paper at that time, and, as a eorollary, the heavy commitments of those publishers who were interested in the matter, caused it to be abandoned. This earlier proposal was for the publication of a selection of Rutherford's most important papers. The present venture was, from the outset, eoneeived on different lines. It was proposed to include every scientific paper which Rutherford had published either alone or with collaborators; and also a number of general artic1es, formal public lectures, letters to editors, and other communications which seemed worthy of preservation. This comprehensive and, indeed, ambitious scheme was brought to my attention in the autumn of 1956 by Dr Paul Rosbaud. I readily agreed to give it my full support and to aet as scientific editor. This publication of Rutherford's Collected Works will consist of four volumes, of which the first three will contain the papers published in the usual way in scientifie journals. The first volume includes his work in New Zealand, at the Cavendish Laboratory and in Montreal, eovering the years from 1894 to April 1907; the second volume will contain the papers of the Manchester period, 1907 to 1919; and the third volume will cover his period as Cavendish Professor from 1919 to 1937. The fourth volume will include miscellaneous articles, public leetures, letters to editors and, in addition, some obituary notiees of Rutherford. It will also include a bibliography whieh, it is hoped, will be complete. Each of the four volumes will contain aceounts of personal reeollections and appreeiations by some of his friends and colleagues and also portraits and photographs of historical interest. Some of Rutherford's papers were published at about the same time in German or in Freneh as weIl as in English journals. When the two versions are identical, the English version has generally been chosen for publication here. There are, however, oceasions when the German or French version contains additional material, and in these cases that version has been adopted. The primary purpose in this publication of Rutherford's Colleeted Works is, of course, to set up a visible memorial to one of the greatest figures in the history of science; and, at the same time, to make it readily possible for the succeeding generations of young scientists to see what he did, to follow the development of his ideas, and to get at first hand some idea of the magnitude of his contribution to our knowledge of the physieal world. No reader of these volumes ean faH to be impressed by the vigour and direetness of Rutherford's mind, or fail to beeome aware that the pursuit of seientific truth was to hirn an aetivity of the highest intensity, and also a very personal activity. Jt has been the wish and endeavour of an eoncerned with this project that

8

The Collected Papers of Lord Rutherford

these volumes should be produced at the lowest price consistent with good printing and reproduction. In furtherance of this aim, the publishers, Messrs George Allen & Unwin, have been most generously helped by large grants towards the very substantial cost of a four-volume publication and the work of preparation and revision which it entailed. lt is with deep appreciation and gratitude that such aid is acknowledged from the Government of New Zealand, the National Research Council of Canada and to the Leverhulme Trust. The copyright of Lord Rutherford's publications is held by his grandchildren, Dr Peter Fowler, Mrs Elizabeth Rutherford Taylor, Mr Patrick Fowier and Dr Ruth Edwards. We are indebted to them for permission to publish. In addition to those who have contributed in writing to this volume, many former colleagues and friends of Rutherford have given encouragement and help which have lightened the burden of its production. I thank them all most warmly; and I must record here my debt to Sir Henry Dale, who provided the photo graph of the young Rutherford and the 'Den', to Dr F. R. Terroux, who sent me many photographs of apparatus used by Rutherford in Montreal, to Mrs Muriel Howorth for the photograph of Professor Soddy and to Professor N. F. Mott and officers of the Cavendish Laboratory, who made available to me the records of the Laboratory. I wish also to express my appreciation of the help given in other ways by Dr E. W. Steacie, President of the National Research Council of Canada, Dr O. M. Solandt of Ottawa and Sir Miles Clifford, Director of the Leverhulme, Trust and especially by Dr Paul Rosbaud, but for whose initiative and sustained interest this publication would not have been undertaken. J. CHADWICK

CONTENTS NOTE The papers appear here in the chronologicalorder 0/ publication. No attempt has been made to impose uni[ormity in the use 0/ abbreviations, in the quotation o[ re[erences, in the consisteney o[ the use o[ symbols, ete. The sequence o[ the illustrations and figures in the original papers has been maintained. In consequenee, some repetition and some ineonsistencies may beeome apparent when one paper is eompared with another. In some eases, however, obvious misprints and errors have been eorreeted; others may have eseaped detection.

Foreword by Sir James Chadwick

page

The Young Rutherford by Sir E. V. Appleton

7

15

1894

Magnetization of Iron by High-frequency Discharges

25

1895

Magnetic Viscosity

58

1896

A Magnetic Detector of Electrical Waves and some of its Applications On the Passage of Electricity through Gases exposed to Röntgen Rays

80 105

1897

On the Electrification of Gases exposed to Röntgen Rays, and the Absorption of Röntgen Radiation by Gases and Vapours (with a Note by J. J. Thomson) 119 The Velocity and Rate of Recombination of the Ions of Gases exposed to Röntgen Radiation 132 1898

The Discharge of Electrification by Ultra-violet Light 149 Some Reminiscences of Professor Ernest Rutherford during his time at McGill University, Montreal, by H. L. Bronson and Otto Hahn 163 1899

U ranium Radiation and the Electrical Conduction produced by it 169 Thorium and Uranium Radiation (R. B. Owens) 216 A*

10

The Collected Papers

0/ Lord Ruthe/ford

1900

page

A Radioactive Substance emitted from Thorium Compounds 220 Radioactivity produced in Substances by the Action of Thorium Compounds 232 1901

Energy of Röntgen and Becquerel Rays, and the Energy 260 required to produce an Ion in Gases Einfluss der Temperatur auf die "Emanationen" radioaktiver Substanzen 296 The New Gas from Radium (H. T. Brooks)

301

Emanations from Radio-active Substances

306

Dependence of the Current through Conducting Gases on the Direction of the Electric Field 309 Transmission of Excited Radioactivity 325 Discharge of Electricity from Glowing Platinum and the 331 Velocity of the Ions 1902

Ubertragung erregter Radioaktivität

351

Erregte Radioaktivität und in der Atmosphäre hervorgerufene Ionisation (S. J. Allen) 360 Versuche über erregte Radioaktivität

370

The Radioactivity of Thorium Compounds, I (F. Soddy)

376

The Existence of Bodies Smaller than Atoms

403

Penetrating Rays from Radio-active Substances

410

Comparison of the Radiations from Radioactive Substances (H. T. Brooks) 415 The Radioactivity of Thorium Compounds, II (F. Soddy)

435

Deviable Rays of Radioactive Substances (A. G. Grier)

457

The Cause and Nature of Radioactivity, I (F. Soddy)

472

The Cause and Nature of Radioactivity, II (F. Soddy)

495

Excited Radioactivity and Ionization of the Atmosphere (S. J. Allen)

509

Note on the Condensation Points of the Thorium and Radium Emanations (F. Soddy) 528

11

COllIents

1903

page

Excited Radioactivity and the Method of its Transmission The Magnetic and Electric Deviation ofthe Easily Absorbed Rays from Radium A Penetrating Radiation from the Earth's Surface (H. L. Cooke) Radio-activity of Ordinary Materials The Radioactivity of Uranium (F. Soddy) A Comparative Study of the Radioactivity of Radium and Thorium (F. Soddy) Some Remarks on Radioactivity Condensation of the Radioactive Emanations (F. Soddy) Radioactive Change (F. Soddy) The Amount of Emanation and Helium from Radium Heating Effect of the Radium Emanation (H. T. Barnes) Heating Effect of the Radium Emanation (H. T. Barnes) Radioactive Processes

529 549 558 560 561 565 576 580 596 609 611 613 614

1904

Does the Radio-activity of Radium depend upon its Concentration ? Heating Effect of the Radium-emanation Heating Effect of the Radium Emanation (H. T. Barnes) Nature of the Gamma Rays from Radium The Radiation and Emanation of Radium, I The Radiation and Emanation of Radium, 11 Slow Transformation Products of Radium The Succession of Changes in Radioactive Bodies The Heating Effect of the Gamma Rays from Radium (H. T. Barnes) Der Unterschied zwischen radioaktiver und chemischer Verwandlung Les Problemes Actuels de la Radioactivite

618 620 625 640 641 650 658 671 723 725 746

1905

Radium-the Cause of the Earth's Heat Slow Transformation Products of Radium

776 786

12

The Collected Papers of Lord Rutherford 1905

Charge carried by the Alpha Rays from Radium Heating Effect of the Gamma Rays from Radium (H. T. Bames) Note on the Radioactivity of Weak Radium Solutions The Relative Proportion of Radium and Uranium in Radioactive Minerals (B. B. Boltwood) Some Properties of the Alpha Rays from Radium Charge carried by the Alpha and Beta Rays of Radium Slow Transformation Products of Radium

page

789 792 799 801 803 816 830

1906

Some Properties of the Alpha Rays from Radium (Second Paper) Magnetic and Electric Deflection of the Alpha Rays from Radium The Retardation of the Velocity of the Alpha Partic1es in passing through Matter The Relative Proportion of Radium and Uranium in Radioactive Minerals (B. B. Boltwood) Retardation of the Alpha Partic1e from Radium in passing through Matter Distribution of the Intensity of the Radiation from Radioactive Sources Absorption of the Radio-active Emanations by Charcoal The Recent Radium Controversy The Mass and Velocity ofthe Alpha Partic1es expelled from Radium and Actinium Mass of the Alpha Particles from Thorium (0. Hahn)

843 852 854 856 859 870 876 878 880 901

1907

Production of Radium from Actinium 907 The Velocity and Energy of the Alpha Partic1es from Radio910 active Substances Some Cosmical Aspects of Radioactivity 917

PLATES Rutherford at McGiIl University, 1906

jrontispiece

jacing page 16 Ernest Rutherford in 1892 'The Den', Canterbury College, Christchurch, New Zealand Rutherford's magnetic detector for electric waves, 1896 (Cavendish Laboratory)

96

Physics Research Group, Cavendish Laboratory, lune 1898

128

Professor Rutherford at McGill University. Pastel by R. G. Matthews, 1907

192

Platinum grid bolometer for studying the energy of Röntgen and Becquerel rays 288 Diffusion apparatus for determining the molecular weight of radium emanation Apparatus for studying the ionization produced by Röntgen and Becquerel rays 320 Frederick Soddy

384

Apparatus for measuring the rate of decay of thorium 480 emanation Apparatus for studying the magnetic and electric deviation of rays from radium Differential air calorimeter for determining the heating effect of radioactive emanation 640 Research Group, Montreal, 1905-6

832

Apparatus used for electrostatic deflection

896

RUTHERFORD AT McGILL UNIVERSITY, 1906

This page intentionally left blank

This page intentionally left blank

The Y oung Rutherford by

E. V. APPLETON, F.R.S.

IT is a striking illustration of Ernest Rutherford's versatility and intellectual 'Vigour that, during the space of only six years, 1893-98, he successfully pursued experimental enquiries in as many as three branches of physics. As a young investigator, his mind having turned to the matter during his period of instruction as a student, he worked, first, in New Zealand, on the subject of high-frequency magnetization and electrical oscillations. On arriving in Cambridge, the same topics continued to provide him with fruitful lines of research until he became associated with J. J. Thomson in even more pioneer investigations on the nature of the electrical conductivity of gases. Also, before he left Cambridge in 1898, to occupy his first Chair of Physics at McGill University, Montreal, his interest had already become engaged and held in what turned out to be, for him, the still more fertile subject of radioactivity. However, since Rutherford, so to speak, took the subject of radioactivity with him to Montreal, and his first paper on the subject did not appear until he had been there some months, the group of six papers which follow relate only to his work on high-frequency phenomena and on the temporary electrical conductivity of gases initiated by X-rays and ultraviolet radiation. Since Rutherford remained, to the end of his life, the individual investigator-even when he had become the organizer and director of the work of other people-it is a fascinating experience to study his very first essay in research. We look, naturally, for signs of the outstanding characteristics of his later work, the ability to advance scientific knowledge by asking Nature simple and uncomplicated questions ; and we are not disappointed. His work on the high-frequency magnetization and demagnetization of iron was carried out during his fifth year at Canterbury College, Christchurch, New Zealand, and is the subject material of Paper I of the present series. It is entitled 'Magnetization of iron by high-frequency discharges', and, like his second publication, Paper H, on 'Magnetic Viscosity', was published in the Transactions of the New Zealand Institute, the relevant dates being 1894 and 1895 respectively. But, as we shall see, the young experimenter displayed, even in his earliest research, evidence of the directness of his approach, and of his ability to go to the heart of the matter, which characterized all his work hereafter. It is not possible to describe the object, and the results, of Rutherford's first research better than in his own words. In his first published paper he writes: 'Before starting this research I was uncertain whether iron was magnetic in very rapidly-oscillating fields or not. . . . What experimental evidence there was seemed vague and contradictory.' This uncertainty

16

The Collected Papers 0/ Lord Ruther/ord

concerning an important physical characteristic of matter was dispelled by his own experiments, which led him to conc1ude that 'iron is magnetic for frequencies up to 500,000,000 per second'. Rutherford's first experimental researches were carried out in a cellar, normally used as a c1oakroom by students, in Canterbury College. The apparatus then available was primitive, when judged by modem standards; and he has told us, * in his own words, of some of the difficulties he encountered, due to the absence of a central battery of accumulators. 'I found it necessary', he has said, 'to prepare each morning a battery of about a dozen Grove cells. This involved the c1eaning and amalgamation of the zinc plates and adding the necessary acids. . . . I found this battery of low internal resistance a very convenient means of obtaining substantial and steady currents, but, after several hours' work, the battery showed obvious signs of exhaustion and accurate work with it was impossible. ' Nevertheless as Paper I on the 'Magnetization of Iran by High-Frequency Discharges' (p. 25) so amply shows, he made his limited apparatus serve all the tasks he had undertaken. The experimental section of the paper opens with adescription of some tests invested with a notably simple elegance. Having found that high-frequency magnetic forces did leave a steel needle with a residual magnetization, he then went on to show that such magnetization was approximately 'proportional to the surface of the iron, and not to its sectional area, as it is for steady currents'. In other words, due to 'skin effect'; it was the outer layers only of the specimen which had been magnetized. But the matter was not left there. The youthful experimenter, displaying even then his characteristica1ly simple directness, put his magnetized specimen in hot nitric acid and showed that he could dissolve its superficial magnetization away. The controlled fashion in which this was done, however, immediately led to an unexpected result. By measuring the magnetic effect of the specimen, as its dissolution proceeded, using a neighbouring magnetometer, Rutherford was able to show how the intensity of magnetization varied, in sign and magnitude, from the surface inwards; and the result was surprising. Expressed in his own words, he found that: 'All the needles used gave the same result-viz., a thin surface-Iayer magnetized in one direction, and a thicker interior Iayer magnetized in the opposite direction.' Both layers were confined to a depth of rto inch from the surface, the outer magnetic layer occupying onIy a quarter of that depth. 'It is evident from the manner in which the magnetization varies inwards', the young author Iaconically remarks, 'that the iron has been under the influence of an oscillatory discharge. ' Rutherford's explanation of the two oppositely polarized Iayers of surface magnetization which he had discovered was based on his assumption that • Sdence in Development, by Lord Rutherford: The Norman Lockyer Lecture, 1936: British Association for the Advancement of Science, London, W.1.

ERNEST RUTHERFORD IN

1892

'THE DEN', CANTERBURY COLLEGE, CHRISTCHURCH, NEW ZEALAND

This page intentionally left blank

The Young Rutile/ford by Sir E. V. Appleton, F.R.S.

17

the osciIlatory magnetizing field employed was heavily damped. The effects of the first and second half-cycles of that alternating field, which alone he counted appreciable, penetrated to different depths and therefore determined the signs of the magnetization of the inner and outer layers respectively. He could, of course, have associated these two directions of magnetization with the two directions of current flow in his magnetizing solenoid; but, if he did make this check, he does not mention the matter. Tests on the effect of an oscillatory magnetizing field on unmagnetized needles were followed by similar experiments with needles which had previously been magnetized to saturation. Here it was found that the effect of the oscillating current in the magnetizing solenoid was always to diminish the magnetization of the specimen, though that effect was found to be specially marked when the first half-cyc1e of that current tended to cause demagnetization anyway. This result was at once recognized by Rutherford as providing the essential physical basis of an oscillation detector, and therefore of a receiver for electric waves. Further, his diagnosis of the nature of the effect led hirn to the achievement of optimum operating conditions for such a detector. This particular matter is, however, most perspicuously dealt with in Paper III-rather than in Paper I-where Rutherford explains how the natural demagnetizing effect of the specimen plays an important part in the process. Certainly he found that a bundle of short needles, where the natural demagnetizing effect would be marked, proved a notably sensitive detector of rapidly alternating magnetic forces. I t is scarcely necessary here to continue in corresponding detail with the later examples of Rutherford's magnetic experiments, described in Paper I. But we may note, in both Papers land II, the ease with which he deals, quantitatively, with the theory of alternating current phenomena. We may note, too, the elegant design of his apparatus (Paper I) for measuring the decrement of a damped oscillation, based on his earlier study of the highfrequency demagnetization of magnetically saturated steel needles. But it may not be superfluous to stress that one item in this first schedule of experiments, the work on the magnetic detector, proved the starting point ofmuch further work, by other people as well as by Rutherford. Rutherford's second New Zealand paper, 'On Magnetic Viscosity' (p. 58), is noteworthy as containing a description ofwhat he called a 'time apparatus', an ingenious mechanical contrivance by which he could measure small intervals of time down to less than one hundred-thousandth of a second. With this apparatus he measured the time of rise of a magnetizing current in solenoids with iron and steel ring cores. The object of the research was to discover what differences existed between static and dynamic magnetic hysteresis curves. Rutherford's conc1usion was that 'soft iron and steel exhibit quite appreciable viscosity in rapidly-changing fields. The effect is far more marked in the case of steel than in soft iron.' However it is quite c1ear that he found the interpretation of some of his experiments in this research somewhat difficult and not entirely in unqualified conformity with

18

The Collected Papers 0/ Lord Ruther/ord

the conc1usions of his previous paper. * At any rate he did not inc1ude the subject of magnetic viscosity when he came to write Paper In for the Transactions of the Royal Society, and when he obviously selected for inc1usion the high lights of his New Zealand work. Rutherford's magnetic investigations, carried out in Canterbury College, occupied hirn from 1893 to 1895, in which latter year he proceeded to Cambridge with a scholarship awarded by the 1851 Royal Commission. It was characteristic of Professor J. J. Thomson, under whom Rutherford worked as a 'Research Student' in the Cavendish Laboratory, that he should encourage the young physicist first to continue his attack on a problem which was his own. 'J. J.' also generously organized various occasions by which Rutherford's experiments on the emission and reception of electric waves could be made known to a wider scientific circ1e in Britain. Not only was Rutherford invited to demonstrate his experiments to the Cambridge Physical Society, within a few weeks of his arrival in Cambridge, but, in the following year (September 6, 1896) he also exhibited his magnetic detector to a more general audience at the British Association. But undoubtedly Rutherford must have feIt most encouraged of all when 'J. J.' offered to communicate his first paper to the Royal Society. This is Paper In (p. 80), to which we now turn. The careful reader will note that Paper In, entitled 'A Magnetic Detector of Electric Waves and some of its Applications', which was read before the Royal Society on June 11, 1896, inc1udes a substantial amount of the material already described in Rutherford's first New Zealand paper, although it deals much more fully with his work on the detection of electric waves at a considerable distance from their origin. Especially will it be noted, however, that the description of the earlier work has been pruned of much superfluous matter; and we are struck, for the first time, by that economic and masculine style which became characteristic of Rutherford's writings generally. An outstanding feature of Rutherford's continuation of his radio work at Cambridge was the way in which he was able to increase the distance between his sender and receiver. He first experimented with his receiving equipment in a house on Park Parade, the sender being installed on Jesus Common. This gave 'quite a large effect' over a distance of a quarter of a mlle. Later, the receiver was installed on the top floor of the Cavendish Laboratory with the sender in Park Parade, and communication was then established over a distance of more than half a mile. The fact that aseries of walls proved little or no impediment to the propagation of the waves was also recorded. In all these experiments Rutherford employed the high-frequency demagnetization of steel as the intrinsic physical process of detection. The oscillating

* I am informed by two authorities on the subject, Professor L. F. Bates, F.R.S., and Dr. N. Davy, of Nottingham University, that, while it is now believed that the domain boundaries of a magnetie material can move, when prompted, with great rapidity. statie magnetie eharaeteristie eurves are not entirely reliable guides in predicting dynamie magnetie behaviour at a frequeney of, say. 100 Me/s.

lIu: Youilg Rutlteljord by Sir E. ~ '. Appletoll, F. R.."l'.

19

currents stimulated in the receiving equipment by the incoming radio waves passed through a coil in which there was placed a bundle of magnetized steel wires 1 cm. long. The received oscillations caused a diminution of the surface magnetization of the steel wires, which diminution was detected by a neighbouring magnetometer needle. We may here remark that Rutherford stopped short of turning his science into technology. From the standpoint of practical radio-communication his apparatus had one primary failing: it would not transmit the Morse Code. In the language of modern information theory, it would transmit one 'bit' of information, that the sender was 'on' or 'off'; but it would not transmit dots and dashes. However other workers, with notably practical interests, took the problem over. The requirement was, of course, that of restoring the magnetization which had been lost as a result of the first signal, in order to confer readiness on the apparatus to receive a second one. This was first accomplished by E. Wilson who invented apparatus which rendered such re-magnetization automatie. But it was left to G. Marconi to invent the type of magnetic detector which gave the most extensive service in practical use. But Rutherford had not come to the Cavendish Laboratory only to continue his researches on high-frequency electric phenomena. Only a month after he had arrived in Cambridge, W. C. Röntgen had announced to the world his discovery of X-rays. This discovery had excited quite feverish interest in the minds of both scientists and laymen. But in no circ1e was that interest destined to be of greater profit to science than in the Cavendish Laboratory. Using tubes made by his famous assistant E. Everett, J. J. Thomson had hlmself immediately verified that Röntgen rays conferred on gases at ordinary pressure a temporary electrical conductivity. Here was, then, the 'break-through', inviting rapid exploitation and consolidation. As Rutherford himself has said, 'It was essentially aperiod of pioneer advance into a new and fertile territory, when new ground was broken day by day and when discovery after discovery followed in quick succession'. Mter only two terms working, largely by himself, on magnetism and radio, we can therefore weH understand why Rutherford joined J. J. Thomson in this even greater adventure. The first sign of the infiuence of Cambridge and of Professor J. J. Thomson in diverting Rutherford's interests from magnetism to atomic physics is to be found in the joint investigation, made by the Professor and his research student, on the electrical conductivity conferred on gases by exposure to X-rays. This work is described in Paper IV (p. 105), entitled 'On the Passage of Electricity through Gases exposed to Röntgen Rays' and published in the November 1896 issue of the Philosophical Magazine. It is a fair assumption that the theoretical ideas disclosed in this paper were due to J. J. Thomson, while the actual experimental investigations were performed by Rutherford. However in any case the paper can be regarded as marking an epoch, and laying the foundations of the subject with which it deals. Inftuenced by the analogy presented by the conducting gas with the electric

20

The Collected Papers ofLord Rutherford

behaviour of an electrolyte, the authors picture the electric currents flowing through agas exposed to X-rays as consisting of the movement of electrified partic1es of near-molecular size. Such partic1es are called 'conducting partic1es' or 'positively and negatively charged particles'; the terms 'positive ion' and 'negative ion' had not yet been generally adopted. It is significant that the paper describing this work contains what must have been the first formulation of the continuity equation which expresses a kind of profit and loss account for the electrical charges occupying a given volume. Writing it in its most general form we have

dn

di =

q-

M 2 -

div(nv)

.

(1)

where n is the number of charges of either sign per ce. at time t (the medium being considered electrically neutral), q is the rate of charge generation per cc., cx is the recombination coefficient, while (-div (nv» represents the number of ions gained or lost by transport per second by the cubic centimetre under consideration. In this case, the transport was brought about by the applied electric field. According to a later statement of Rutherford's it was J. J. Thomson who first recognized that the rate of recombination should be proportional to n 2 , since the rate of loss of ions by this means was seen by him to be proportional to the rate with which pairs of ions collided. In choosing the symbol cx for the recombination coefficient J. J. Thomson initiated a nomenc1ature which has remained in use ever since. In what is now a classical piece of simple physical theory the relation between the current and voltage is deduced for a slab of air rendered conducting by Röntgen irradiation. When the voltage applied across the slab is weak the current-made up, of course, by the movement of the conducting partic1es-is so small as not materially to deplete the slab ofionization. Under these conditions theory predicts that the current passing through the slab of conducting air should be approximately proportional to the applied voltage, as was indeed experimentally found to be the case. As the voltage applied was increased, however, this relation ceased to hold, since the current taken from the slab of air was taking its toll of the volume ionization. With a still higher voltage the conducting partic1es were swept away at the rate at which they were generated, substantially no recombination having taken place. A saturation current, independent of the magnitude of the applied voltage, was then obtained. Using, again, their recognition of the analogy between electrolytic conduction and gaseous conduction, the authors went on to estimate roughly the number of charged molecules which were reaching the electrodes under saturation conditions. This was certainly an adventurous calculation in those days, but the result can be checked using later knowledge and be shown to have given the correct order of magnitude: the fraction of the gas 'electrolysed' was 3 X 10- 13•

Tl!c YOlll1g Rutlle/Jord hy Sir E. V. Appletol1, F.R.S.

21

The investigation further proceeds with a study of what Rutherford later called 'after-conductivity', the study of the passage of current through a slab of air after the rays which have rendered it conducting are abruptly stopped. Here the conception of ionic mobility is introduced, and the assumption made that, due to the collisions with neutral molecules, ions in movement would immediately reach a terminal velocity which would be proportional to the applied electric fie1d. The method by which such a mobility was first determined is based on an elegant theoretical relationship which received its first formulation. It is to the effect that, T seconds after the ionizing radiation has been switched off, the current passing through the si ab bears the same ratio to the saturation current as does the quantity T to the time the ions take to travel across the slab. Experimental results in the light ofthis relationship yie1ded a value for the ion mobility of ab out 0·33 cm/sec per volt/cm. This velo city was noted as being large compared with the estimated velocities of electrolytic ions. The work conducted jointly by the Professor and his research student was immediately continued by the latter. As a new subject had been opened one can weIl imagine that it was not a case of finding something to do, but of deciding what to do next. A variety of new experimental arrangements were set up by Rutherford, the results of which yielded ample confirrnation of the discoveries he had previously made in collaboration with J. J. Thomson. These results were twofold and, in Rutherford's economic style, were stated to be simply that 'A gas becomes a temporary conductor under the influence of the Röntgen rays, and preserves its power of conducting some short time after the rays have ceased to act'; and 'the conduction in the gas is probably due to the convection of charged partic1es which travel through the gas with a velocity of the order of I cm. a second for a potential gradient of one volt per cm'. However Rutherford broke quite new ground in his examination of what he calls 'the opacity of gases for Röntgen radiation'. In the course of this work he was able to show that the attenuation of X-rays in passing through the volume of agas varied directly with the electrical conductivity conferred on the gas by the passage of the rays through it. This led J. J. Thomson, in a postscript added to Paper V (p. 119), which he communicated for the young author to the Philosophical Magazine, to speculate on the nature of the process of ionization. Assuming that 'Röntgen rays so far resemble light as io be of the nature of an electromagnetic wave or impulse', he pictures a type of quantum mechanism by which a discrete quantity of energy is absorbed, and the wave action correspondingly quenched, when one molecule is dissociated with 'the production of one positive and one negative ion'. Here we may note the introduction, for the first time, of the term 'ion' in this Cambridge series of scientific papers. In Paper VI (p. 132), entitled 'The Velocity and Rate of Recombination of the Ions of Gases exposed to Röntgen Radiation', Rutherford pursued his study of the volume ionization engendered in air and other gases by X-rays.

22

The Collected Papers of Lord Rutherford

Here he is concemed with the ensuing phenomena during the short period after the ionizing action of the rays has been terminated. Two aspects of these phenomena are investigated: the recombination of the oppositely charged positive and negative ions, and the measurement of the speed with which such ions move under the influence of an electric field. By studying what he graphically calls 'the duration of the after-conductivity' when the ionizing X-rays have been cut off, Rutherford was able to subject to a quantitative test the law of recombination, namely dn dt = - rx,n2 • (2) the symbols having their usual significance. His results indicated that the recombination rate was proportional to the square of n, as (2) requires, thus confirming the theory of J. J. Thomson that recombination rate should be proportional to the number of collisions occurring between oppositely charged ions every second. However we must note here that the experiments were developed only far enough to yield values for the quantity rxn, and not the values of rx and n separately. In modern phraseology we would say that the time of relaxation

(2~)

of the ionized medium was determined, for a number of different

gases at atmospheric pressure. Taking the case of air it was found that the after-conductivity fell to half its initial value in a time interval T of 0·3 seconds, from which we may conc1ude that rxn was 3·3 sec. -1. Employing the results of later work, also done in the Cavendish Laboratory, by H. Thirkill and by J. A. Sayers, we may take rx, the recombination coeflicient of ions in air at atmospheric pressure, as approximately 2 X 10- 6 cm. 3 sec.- 1 which indicates that under the condition of Rutherford's experiments the number of ions of either species, positive or negative, was about 106 per cc. Paper VI also contains an account of the first extensive investigation ever made of ion mobilities. In addition to the method used earlier by J. J. Thomson and Rutherford, another one of notable directness was also employed. Here the apparatus was designed so that only, say, the right-hand half of the slab of air between two electrodes was ionized by the incident X-rays. By measuring the brief time required for ions to travel across the dead space to the left-hand electrode, immediately after the X-ray tube had been switched on, it was possible to determine their mobility. Evidently the apparatus could be used for ions of either sign. In both cases the mobility was found to be I· 5 cm per sec. for unit voltage gradient. The work on volume electrification of gases by X-rays was later supplemented by experiments on electrification produced by ultra-violet light on metal surfaces. These researches are described in the last paper ofRutherford's first Cambridge period Paper VII, which is entitled 'The Discharge of Electrification by Ultra-violet Light', published, again, in the Phi/osophical Magazine. This communication is notable as describing some of the most

7he Young Ruthel:ford by Sir E. V. Appleloll, F.R.S.

23

accurate measurements of gaseous ion mobilities which had yet been made. It is also interesting as showing Rutherford's own views concerning the precise physical nature of the photo-electric effect-views which have since been discarded. At the time the work was described, February 1898, the photo-electric phenomenon itself had, of course, become well known, principa1ly through the work of W. Hallwachs. It was usually demonstrated by showing that a negatively charged metal plate tended to lose that charge when irradiated by ultra-violet light. Rutherford's stated objective was to discover the nature of the electrical conduction, and especially of the electric carrier, by which the negatively charged plate was discharged under such irradiation. In Rutherford's first experiment the irradiated plate was positioned on one side of a kind of wind tunnel with a positively charged plate situated on the other side. Negatively charged ions, originating in the immediate neighbourhood of the photo-electric emitter, were therefore subject to two influences: they were drawn across the wind tunnel by the electric field and they were drawn along it by the moving air. From the geometry of the system Rutherford was able to measure the trans verse speed of the ions, promoted by the electric field, in terms of the known longitudinal speed, promoted by the air current flowing through the apparatus. Although Rutherford counted this method incapable of great accuracy, because of the non-uniformity of the wind speed across the cross-section of the tunnel and the effect of eddying motion of the air at the higher speeds, he nevertheless obtained an entirely satisfactory value for the mobility of the ions in air at atmospheric pressure, namely 1·5 cm per sec. for an electric field of 1 volt per cm. However his dissatisfaction with the wind tunnel method had the useful effect of prompting hirn to design another method of mobility determination, of notable elegance and simplicity. In this case the negative ions engendered near the irradiated plate were drawn away by an electric field of an alternating character supplied from the Cambridge town mains. When the electric field was favourable in sign the negative ions moved away from the plate only to reach a certain limiting distance when they started to return under the influence of the revers al of the field. By measuring the maximum distance travelled by the negative ions under an alternating field of known characteristics it was possible to calculate the velocity an ion would attain under an electric field of one volt per cm.-in other words, the ion mobility. For air the velocity was found to be 1·45 cm. per sec.; and this velocity was found to be invcrsely proportional to the air pressure so long as that quantity was not too low. Othcr results obtained with the same apparatus related to the mobilities of ions of different gases. The mobility of the hydrogen ion was found, for example, to be nearly three times that of the atmospheric ion, while that of carbon dioxide was slightly more than half the same quantity. The paper on the nature of the discharge from a plate by way of the photoclectric effect is, however, also notable because it contains Rutherford's own

24

The Collected Papers of Lord Rutherford

views on the physical nature of the discharge process. He noticed that the mobilities found in the case of what he called 'ultra-violet light conduction' were sensibly the same as those he had measured in 'Röntgen conduction', which led him to suspect that 'possibly the carrier is the same in the two cases or in any case not greatly different'. However he then went on to make the erroneous deduction that it is 'the gas near the surface of the negatively electrified plate' which 'is ionized by the action of ultra-violet light'. Later experiments by others have, however, shown that the photo-electric effect can be observed under the highest vacuum conditions, electrons being liberated by the plate itself when appropriately irradiated. In Rutherford's case the liberated electrons must have rapidly attached themselves to gaseous molecules or clusters, and so become negative ions responsive to the influenee of an electric field. There would seem to be no corresponding source of positive ions since ionization by collision was prevented by the high pressures used. We note here, then, a difference between 'ultra-violet light conduction' and 'Röntgen conduction': in the former case we are concemed only with negative ions, while, in the latter case, ions of both signs are of importanee, being engendered, in equal numbers, by the X-rays. At the end of his paper on the discharge of electrification by ultra-violet light, Rutherford promises the continuation of work on the same subject. But even this paper contained mention of the ionizing effect of 'uranium radiation' which we may take as foreshadowing the diversion of his interests to the subject of radioactivity, the unravelling of which was one of his highest achievements. Certain it is that the movement from Cambridge to Montreal coincided with the change in the problems which preoccupied his attention. He might, in a popular lecture, return to his old experiments on radio waves, or even conduct some experiments on the vibrations of buildings in connection with a law case. But these were merely diversions. As the central absorbing subject of his interest there was the physics and chemistry of radioactivity as the great and unknown territory before him. Perhaps he then expressed himself as he onee did in Cambridge, many years later in another connection, when he declared: 'I think this is a grand subject because there is so much in it that we don't know.'

Magnetization of Iron by High-frequency Discharges by E. RUTHERFORD, M.A. From Transactions ofthe New Zealand Institute, 1894, vol. xxvii, pp. 481-513 (Read before the Philosophical Institute of Canterbury, November 7, 1894)

I. Magnetization o/Iron by Leyden-jar Discharges

THE subject of the magnetization of iron in very rapidly-varying fields has been touched upon more or less fully by several different scientists, notably Dr Lodge, Professor J. J. Thomson, Hertz, and a few others. In Dr Lodge's Modern Views 0/ Electricity we find the following: 'But in the case of a discharge of a leyden-jar iron is of no advantage. The current oscillates so quickly that any iron introduced into the circuit, however subdivided into thin wires it may be, is protected from magnetism by inverse currents induced in its outer skin, and accordingly does not get magnetized, and, so far from increasing the inductance ofthe discharge circuit, it positively diminishes it by the reaction effect of these induced currents; it acts, in fact, much as a mass of copper might be expected to do.' In Fleming's Alternate Current Trans/ormer, vol. i, p. 398, there is a description of Dr Lodge's experiments on the effect of iron in rapidlyvarying fields: 'With respect to the apparent superiority of iron, it would naturally be supposed that, since the magnetic permeability of iron bestows upon it greater inductance, it would form a less suitable conductor for discharging with great suddenness electrical energy. Owing to the fact that the current only penetrates into the skin of the conductor there is but little of the mass of the iron magnetized, even if these instantaneous discharges are capable of magnetizing iron . .. the electro motive impulses, or sudden rushes of electricity, do not magnetize the iron, and hence do not find in it greater self-inductance than they would find in a non-magnetic but otherwise similar conductor.' Dr Hertz, from his experiments on oscillating circuits, came to the conclusion that iron was not magnetic for very rapid frequencies; and, to quote from Fleming's abstract of Hertz's researches (vol. i, p. 416), 'Hertz supposed that, as the sclf-inductance of iron wires is for slow alternations from eight to ten times that of copper wires, therefore a short iron wire would balance a long copper one; but this was not found to be the case, and he concludes that, owing to the great rapidity of the alternations, the magnetism of the iron is unable to fo]]ow them, and therefore has no effect on the selfinduction.' And again, p. 423: 'When the wire was surrounded by an iron

26

The Collected Papers 0/ Lord Rutherford

tube, or when it was replaced by an iron wire, no perceptible effect was obtained, confirming the conc1usion previously arrived at that the magnetism of the iron is unable to follow such rapid oscillations, and therefore exerts no appreciable effect.' Stefan has, however, shown that we could expect very Httle alteration in the inductance of a wire, even if it were magnetic, on account of the greater concentration of the current in a magnetic conductor on the surface of the wire. Professor J. J. Thomson (Recent Researches, p. 322, and Philosophical Magazine, 1891, p. 457) has shown that an iron cylinder placed in a solenoid absorbs considerably more energy than a similar non-magnetic conductor of equal conductivity, on account of the higher permeability of the iron. J. Trowbridge ('Damping of Electric Oscillations': Phi!. Mag., December 1891) has shown that the resistance of iron wires damps electrical energy very considerably, and has deduced that iron must have a fairly high permeability to account for the effects observed. Lastly, we have the statement, in the last page of Gray's Absolute Measurements, that the damping of oscillations in a resonator is greater when the wire is of iron than when it is of copper. In order to investigate the effect of 'magnetic penetration' in iron for fields varying very much more rapidly than could be obtained with the use of the 'time apparatus', the readiest means to hand for obtaining a very rapid oscillatory current was the ordinary leyden-jar discharge. The subject of the magnetization of iron in these fields has been very little touched upon since the time that Henry experimented on the effect of leydenjar discharges on the magnetization of steel needles. In the experiments that follow it will be shown that iron is strongly magnetic in rapidly-varying fields, even when the frequency is over 100,000,000 per second. A solenoid was wound on a small glass tube, sixty turns of wire, seven turns to the centimetre. A leyden jar, charged up to a convenient potential by a Voss machine, discharged through this solenoid, and any iron, whether solid or finely divided, placed inside the solenoid was always more or less magnetized by the discharge. Fig. 1 (page 56) C, ordinary leyden jar; A is solenoid; S, spark gap. The whole of the discharge passed through solenoid A. After the discharge had passed the needles were examined by means of a small mirror magnetometer. As this magnetometer is used in all future experiments for testing the magnetization of needles, the construction is briefly explained. It was made on the pattern set forth in Gray's Absolute Measurements, vol. ii, p.79. The needle was small, and arranged in a cavity, so that it was nearly dead beat. The deflection was increased by means of a lamp and scale in the ordinary way. The value of the horizontal component at the needle

.Hagncli::ation

(~rlroll

by lIig":/i"cljuency Discharges

27

was O· 22, and remained practically constant, as there were no masses of iron in the vicinity. It was first settled that the needle placed in the solenoid was unaffected by the charging current from the Voss. The Voss was turned so as to charge up the jars just below the potential necessary to spark across knobs at S. The needle was then removed and tested by the magnetometer. No effect was observed. The effect of discharges on needles of different diameters was first investigated. Length of needles, 7 cm. : (1) Part of steel knitting-needle, diameter 0·103 in.: Deflection 112, at 9 cm. distance from magnetometer needle. (2) Pianoforte steel wire, diameter 0·032 in.: Deflection, 40; distance from magnetometer, 9 cm. (3) Thin steel wire, diameter 0·008 in.: Deflection, 10; distance from magnetometer, 9 cm. Diameter

Deflection

0·103 in. 0·032 in. 0·008 in.

112 40 10

It will be observed from these experiments that the deflection is very nearly proportional to the diameter of the wire. This is to be expected, as the magnetizing forces are confined to a thin skin of the substance. The amount of the magnetization of the wire is proportional to the surface of the iron, and not to its sectional area, as it is for steady currents. In order to show that the effect was a surface one, and did not penetrate any depth, a cylinder of thin copper was placed over the needle. The needle gave no appreciable deflection, showing that the copper cylinder completely screened off any effect on the iron. A thin extern al iron cylinder gave the same effect. In order to determine with accuracy the state of a needle which had been under the influence of discharge, recourse was had to a method of solution of the iron. After several preliminary experiments, dilute HN0 3 at a temperature of boiling water was found to give the most reliable results. In order to test the rate at which the iron was eaten away, a piece of pianoforte-wire 6·5 cm. long, 0·032 in. diameter, was taken and placed inside a solenoid, and subjected to a steady field of 100 C.G.S. units. The needle was then assumed to be magnetized uniformly throughout its section. Fig.2 (page 56) EHFis a glass vessel, inside another glass vessel, ABCD, which is supported on a tripod of copper. Water is kept boiling in the outer vessel by a burner, K. Inside the inner vessel, but not touching it, is the needle, firmly fixed by the ends in a light frame. This frame is supported clear of the vessels by the stand, S. The needle is fixed horizontally at a distance from the magnetometer, R,

28

The Collected Papers of Lord Rutherford

to give a convenient deflection on the scale. As the water is heated up to boiling point the deflection due to the needle decreases slightly, due to the effect of temperature on the magnetic moment of the needle. At a stand alongside, the dilute HN03 is kept in a beaker of boiling water, and when all is ready the HN03 is quickly transferred to the vessel EHF, taking care not to disturb the needle. The moment the HN0 3 reaches the level of the needle in the vessel the time is noted, for at that instant the needle commences to dissolve. Sufficient HN0 3 is poured in to cover the need1e half an inch. As the needle is dissolved the deflection falls, and the deflection at different intervals is carefully noted. This method of fixing the needle first and then pouring in the acid was unavoidable, as the maximum deflection due to the needle could not otherwise be obtained. By keeping the HN0 3 at 100oe. and rapidly transferring it to the vessel (itself surrounded by boiling water) we insure that the needle is covered by HN0 3 at the same temperature during the whole time of solution. Since the amount of acid was large compared with the size of the needle, the effect of solution of the iron would not materially alter the rate at which the needle was dissolved. A uniformly-magnetized steel needle was found to dissolve very regularly till it was reduced to an extremely fine filament, which did not break up until the magnetometer deflection had faUen within 3 div. of zero. The following is the result of an experiment on a uniformly-magnetized needle (needle 0·032 in. in diameter; steady deflection just before acid is poured in = 222): Time in Seconds after Solution begins

o

30

49 56

Deflection

222 217 195 177

373 414

157 147 137 107 97 77 57 47

566

17

90

115

139 206 246 311

454

638

37

7

It. would be expected that the rate of solution of the metal at any instant would be : :&1 to the surface of the metal at that instant-that is, to the

.\!agl1('li=atiol1 (?f fron h.l' 1I(f!h-{requeI1CY Dischmges

29

radius of the wire. This is very accurately the case in the above experiment. The deflection of the magnetometer at any instant is proportional to the sectional area of the wire-Le. to the square of the radius. The radius of the wire at any moment is therefore known. If a curve is constructed whose abscissae represent time, and ordinates the radii of the wire at different intervals, it will be found to be nearly a straight line, with the exception of an irregularity in the beginning of the curve. A needle of steel 0·032 in. in diameter was then taken, and magnetized by passing the discharge from four leyden jars in parallel through the solenoid. Spark-Iength, ,\ in. A correction was made for the fall of deflection when needle was immersed in dilute HN03 at 100°C. Fig. 3 (page 56) The following is the table of observed values of time and deflection. The values of time and deflection are reduced for convenience in plotting curves: Abscissae

Ordinates

0·0 0·7 1·7 3·2

8·5 12·5 14 15 15·3 15·6 15·5 15 14 12 10 8

4·2

5·2 6·2 7 8

9·2

10·5 11·7

12·9 14·4 16·2 18·2 20·2

6 4

2 1 0·5

The steady deflection at first was 85. As the iron commenced to be eaten away the deflection rapidly rose, and reached its maximum, 156. It remained stationary for a short interval at its maximum value, and then rapidly decreased down to zero. When the deflection had fallen to zero the needle was removed, its diameter measured, and found to be 0·013 in. The depth of magnetic penetration was therefore about 0·0095 in. Now, from the results of experiments on the eating-away of uniformlymagnetized needles, we see that the depth to which the iron is dissolved is proportional to the time. Since in 200 sec. the depth dissolved was 0·0095 in., the rate of solution = 0·000047 in. per second.

30

The Collected Papers

0/ Lord Rutherford

If 1 represents intensity of magnetization of a thin circular shell, distance

r from centre of the needle, and M the deflection of the magnetometer at

any instant-

fl.27Tr.dr is : :al to M;

Then

. 1r

I'S •• .al •

•••

to dM .J ur'

If a be radius of needle at first, it has been shown that (a - r) is :

t (the time of action of acid).

Let

a - r

then

- dr

:al

to

= k. t, =

kdt,

and, substituting in equation (1), we get:

1 (a - kt) is : :al to d: ,since dr is : :a1 to dt;

Now,

d:

• ' •• al .. llS ..

1 dM to a - k'-d' t t

is the value of the tangent of the angle that the tangent at any

point of the curve B (see Fig. 3) makes with the axis of abscissae.

~~

is therefore known from curve B. We can consequently determine the curve of variation of 1 from the surface to the centre, although there are not sufficient data to actually calculate 1 in absolute measure. Fig. 4 (page 56) The curve in Fig. 4 is an approximate representation of the magnetization from the surface inwards. The ordinates represent I, the intensity of magnetization. The abscissae represent the distances from the extern al surface of wire. It will be observed that the surface layer is magnetized in an opposite direction to the main part of the magnetized metal. As we go inwards from the surface the intensity of magnetization rapidly decreases till at the point A there is a portion of the meta! which is not magnetized. This will be called the 'neutral point'. On penetrating still further the magnetization changes sign, and rapidly rises to a maximum, which most probably represents an intensity corresponding to the saturation point of steel. The intensity then remains practically constant till at D it decreases very rapidly down to zero. It is evident from the manner in which the magnetization varies inwards that the iron has been under the influence of an oscillatory discharge. The

Magneti=ation of Iron by High-ji-equenc)' Discharges

31

first half-oscillation penetrated to a depth of i"~O in., which is represented by the length OB in the figure. The neutral point A is at a depth of about aho in. from the surface. The second half-oscillation has evidently decreased in amplitude considerably, since the depth of penetration is only a quarter that of the first discharge. In this experiment there is only evidence of two half-oscillations. Several needles were examined which had been magnetized under the influence of various fields and different lengths of spark gap, but the existence of the return oscillation could not with certainty be detected. All the needles used gave the same general result, viz. a thin surface layer magnetized in one direction, and a thicker interior layer magnetized in the opposite direction. In one case examined the depth of penetration of the first discharge was considerably less than doo in. The effect of varying the capacity of the condenser and keeping the selfinductance and the spark gap constant was also investigated. In the first experiment four leyden jars were placed in parallel; the depth of penetration was found to be l~ ()

l>

bchr/irl! =

l>

5

l>

'000"'7 in. 10

l)epth uf penetr.;tion

15

20

l> l>

Poles ur Voss or Intlu{;ti()l/ MlIchi"e C

Magnetometer ale

5

Sc

COhrlenser

M

B s Spark G.;p

A

Termihilk (lf Imluctioh Coi/

S Spirkpp A

B

c A

6

8 ])ec.ay ,;osci//':l't/on '7.

Magnetisation E.R. delt.

B

()f'

Figs. 1-8

Sf)/en(1/g

lI'on W. de R.B. lith

s

8

0

9.

A'

~

F ~

ISprk

.~

.

,8

~

Z' ...

~

...8

~p

E

10.

S"IIf!D;rJ ob

E

d:

."

~

A

A

C

-:0

C

12. B

c

»

d

M

~

.htector Solenoid 0

0

B

Wt.r (11

13

, CtJil

A

~rA'

6t1p

C S"leA()/t/

» c;:i

0

+

C

,

.~

0

Y

~

A

0

15

~~

B

.p

D

CI)

~ :::,.. C

1)

[lJ

A

B

17.

Coil

Milgnetis.ztion of Iron W. de R.ß. 11th

B.R. delI.

Figs. 9-17

Magnetic Viscosity by

E. RUTHERFORD, M.A., B.se.

1851 Exhibition Science Scholar

From Transactions 01 the New Zealand Institute, 1895, vol. xxviii, pp. 182-204 (Read before the Philosophical Institute of Canterbury, September 4, 1895)

research was undertaken to see if steel or soft iron exhibited any appreciable magnetic viscosity when under the influence of very rapidlychanging fields. Ewing had shown that there was a slow creeping-up of the magnetization for some seconds after the magnetizing force had been applied; but considerable difference of opinion has been expressed as to whether the area of the hysteresis curve would be less for a slow cyc1e than for a very rapid cyc1e of less than 10100 of a second. I had already designed the apparatus and the method of reducing the experiments before a copy of the Proceedings of the Royal Society, April 20, 1893, reached New Zealand. I there found an account of experiments by Messrs Hopkinson, Wilson, and Lydall, which in a great measure anticipated what I had intended doing. Later, when I received a copy of Gray's Absolute Measurements, I found an account of recent researches on the same subject (vol. ii, 753-758). Messrs Evershed and Vigneroles had shown that there was very little difference between the energy lost in magnetic hysteresis at periods from 2 seconds to 1~0 of a second. Hopkinson had obtained quite a marked difference between a slow and a rapid cyc1e, and had conc1usively shown that the difference observed was not due to any time effect on the ballistic need1e (Proc. Roy. Soc., April 20, 1893). As the subject of the dissipation of energy due to magnetic hysteresis with varying periods is one of considerable interest, I determined to continue my experiments on the subject, especially as I was enabled to deal with intervals of time much shorter than those in Hopkinson's experiments. In order to carry out these experiments, a special form of apparatus for measuring short intervals of time was designed. It was necessary for the research to be able to measure the times of rise of currents in circuits whose self-induction was chiefly due to the amount of iron in the circuit. The 'time-apparatus' was found to work very satisfactori1y, and by its means time-interva1s of 1ess than loJ-OOO of a second cou1d be with certainty determined. THIS

59

·\1agnetic Viscosily

Description

0/ the Time-apparatus

AB, CD were two solid copper levers, pivoted at A and D respectively. The lever AB was kept pressed against a copper rod F by means of the spring H. The lever CD was kept pressed against the point E of a screw S

,w

L

s H A

K

F

s

o

B C G

M

Fig. 1

by means of the spring K. A vertical nickel wire WG, of length 6 ft., passed between the extremities B, C of the levers, and was tightly stretched. A falling weight LM slid freely on this vertical wire. The shape of this weight is shown on the right-hand side of Fig. 1. A hole passed 10ngitudina1ly through the falling weight, and, in order to prevent undue friction, the hole in the centre of the mass of metal was larger than at the ends, so that the wire could only touch the metal at the extremities LM. In order to hold up the falling weight at any height on the vertical wire an electromagnet was made to slide on the wire, and was held in position at any point by a screw. On turning off the current the weight fell instantly without communicating any movement to the wire. When the ends of the levers B, C were exactIy in the same horizontal plane the falling weight knocked the levers from E and F simuItaneously. If by means of the screw S the lever CD was depressed below AB, the falling weight reached the lever AB first, and after a certain definite interval the lever CD. The interval of time was caIculated as follows:

CM A

B

c.1

e

c.1

Fig.2

D

60

The Collected Papers of Lord Rutherford

Let AB, CD be the two levers when horizontal. Let the screw be given n turns, so that the lever CD is then in the position C' D. Let E and E' be the ends of the screw in the two positions. Let () be the angle CDC'. Let h = height of weight above the first lever. The velocity with which the weight reaches the lever AB is given by y2iii, assuming the body falls freely under the influence of gravity. As the distance between the levers was never greater than ~ in., and h was generally 3 ft., we may assurne the velocity to be sensibly constant over the interval. Let d = distance between threads of screw: Then EE' Let CD levers:

=

I; ED

=

=

nd

11 ; let C'M be vertical distance between the two

C'M = C'D sin ()

=1

nd

V l +n d 2 12

2

2 2

1 nd { l-~V nd = Ir'

+etc. }

Now, nd in these experiments was never more than ~ in., and /1 = 4·81 in. The value of the correction due to C moving over the are of a circ1e may therefore be neglected. The time taken to move over the vertical distance C'M, assuming velocity constant, is given by ndl

t=

In the actual experiments

ltVfih

l' d = 40 10.; 1= 6·125 in.;

/1 =

h=

4·81 in.; 3 ft;

:. t = n X 0·000192 nearly. .'. the time to cross over an interval corresponding to one turn of the screw is 0·000192 sec. The screw-head was divided up into twenty divisions, and the apparatus was quite delicate enough to show a difference for every division of the screw-head when determining the time of rise of currents of very short duration. The apparatus could, therefore, readily measure intervals of time up to 100~OO of a second. Now, this gives the time-intervals as derived from theory. In practice the time-intervals corresponding to one turn must be slightly greater, due to

Ci!

M agile! ic Viscosif)'

the retardation of the falling weight. The eauses of the retardation are-

(I) frietion of the wirc against falling weight; (2) the work done in knocking

away the first lever; (3) frietion of air, ete. As the wire was weIl oiled and placed exactly vertical, cause (1) is very small; as the weight was very heavy compared with the lever AB, the correction for (2) cannot be very great; and (3) is quite insignificant. Later, experimental verification will be given that the calculated values are very nearly the same as the true values. F or the suceess of the experiments it was not necessary that the absolute values of the time-intervals should be known, but only that successive turns of the screw should correspond to equal intervals of time, and this, from the nature of the instrument, is very nearly true. In order to determine the hysteresis curve for soft iron and steel when the current varied very rapidly, the time of rise of the magnetizing current for soft iron and steel rings was obtained by use of the time apparatus.

Arrangement 0/ Experiment A battery of five Grove cells was connected to the binding screws A and B of the time-apparatus. A wire led from B through a non-inductive resistance B01//ist/c Ga/y &t!ery

Contlenser

s

ALeve r

L

R

R

Be

E

Letter

.R 1

M

0

s Contlenser

Ion Rr"rfg

Fig.3

r, thence round the iron ring which is to be experimented on, and back through a resistance-box R to the binding screw A. From one terminal L of the non-inductive resistance r a wire was taken to the back of the screw S. From the other terminal M a wire was led through a resistanee-box R, and thence to one terminal of a j microfarad condenser, the other terminal of which was connected to the binding-screw D in the lever of the time-apparatus. A ballistic galvanometer was connected to Sand D. Since the levers AB, CD were of solid copper they acted as very low resistance shunts to the circuit RML and the ballistic galvanometer

62

The Collected Papers of Lord Rutherford

respectively. When the battery current is turned on only a very minute amount of the current passes round the circuit LMR, since its resistance is many thousand times greater than that of the lever AB. We may therefore assurne, for all practical purposes, that when the shunt AB is in position there is no curTent round the circuit LMR. (1) Suppose the two levers to be exactly level, so that the falling weight knocks them from their contacts simultaneously: When the shunt AB is removed the current commences to rise in the circuit AMB, the equation of rise being given by CR=E_ dN dt

where C = current at any instant; R = total resistance in the circuit; E = total E.M.F. of battery; N = total induction through the iron ring. In all experiments the inductance of the connecting wires was very small, and can be neglected. The E.M.F. at the terminals of the non-inductive resistance r is given at any instant by

e= Cr. Since the shunt ED is knocked from its contact E at the same instant as AB, the whole quantity of electricity required to charge up the condenser to the steady difference of potential between the terminals LM of the noninductive resistance r flows through the ballistic galvanometer. The throw of the galvanometer needle is therefore proportional to the maximum E.M.F. between the terminals L, M. (2) Now, suppose the lever CD is depressed by giving one turn to the screw: On releasing the weight, the lever AB is knocked from B a certain definite interval before the lever CD is reached. During the interval the current has been rising steadily in the circuit BMR. The condenser is charged through the shunt ED, the E.M.F. e between ts coatings at any instant being proportional to the current in BMR. (The wires connecting the non-inductive resistance r to the condenser were short, so that we may assurne, without any sensible error, that the difference of potential between the coatings of the condenser at any instant is equal to the E.M.F. between the terminals Land M of the resistance r.) When the lever CD is reached, the remainder of the quantity of electricity required to charge up the condenser to the steady difference of potential passes through the galvanometer, The throw of the galvanometer is therefore proportional to the value dN h . of dt at t at Instant.

Magf1ctic Viscosity

63

By gradually increasing the distance between the levers by turning the serew we get aseries of values eorresponding to

~ for

different values of

time. When the eurrent has fully risen

~=

0,

so that we then get no throw

in the galvanometer, as the whole quantity flows through the shunt. Sinee the value of ~ is known at any instant, the induetion N through the iron for that instant may be ealeulated, and, sinee thc corresponding current is known, we have all the data required to plot out the hysteresis curve for a very rapid cyc1e. Experimental Verification In order to see 10 what degree of aeeuraey the time-apparatus eould be depended on, the time of rise of the eurrent in a coil 0/ knOl'rn self-inductance was eompared with the theoretical time of rise as determined from the equation E ( l-e -~.t ) C="R L The coeffieient of self-induetion L was very aeeurately determined. The mean value of L was found to be 2·315 X 107 em. The period of the ballistic galvanometer needle in this and all sueeeeding experiments was 7 sec. The sensitiveness of the shunt-levers of the time apparatus was tested, and they were found to work perfeetly, no eorreetion having to be made. The resistanee of the whole cireuit was 15· 65 ohms, and a battery of two Daniells's eells was used in this case. The following are the results of aseries of observations of the deflection of the galvanometer, and the number of turns of the screw at which the defleetions were observed. Eaeh observation is a mean of two experiments at least, and thc curves in many cases were determincd several times: Turns of Screw

Throwof Galvanometer

0 1 2 3 4 6 8 10 13 17 20

99182 74 65 58-!46! 34* 25 19;\-

12i 9J

64

The Colleeted Papers ofLord Rutlzeljord

The current has here only risen to nine-tenths of its maximum value. It was not convenient to have the levers separated by more than twenty turns, so that the whole curve is not completely determined. It has been shown that the throw of the galvanometer at any instant is proportional to dN 1.e. . to L de. . d uctance L. dt: dt In the case 0 f a coil 0 f constant 10 A curve can therefore be constructed whose abscissae represent time and ordinates current. The theoretical 50 curve of rise, calculated from

45 40

Curve l

the equation eR = E - L

~~,

is

plotted alongside the experimental curve. The close agreement between ~30 the two curves shows that the ~2S time-apparatus may be relied on ~ to give very accurate results. It ~2G also shows that the time-intervals ü,It Yt'rtic.llirr. .. fD PIIIX'!' Cflrrtnt .5 theoretically calculated are the Mit ",til.dJlt~ -1I10196s«rmf/r true intervals, and that successive 10 turns of the screw correspond 5 very accurately to equal intervals of time. o 5 ID /S 20 15 30 35 ." In the experiments on magnetic TIME viscosity rings of soft iron and steel were taken, and the times of rise of the magnetizing current determined as explained previously. 55

TABLE FOR CURVE

1

(Dotted curve is the theoretical curve, and the other the experimental curve) Turns of Screw

Observed Values

Theoretical Values

2 4 6

8·8 12·8 17·3 20·7 26·8 32·9 37·6 40·5

6·1 11·45 16·15 20·3 27·1 33·3 36·4 40·8 44·5 46·3

8 12 16 20 26 34 40

44·1

45·5

65

A1aglletic Viscosity Particulars of Soji-iron Ring

Composed of iron wire 0·008 in. in diameter, wound into a ring and thoroughly insulated from eddy currents by shellac varnish. Mean diameter of ring, 8 cm. Sectional area of ring, 0·079 sq. cm. Wound with three sets of coils of 511 turns altogether. The magnetizing force corresponding to one ampere of current round the ring was 25· 5 C.G.S. units. Part;eulars of Steel Ring

Composed of fine steel wire 0·01 in. in diameter, insulated with shellac varnish. Mean diameter, 8·3 cm. Sectional area of ring, 0·14 sq. cm. Wound with two sets of coils; total, 365 turns. The static hysteresis curve for the soft iron and steel was very accurately determined. A special method was used, which allowed each individual point in the curve to be determined several times in succession. From experiments with the time-apparatus it will be seen that the current rose to a maximum in the ring in about lO~ of a second; so that the secondary current must have all passed through the ballistic galvanometer long before there could have been any appreciable movement of the needle. The hysteresis curve for the very rapid cycle was determined for the same maximum values of induction as the static curves, and under exactly the same conditions. Curve 2 (AA) represents the relation between thc values of

~

and

t

(time) for the soft-iron ring. Since

~~*

may be called the back E.M.F. in the circuit at any instant,

when, t

= 0,

dN dt = E, the total E.M.F. of the battery. The value of the

current flowing in the circuit at any instant is therefore known when

~~ is

known. It will be observed that the value of

~ changes rapidly at

the beginning, and then very slowly when the steep part of the hysteresis curve is reached. The value of

~~ changes

again very rapidly at the point

. dw dn * In figures of curves erroneously pnnied as dt or dt'

c

The Collected Papers o[ Lord Rutherford

66

where the hysteresis curve bends over, and gradually falls to zero as the iron reaches its saturation-value for the maximum magnetizing force. Curve 2 (BB) is deduced from the curve AA. If we take any point in the curve A, the magnetizing force is known, and the value of the total 50

8 8

10

Curve2

10

beh di'!'! Dfrtinites:· 000038'JSI!(,.

E,cn diY' ilJscissil corresponds, fJj

20

-775 vnil4,ofm.ign~tisi,!g force

8

10

5

tlw

3D

3D

3D 3D

ii

3D 3D

3D

3D 3D

3D

3D

3D

induction through the iron corresponding to that magnetizing force is proportional to the area of the curve included between the axes, the curve AA, and the abscissa drawn through the point. 50 8

10

C(Jl'~3

Eilen ver/ic tI;Y'.PPPPflsmnIJ

A

.[at" "".il (/i~ ClfrtSlonis j "467Qllits ITMJltIfs)ng foret

TIME

10

10

A 10

B tlw

o

10

ii

20

3D

3D

3D

The values of Band H for any point may thus be determined. The ordinates of the curve BB are drawn proportional to the induction, and the abscissae to the magnetizing force.

67

Magnetic Viscosity

From the curve 2 (BB) curve 5 is plotted, showing the relation between Band H for the rapid cyc1e. The static ballistic curve is drawn alongside for comparison.

s.

Curve 4

TIME

30

Eq,fI d(V'! ='t(}00048 s.e,otids

30

A

A

klthQriz.flMJ'JI4uilib - ofIMgIIdisill! fom

30

30

A

B dn

o

10

30

dt

48

40

30

Curve 3 shows the corresponding relations for the steel-wire ring as curve 2 for the soft iron. Curve 6 shows the hysteresis curves for the slow and rapid cyc1es for soft steel. Curves 4 and 7 show the relations for a softiron ring when the maximum magnetizing forces is much lower than for the first two sets of curves. The value of H in this case was just sufficient to carry the magnetism of the iron up the steep part of the hysteresis curve.

15000

Curve 5

121••

9000

8

6000

3001

H 5

-3000 -6000

-9000 -12000

ID

2D

30

40

The Collected Papers of Lord Rutherfol'd

68

TABLE FüR CURVE

Static Ballistic Curve Magnetizing Force Total Induction =H =B

0 2·35 2·75 3·76 4·3 6·21 8·1 10·53 16·62 21·05 29·14 44·4

5

Rapid-cycle Curve Magnetizing Force Total Induction =H =B

12,936 9,834 8,316 264 5,808 10,032 12,016 13,464 15,492 15,840 16,304 17,163

12,936 10,709 8,060 4,930 115 4,467 8,013 11,806 15,117 16,080 16,682 17,103 17,132 17,162

0 2·5 3·46 3·85 4·43 5·39 6·74 9·05 20·21 30·61 36·38 42·54 43·7 44·4

IU~'

/2.000

Curve 4 8.DDD

o -4.808 -8.000 -/0.000

·41

K-J1I2 ~.7A/) IJ/f/1'fI

4.100

rauv .1!.p:~., ~4"IS'

8

·41

H 6~

81

IOD

Magnetic Viscosity

69

Several more curves for soft iron and steel, with different maximum magnetizing forces and different periods, were also obtained, but, as they showed the same effeet as the curves 5, 6, 7, they are not given here. TABLE FOR CURVE

Static Ballistic Curve Magnetizing Force Total Induction =H =B

0 7·8 10·63 12·45 14·94 16·6 18·26 20·75 24·07 29·05 36·52 48·8 96·3

I

6

Rapid-cycle Curve Total Induction Magne~~i~g Force =B

11,076 10,127 9,638 8,946 7,881 6,890 3,817 5,603 10,543 11,608 12,567 13,019 13,952

The general results of these experiments conc1usively show that soft iron and steel exhibit quite appreeiable magnetie viseosity in rapidly-ehanging fields. The effeet is far more marked in the ease of steel than in soft iron. The greatest departure of the slow-eyc1e from the rapidcyc1e eurve is shown at the 'knee' of the magnetizing eurve.

11,076 10,326 8,826 7,474 4,076 976 2,174 5,274 8,374 10,224 12,374 12,724 13,174 13,524 13,774

0 9·54 15·91 20·1 23·22 24·34 25·4 27·52 28·5 30·73 54·1 64·7 75 83·5 88·5 15,000

I2DIl

9.0

B

I.IIQI

,t.. ~ ~ ~ 't::i

3.000 ~ .~ '§

t

~ ~

4 ,~

-UD -6.00

H

1O

15

C",.,,7

-1.D00 -I2DO

When finely-divided iron or steel is subjected to rapidly-alternating currents the loss of energy due to magnetic hysteresis is greater than for slow

The Collected Papers 0/ Lord Rutherford

70

cyc1es. In the case of steel the loss of energy would be quite 10 per cent more than for slow cyc1es, and in soft iron not so much. In later experiments it was shown that the effect observed was in no way due to any screening of the interior mass of metal from induction. The iron wire of which the ring was composed was of too small diameter to exhibit any appreciable screening effect, due to induced currents, for the period investigated. TABLE FOR CURVE 7 Static Ballistic Curve Magnetizing Force Total Induction =H

0 2·1 3·21 3·68 4·35 5·29 6·7 9·25 14·44

=B

11,952 9,992 5,680 692 5,492 7,844 10,045 12,548 14,704

Rapid-cycle Curve Magnetizing Force Total Induction =H

0 2·61 3·71 4·01 4·33 4·65 5·2 6·7 8·08 10·31 14·44

=B

11,952 9,845 8,546 4,602 1,204 1,794 4,805 7,810 10,186 12,050 14,704

In my paper published last year (Trans. N.Z. Inst., xxvii, art. lix) it was shown that iron could be magnetized and demagnetized when the magnetism was reversed more than 100,000,000 times per second. Soft iron and steel exhibit the effect of magnetic viscosity quite strongly for a frequency of 1,000; but whether the 10ss of energy due to hysteresis increases with the period is not yet known. The molecule of iron can swing completely round in less than a hundred-millionth part of a second; but it is quite probable that the magnetizing force required to produce any given induction is considerably greater for a frequency of 100,000,000 than for a frequency of 1,000. For very rapid frequencies the screening effects are so great that only a very thin skin of the iron is magnetized, and the effect of successive oscillations makes the interpretation of the results very difficult.

Various Uses o/the Time-Apparatus Not only was the time-apparatus a very simple means of determining the times of rise of currents in circuits when a steady E.M.F. was applied, but with different connections the duration of secondary induced currents at

Magnetic Viscosity

71

make and break of the primary could be examined under any conditions required. Very interesting information in regard to the screening effects of solid iron in rapidly-changing fields was deduced, and the subject of the gradual decay of magnetic force in magnetic and non-magnetic conductors, when the magnetizing force was removed, was experimentally verified. The behaviour of the magnetic metals when subjected to rapidly-changing fields is of great practical importance, and the need of very fine lamination of the iron for high rates of alternation is clearly shown in all the experiments. The principle of the time-apparatus can also be used to determine the velocity of projectiles at various points of their path. If two conductors, acting as shunts to the battery and galvanometer circuits respectively, be placed in the path of the projectile at a convenient distance apart, the time taken to traverse the distance between the two could be readily determined by observation of the amount of rise of the current during the interval. In a circuit of known inductance and resistance, the observed deflection of the galvanometer would be proportional to e -

~.t;

and, since

~

is a

constant for the circuit, t could readily be determined, and thus the velocity known. This method is purely electrical, and is capable of great accuracy. The determination of the constants of the circuit is a simple matter, and there are no sources of error introduced.

Time

0/ Rise 0/ Currents in Various Circuits

In the experiments on magnetic viscosity the times of rise of currents in circuits containing iron were determined. It was observed that the nature of the curve of rise varied greatly with the maximum current, and also depended on whether the iron in the circuit was solid or finely divided. To illustrate the difference between the curves of rise for different maximum currents, curve 8 is appended. In curve 8 (A) the maximum magnetizing force is 132·6 C.G.S. units. After the steep part of the magnetizing curve is passed the current rises extremely rapidly, as is evident from the almost vertical line. Time of rise = 0·00173 sec. Curve 8 (B): Maximum magnetizing force, 38· 7 units. None of the changes are so sud den as in the first curve. Time of rise = 0·00192 sec. Curve 8 (C): Magnetizing force, 15 units, which is just sufficient to ascend the steep part of the hysteresis curve. The current rises very gradually, and there are no sudden changes in the curve. The times taken by the currents to rise in the three cases are nearly equal, notwithstanding the fact that the resistance in one case is nearly nine times that of the others. In the above curves the iron was finely laminated, but when the iron is solid the current rises very rapidly for the first few ten-thousandths of a second, and then increases very slowly to its final value. This is due to the

72

Tize Collected Papers of Lord Ruthel/ord

fact that only the surface-Iayers of the iron are magnetized at first, and the induction penetrates but slowly into the mass of the metal, due to the screening effect of induced currents. With large solid electro-magnets the current takes in many cases over a second to rise to its maximum, and after the first 50~O of a second the curve of rise is nearly a straight line. $0

CIJl rd Eil&. ~ VI. ficI. dir/) 2·6j IJInit. 40 n. rgne isin fgl 'Ce

so

EilG

n,r. 1.m ':='0. 004, ~st&

~

B

~

20

B

;! c;..

B

10

B

o

10

B

B

~

20

TIME

30

40

50

The curve of rise in the case of short cylindrical iron rods like the cores of induction-coils resembles very closely curve 1, for the inductance is sensibly constant. If a closed secondary is wound over the primary the current rises much more rapidly than when the secondary is open, as we should expect from theory. Duration of Induced Currents at Make and Break

The time-apparatus could not only be used for determination of times of rise of currents in various circuits, but also for determining the duration of the current in the secondary at make and break. The method is a very simple one, and the duration of the secondary current may be determined under whatever conditions we please, since the resistance and inductance of the galvanometer does not affect the duration of the current in the circuit which is being experimented on. One terminal of the battery is connected to F, and when the lever AB is in position the current passes along the lever BA, through the primary P, and through a resistance-box back to the other electrode of the battery. The secondary circuit is connected through a resistance-box Rand the shunt-lever CD. The ballistic galvanometer is a shunt off the lever ED.

Magnetic Viscosity

73

The resistance in the secondary QEDR may be adjusted to any required value. When the falling weight is released, on reaching the lever AB it breaks the primary. The induced current at break commences to circulate in the secondary round the circuit QEDR. No appreciable part of the current flows through the galvanometer, as the resistance of the lever CD is extremely low. When the weight reaches the lever CD it breaks the secondary circuit QEDR, and the remainder of the quantity of electricity induced at break flows through the ballistic galvanometer. Bal/istic Ga/v. Bal/istic

Lever F

A

RBox

P

Pritniry.

LlNet"

8 C.

E

o

Q.

,S«iinrJart ·R

80x

Fig.4 By varying the turns of the screw-i.e. the interval between the break of the primary and secondary-the quantity of electricity wh ich has passed through the secondary during the different intervals is easily determined. It TIlUst be noted that the galvanometer does not influence the curve so obtained, as the deflection of the galvanometer is proportional to the quantity of electricity which has passed be/ore the galvanometer is placed in the circuit. The duration of the induced current in the secondary is dependent on the self-induction and resistance: the greater the resistance the shorter the duration, and the greater the inductance the more prolonged the duration. Let Land N be the self-inductance of the primary and secondary circuits respectively, and M the coefficient of mutual induction; let Rand S be resistances of primary and secondary; let x and y be the currents in primary and secondary; If E be the E.M.F. of the battery, the equation of rise in the primary is given by dx dy L-+M-+Rx=E' dt dt ' and the equation of rise in the secondary dy N dt

c*

dx

+ M dt + Sy = O.

74

The Collected Papers of Lord Rutherford

From these two equations x and y may be found when L, M and N are constants. When iron, solid or finely divided, is in the circuit, the values of L, M and N are variable, and the values of x and y cannot be determined. The duration of the current in the secondary was determined under varying conditions of lamination of the iron, and a few of the more important results are given. The duration of the induced current at break, when there was no iron in the circuit, was first examined. Two solenoids were wound over one another, and the secondary was of sufficient number of turns to give a convenient defiection in the ballistic galvanometer when the current was broken. Curve 9 (A) shows the quantity of electricity that has passed in the secondary for different intervals of time. Curve 9 (B) is the current-curve, and is deduced from 9 (A); for the . 10 . the ClfCUlt . . at any IOstant . . . b C = - dt' dQ where IS glVen y current fi OWlOg

Q is the quantity of electricity that circulates in the secondary. SD

Carve 9 f.;rn nl1f'iz,rliv!'='OPPP24set

SD

A

t: ~2.

A

~

t:)'

10

8

Cu

B

er

ID

rre

nt C

TIME 30

urv

e 3'0

SD

It will be observed that the current rises rapidly to a maximum, and then slowly decreases in value through a long interval of time. When more than two-thirds of the quantity of electricity had already passed through the secondary, the slightest variation of the screw often caused large alterations in the deflection. This irregularity in the defiections was evidently due to the fact that the current in the secondary was oscillating very rapidly. It was not thought necessary to investigate the oscillations further , as the subject has been treated experimentally by Heimholtz, Schiller, and others. Curve 10 shows what a marked difference there is in the current-curve in the secondary when iron is in the circuit. A secondary was wound over the laminated core of a small inductioncoil, and the duration of the secondary current determined. The currentcurve exhibits two maxima, and is far more irregular than curve 9 (B). This

75

Magnetic Viscosity

was first thought to be due to some experimental error, but further investigations showed that the same peculiarity was exhibited by aU the curves obtained. Curves were also obtained when finely-Iaminated iron and steel rings were used. The duration of the secondary could be varied by altering the resistance. When a large resistance was placed in the secondary the duration was very short. In the cases above considered the induced currents lasted about l~ of a second. 50

Cvry~ '"ibo'uc~d

50

~ ~

IÜxtpJ'l =iJAfPJlljiA ",Je]

;!

IJNU'(llJ

c;..

~

o

re

JD

Cur

Curren

t

'NS tlIOIJOO.-;,A'P)l!J0'l l(Je/

50

Cbr~ /0

cvrr'~t oif /bru~

40

TllMl 10

'20

30

40

50

When solid iron is in the current the duration of the secondary is greatly prolonged, and is independent in a great measure of the resistance of the secondary. Asolid iron ring was taken and wound with appropriate magnetizing and ballistic coils. It was found that the secondary was of long duration. When 1,000 ohms were added to the current very little difference was observed. If the lines of force had passed suddenly out of the primary, as in the case of the laminated core, the duration of the secondary induced current would have been diminished by increasing the resistance in the secondary; and yet in the case of the iron ring the effect was scarcely appreciable. Clearly, then, the lines of force must pass out of the primary very slowly to account for the observed effect. The magnetic force in the iron changes very slowly when the current is broken, on account of the induced currents in the mass of the metal tending to prevent the decay of magnetic force through the iron.

Decay 0/ Magnetic Force in fron and Copper Cylinders The very slow rate of decay of the magnetic force in an iron cylinder, which was observed in the experiments on the induced current at break, led to a

76

The Collected Papers of Lord Rutherford

series of more detailed experiments on the rate of decay of magnetic force when a uniform field was suddenly removed. The subject is treated mathematically, pp. 352-358, in Thomson's Recent Researches, but I am not aware that the subject has been experimentally verified. Suppose a meta! cylinder be placed in a solenoid, and a steady current be sent round the solenoid. If the current is suddenly broken there are induced currents in the mass of the metal tending to maintain the original state of the magnetic field, and instead of sinking abruptly the field decays very slowly. In order to experimentally test the rate of decay of induction in such a cylinder, a solenoid 10 cm. long was wound uniformly with wire, ten turns to the centimetre. A secondary coil was wound over the primary, sufficient to give a convenient deflection in the galvanometer. On breaking the steady current flowing through the primary an induced current circulates through the secondary, and the duration of this secondary current depends on the resistance and inductance in the secondary circuit. If sufficient non-inductive resistance be added in the secondary the duration of the induced current may be readily reduced to less than ~oo of a second. If the copper cylinder be now introduced into the solenoid the duration of the secondary is considerably prolonged, and its curve of rise and decay may be determined by the same method which has been used before. The arrangement for the experiment is exactly the same as in Fig. 4. 1,000 ohms non-inductive resistance was added in the secondary circuit, and the duration of the secondary was less than l~ of a second. The solid copper rod was now placed in the circuit, and at break the induced current was found to be considerably prolonged, due to the time taken for the magnetic force in the cylinder to decay. From the fact that when there is no metal the whole current has passed in less than lO~OO of a second, we see that the current circulates in the secondary almost instantaneously after the lines of force pass out of the primary. When the copper cylinder is placed in the solenoid the quantity of electricity that flows in the secondary for any definite interval is proportional to the number of lines of force that have passed out of the primary in that interval. Let N = total induction through secondary; let a and b be the areas of primary coil and copper cylinder respectively: The induction through the copper = The part of the induction N (1 -

~.N. a

~ ) decays very suddenly; but the induc-

tion through the copper decays gradually.

77

Magnetic Viscosity

On pp. 356, 357, Recent Researches, a table is given for the theoretical calculated values of the rate of decay for aseries of values ; where T

47TW2

= --,

a

where r

.. IS

.

radIUs of the cyhnder.

Now, for this experiment, assuming IL = 1,

=

T

G

= 1,600,

0·0069 sec. approximate1y: THEORETICAL TADLE

-Tt

Total Induction

1·0 0·7014 0·5904 0·5105 0·4470 0·3941 0·2178 0·1220 0·0684 0·0384

0·00 0·02 0·04 0·06 \J·08 0·10 0·20 0·30 0·40 0·50 Ig

Cultve /f

E~cIt v~rtit I/~= ifomi;f'inyu&fi~

Ig

dill!~·OOfJD48~ec.

IN/)IJ,CTI()~

Ig

4elt /kJriz

ID

o

Ig

7/ ~E Ig

30 Ig

40 "'"0

Copper Cylinder Curve 11 shows the rate of decay of the total induction through a copper cylinder 1· 875 cm. in diameter. The dose agreement between the theoretical and experimental curves is a confirmation of the mathematical theory, for the difference between the two is quite within the limits of experimental

78

The Collected Papers oJ Lord RutherJord

error. The induction falls rapidly at first, and then very slowly, so that a long interval elapses before the induction has fully fallen. In 0·00074 sec. the induction has fallen to half its original value. SoJt-iron Cylinders Curve 12 shows the rate of decay of induction in soft-iron cylinders of diameter 0·676 cm. and 0·573 cm. respectively. The rate of decay is much slower than in the case of copper, on account of the high permeability of the iron, although the diameter and conductivity are less for the iron than the copper. The greater the radius of the cylinder the longer the induction takes to decay. 40

C, 'I've /2

.",.n ~ o'~ ~=a 11-4fll iulUj !ion

lEich

so

Ead jhori o'iVl. = o(l UOO9 pStC

~

~

t%D " ~ ...

10

o

7/ ~E 20

'0

!O

40

In this case the induction falls extremely rapidly, and in about l~OO of a second has fallen to half its original value. The subsidence of the remainder is much more gradua1. 40 ~ 1'VE /3

3D

~tlJ ~DI"/~ .div. F:l0D ~Og~ sec.

~

i:::

~~

~

==-20 ~ ...

Curre

JO

o

JO

7/ ~E ~o

4CJ

4CJ

Magnetic Viscosity

79

Soft-iron and Steel Rings

Curve 13 shows the fall of induction f0r soft-iron and steel rings of sectional diameter 0·93 cm. It will be observed that the rate of decay of the induction is much slower when the magnetic circuit is complete, as in the iren and steel rings, than in short cylinders of metal. Summary of Results (1) For fine1y-Iaminated iron, the lines of force pass out into the secondary circuit very rapidly after the magnetizing current is broken. It was experimentally shown that the iren did not take more than lO~OO of a second for the rearrangement of the molecules into their final position; so that there is no appreciable time-effect in the demagnetization of finely-Iaminated iron. (2) In solid iron cores the induction decays very slowly compared with non-magnetic metals. (3) In iron and steel the decay is very rapid at first, and then very gradual. (4) The rate of decay of induction is more rapid in a short cylinder of iron than in a ring of the same dimensions, and is more rapid for steel than for soft iron of the same diameter. (5) The decay of induction in iron is purely due to the reaction-effect of ,nduced currents in the mass of the metal, and is in no way due to any true time-effect in molecular rearrangement.

A Magnetic Detector of Electrical Waves and some of its Applications by E. RUTHERFORD, M.A., 1851 Exhibition Science Scholar, New Zealand University, Trinity College, Cambridge From the Philosophical Transactions 0/ the Royal Society, 1897, sero A, vol. 189, pp. 1-24 Communicated by Professor J. J. Thomson, F.R.S. (Received June ll-Read June 18, 1896)

INTRODUCTION

THE present paper deals with the subject of the magnetization of iron by high-frequency discharges, and the uses of magnetized steel needles for detecting and measuring currents of very great rapidity of alternation. It will be shown that these magnetic detectors oifer a very simple means of investigating many of the phenomena connected with high-frequency discharges, and may be used over a wide range of periods of alternation. Not only may these detectors be used in ordinary Leyden jar circuits, but they also oifer a sensitive means of investigating waves along wires and free vibrating circuits of short wave-Iengths. They were also found to be a sensitive means of detecting electrical radiation from Hertzian vibrators at long distances from the vibrator. In the course of the paper the following subjects are investigated:I. Magnetization of iron by high-frequency discharges and the investigation of the eifect on short steel needles. H. Magnetic detectors and their uses.

a. Detection of electro-magnetic radiation in free space. Waves were detected over half-a-mile from the vibrator. b. Waves along wires. c. Damping of oscillations. Resistance of iron wires. Absorption of energy by conductors. d. Determination of the period of Leyden jar discharges and the constants of the discharge circuit. The magnetization of steel needles when placed in a spiral through which a Leyden jar discharge was passed has long been known. In 1842 Professor HENRY was led to suspect from the anomalous magnetization of steel needles that the Leyden jar discharge was oscillatory. Professor HENRY, ABRIA, and several others, used steel needles in their

.1 :\4agnet;c Detectol'

0/

Electrical H'ares amI SOI11(, (?f its Applicatiolls

gl

attempts to determine the direction of induced currents in secondary and tertiary circuits, when the Leyden jar was discharged through the primary, but with confiicting results. Lord RAYLEIGH ('Phil. Mag.', vol. 39, 1870, p. 429) made use of steel needles in a magnetizing spiral in investigating the maximum current of a break for ordinary induction circuits. The general subject of the magnetization of iron, for rapid oscillatory currents, has been worked at by many different experimenters; Lord RA YLEIGH, using oscillatory currents of a frequency up to 1050 a second, showed that iron wires showed considerable increase of resistance, and deduced the value of the permeability of the wire. TROWBRIDGE ('Phil. Mag.', 1891) has shown that iron wires rapidly damp down the oscillations of the Leyden jar discharge, and from his results deduced a rough value for the permeability of the specimens tested. V. BJERKNES ('E1ectrician', November 18, 1892) found that the damping out of oscillations in a Hertzian resonator takes p1ace much more rapidly in a resonator of iron than when it is made of non-magnetic material.

Magnetization o/Iron by a Leydell Jal' discharge If a piece of steel wire, several centimetres in length, be taken and placed in a solenoid of a few turns, on the passage of a discharge the wire will be found to be magnetized. The magnetization is, in general, smalI, and increases slightly in amount when a succession of discharges are passed in the same direction. Fig. 1 shows the arrangement. A and Bare the poles of a Wimshurst machine or an induction coil, C the condenser, S the air-break, and D the solenoid in which the steel wire is placed. The charging-up of the condenser before the spark passes was D found to have no effect in magnetizing the needle. In all experiments to follow, the magnetization of the needles was tested by means of a small mirror magnetometer. The need1e FIg. 1 was either fixed in position in the solenoid and the magnetometer placed beside it, or the needle removed and tested after each experiment. If wires of the same length but of different diameters be taken, it will be found that the magnetization is roughly proportional to the diameter of the wires. This is to be expected, since the magnetizing forces are confined to a thin skin on the surface of the needle, and so the amount of magnetization depends more on the surface than on the sectional area. In order to determine accurately the way in which a piece of steel was

A

B

c

82

The Collected Papers

0/ Lord Rutherford

magnetized from the surface inwards, recourse was had to a method of solution of the iron in acid. The needle to be tested was fixed in a glass vessel before a dead-beat magnetometer. Dilute hot nitric acid was poured in and kept at a constant temperature. As so on as the needle was covered it commenced to dissolve, and the variation of the deflection with the time was noted. In this way the amount and stages of the magnetization of the iron could be completely determined. From preliminary experiments on uniformly magnetized needles, it was found that under the action of the acid the diameter of the wire decreased uniformly with the time. Let I represent the intensity of magnetization of a thin circular shell distant r from the centre of the needle, and M the deflection of the magnetometer at any instant. r

LI. 27fr dr is proportional to M, therefore

Ir is proportional to dM/ dr.

Let r be the radius of the wire at first. It has been shown that a - r is proportional to t, the time of action of the acid. Therefore a - r = Kt where K is a constant, and therefore

- dr = Kdt,

I varies

1

dM

a - Kt' dt'

If a curve be plotted whose ordinates represent the deflection and the abscissre the time of action of the acid, dM/dt at any point is equal to the tangent of the angle which the tangent to the curve at that point makes with the axis of x. The variation of I can thus be completely determined from the experimental curve. The following curve (Curve I) is an example of the magnetization of a piece of pianoforte wire, 4 centims. long, O· 08 centim. in diameter, placed in a solenoid of two turns per centim. The frequency of the discharge was about 3 million per second. The ordinates of the curve represent the deflection of the magnetometer and the abscissre the depth to which the iron has been dissolved by the acid, measuring from the surface inwards. Each division of the ordinates corresponds to a depth - O· 00057 centim. The deflection of the magnetometer at first was 85 divisions. As the needle commenced to dissolve, the deflection increased rapidly to 156, remained nearly steady for a short time, and then rapidly diminished to zero; when this was the case the diameter of the needle was 0·032 centim., so that the magnetization had penetrated to a distance of O· 024 centim.

A Magnetic Defecfor

0/ Electrical

Waves and some ofits Applications

83

If the variation of the value of I, the intensity of magnetization from the surface inwards, be deduced from this curve, it will be seen that the surface layer is magnetized in one direction and an interior layer in the opposite direction. This apparently gives evidence of only two half oscillations, in opposite directions, in the discharge. A large number of needles were dissolved after magnetization under various conditions, and the same peculiarity was observed, although, from other evidence, it was known that there were a large number of vigorous oscillations before the discharge was much damped down.

Oef!ecti'pn.

Curve(/).

OeptJ, oFP~etrt1tiDn Each div!l= -00@57cln6,

When a needle magnetized to saturation was subjected to the discharge, the magnetization of the needle was always diminished, and on solution of the iron the same effect was observed, viz. a surface layer magnetized in the opposite direction to the internal magnetism. Since a Leyden jar discharge in general gives several complete oscillations before it is greatly damped down, it would be expected that the surface layer of a uniformly magnetized needle would either be completely demagnetized or show evidence of several oscillations in opposite directions. The effect observed may be explained when the demagnetizing force of the ends of a short needle on itself is taken into account. The first half oscillation that tends to demagnetize the needle has the demagnetizing force of the needle assisting it, while the return oscillation has it in opposition. The return oscillation will therefore not be able to completely remagnetize the surface layer already affected, but a thin layer will be left in the interior. After the passage of the next oscillation another layer will be added in the same direction, and so on, till the final effect will be that the surface of the needle will be magnetized in the opposite direction to the interior.

The Collected Papers 0/ Lord Ruther/ord

84

If strongly-magnetized needles of the same diameter, but of different lengths, are taken and placed in the same solenoid, it will be found that the reduction of magnetic moment of the needle, due to the discharge, is greater the shorter the needle. This effect is due to the demagnetizing influence of the ends, which is greater the shorter the needle. It was also found that if successive discharges be passed, the reduction of deflection gradually increases, till it reaches a steady state, so that the passage of any further number of discharges has no apparent effect on the magnetism of the needle. The following table shows the effect of varying the length of the needle, the diameter being kept constant, and also the effect of successive discharges in each case. Needle 0·08 centim. in diameter; frequency about 3 millions. Numberof discharges 10· 5 centims. 6· 4 centims.

0 1 2 5 10 20 50

250 204 199 195 190 189 188

250 190 182 175 170 166 162

3· 2 centims.

1· 6 centims.

250 166 155 138 130 125 120

250 150 135 115 107 102 98

0, 7S

centim'l

250 114 88 64 57 54 50

,

I

In the above table each of the needles was placed at such a distance from the magnetometer to give the same steady deflection of 250. The vertical columns show how the deflection fell after the passage of the different numbers of discharges. The vertical columns correspond to needles 10· 5, 6,4, 3'2, 1,6,0,75 centim. respectively. For the needle 10·5 centims. 10ng, the deflection fell from 250 to 188, while for the short needle, 0·75 centim. 10ng, the deflection fell from 250 to 50, although all the other conditions were precisely the same for each. It will be observed from the above that the first discharge is mainly instrumental in reducing the deflection, and that after ten discharges have been passed, the deflection has nearly reached its final value. Whenever a magnetized needle is placed in a solenoid and a discharge passed, there is always a reduction of the magnetization, the amount depending, for any given size of needle, on the intensity of the magnetic force in the solenoid and on the period of the oscillation. This is the case whether we are dealing with Leyden jar circuits, or the free vibrations, such as are set up in Hertzian receivers. An unmagnetized needle, on the other hand, is not appreciably magnetized

A Magnelic Detector 01 Electl'ical Wal'es ami some of its Applicalions 85

when placed in a circuit where the damping is small, for each successive oscillation destroys the effect of the previous one. Soft iron wires exhibit a similar eft"ect to steel, only it is difficult to use wires of sufficient length to retain their magnetization. The effect on the needle is in general a purely surface one, and the amount of demagnetization does not bear any simple relation to the magnetizing force acting on it. After every experiment, the needle was removed and placed in a solenoid and a steady current passed, sufficient to saturate the steel. In this way the needle could always be quickly reduced to the same state after any experiment, and it was found that, using hard steel wires, the results of successive experiments were very consistent with one another. Passage

0/ a

Discharge Longitudinally through a Magnetized Wire

If a piece of pianoforte wire, several centimetres in length, be taken and placed in series with the discharge circuit, in the passage of a discharge, the magnetic moment of the needle is diminished, due to the 'circular' magnetizati on of the wire. If the needle be dissolved in acid, it will be found that there is a thin skin, apparently magnetized in opposition to the original magnetization, due to the resultant action of the demagnetizing force of the needle and the magnetic force due to the current in the wire. The magnetic force H acting at any given point in the wire is given by H = 2y, r

where y is the current through the conductor flowing internal to the circ1e described through the point, and concentric with the surface ofthe conductor. The value at the surface of the wire is given by H = 2" ,

a where a is the radius of the wire. Assuming p., the perrneability of the iron wire, as constant, the maximum value of the current at any point of the conductor decreases in geometrical progression as the distance from the surface inwards increases in arithmetical progression. As will be shown later in the part on 'Resistance ofIron Wires', the current falls off even more rapidly than the theoreticallaw, on account of the increase of the value of p. as the amplitude of the current diminishes in intensity from the surface inwards. For thin wires the magnetic force at the surface of the wire is much greater than for thicker ones. We should, therefore, expect a thick magnetized wire conveying the current to be affected to less depth than a thin one, and this is found to be the case. A thin steel wire, 0·025 centim. in diameter, was completely demagnetized by a discharge. In this particular case the maximum value of the current

86

The Collected Papers of Lord Rutherford

through the wire was about 100 amperes, and the value of the magnetic force at the surface of the wire was, therefore, about 1600 C.G.S. units. A hard steel wire, O· 08 centim. in diameter, was only partially demagnetized, the deflection being reduced from 250 to 116 scale divisions. The following are examples of a few of the experiments on the demagnetization of iron wires when the frequency of the discharge was about 3 million and the value ofthe maximum current about 100 amperes:I. Thin . diameter,' 0 025' . d. Wlre centIm. : compIeteIy demagnetIze Thin softiron} steel 2. Steel wire: diameter, O· 08 centim. : fall of deflection from 250 to 116. 3. Steelwire: ,,0·16centim.: " " 250to184. 4. Steel wire: " O' 25 centim. : " " 250 to 216. 5. Longhollowsoft iron cylinder, imillim. thickand diameter 1· 8 millims. *: fall of defiection from 250 to 230. The same condenser and discharging current were used for all the specimens tested, and it is of interest to observe the depth to which the magnetism of the iron was affected by the discharge, assuming that the final defiection is that due to the mass of iron not circularly magnetized. Wire

Diameter centim.

Hard steel wire Soft steel wire Soft steel Soft iron cylinder

0·08 0·16 0·25 1·8

Depth of penetration of the discharge centim.

0·013 0·011 0·009 0·0011

Experiments of this kind show to what a small depth the current penetrates into the iron wire. Very !arge momentary currents are conveyed througb a surface skin of the conductor, and the intensity of the current diminishes rapidly inwards. A thin magnetized steel wire was placed in the circuit of a small Hertzian plate vibrator. The deflection due to the needle fell from 300 to 250 after a succession of discharges. This shows that the iron was unaffected below a depth of about 0·0011 centim. • Editor's Footnote: Rutherford was a somewhat indifferent proof-reader. There are many instances in bis papers of inconsistencies in nomenc1ature, in abbreviations, and some of arithmetical mistakes, as weil as obvious misprints. In most of these cases, the original form will be retained and a footnote will be added to indicate the correction; in the more trivial cases the correction will be made in the text. Correction: There can be little doubt that the dimensions of the cylinder referred to above should be 'centim.', not 'mi1lim.'

A Magnetic Detector of Electrical Waves and some

0/ its Applications

87

For rough comparisons of the intensity of currents in multiple circuits, the use of the 'longitudinal' detector is often preferable to placing the needle in a solenoid. A thin magnetized steel wire, placed in series with the circuit, is a surprisingly sensitive detector of oscillatory currents of small intensity. In practice copper wires were soldered on to the extremities of the steel needle, which is placed in position before a magnetometer. A magnetizing solenoid is wound over the needle, and after every experiment a steady current was sent through in order to re-saturate the needle. Both the 'longitudinal' and 'solenoidal' detectors may be very readily used to compare the intensities of currents in multiple circuits when the period of oscillation is the same for each. The best form of the solenoidal detector is explained later, and it has the advantage of being able to distinguish between the intensity of the first and second half oscillations. Detection

0/ Waves in Free Space

It has been shown that the amount of demagnetization of a magnetized

needle depends on the fineness of the wire and the number of turns per centim. on the magnetizing solenoid. If a short piece of thin magnetized steel wire be taken, and a large number of turns wound over it, it is a very sensitive means of detecting electrical oscillations in a conductor when the amplitude of the oscillations is extremely smalI. It was on this principle that adetector for electrical waves was devised, which proved to be a sensitive means of detecting Hertzian waves at considerable distances from the vibrator. About twenty pieces of fine steel wire 0·007 centim. in diameter, each about 1 centim. long, and insulated from each other by shellac varnish, formed the detector needle. A fine wire solenoid was wound directly over it, of two layers corresponding to about 80 turns per centim. As the solenoid was ofvery small diameter, about 15 centim. ofwire served to wind the coil. This small detector was fixed at the end of a glass tube, which was itself fixed on to a wo oden base, the terminals of the detector coil being brought out to mercury cups.

D D

DD

D

Fig. 2

S (fig. 2) represents the detector ne edle and the solenoid wound over it. A and Bare the mercury cups. CA and BD were two straight rods which served as receivers, one end of each being placed in the mercury cups.

88

The Collected Papers o[ Lord Rutherford

The detector needle was strongly magnetized and placed before a small magnetometer, the deflection due to the needle being compensated by an auxiliary magnet. Ifthe receivingwires were parallel to the electric force ofthe wave from the vibrator, oscillations were set up in the receiver circuit, the surface layers of the needles were demagnetized, and there resulted a corresponding deflection of the magnetometer needle. The amount of the deflection, of course, depended on the amplitude of the oscillations set up in the receiver, and, therefore, on the distance from the vibrator. Long Distance Experiments

When a Hertzian vibrator was used with plates 40 centims. square, and a short discharge circuit, quite a large deflection was obtained at a distance of 40 yards, the waves passing through several thick walls between the vibrator and receiver. Further experiments were made to see how far from the vibrator electromagnetic radiation could be detected. For the long distance experiments, the vibrator consisted of two zinc plates, 6 feet by 3 feet, and separated by a short discharge circuit of about 30 centims. When large plates were used, a Wimshurst machine was equally efficient as a Ruhmkorff coil for exciting the vibrations. The first experiments were made over Jesus Common, Cambridge, the receiver being placed in one of the buildings on Park Parade. Quite a large effect was obtained at a distance of a quarter of a mile from the vibrator, and from the deflection obtained it was probable that an effect would have been got for several times that distance. When the vibrator was set up in the top floor of the Cavendish Laboratory, a small, but quite marked effect was obtained at Park Parade, a distance of over half a mile in the direct line. In this case the waves, before they reached the receiver, must have passed through several brick and stone walls, and many large blocks of buildings intervened between the vibrator and receiver. The length of wave given out by the vibrator was probably six or seven metres, and a wave of that length seemed to suffer very Httle loss of intensity in passing through ordinary brick walls. From an experiment tried in the Cavendish Laboratory it was found that the effect of six solid walls and other obstacles between the vibrator and receiver did not diminish the effect appreciably. When the vibrator was working in the upper part of the Laboratory, a large effect could be obtained all over the building, notwithstanding the floors and walls intervening. A large number of experiments were made on the effect of varying the length and diameter of the receiving wires. If a fairly dead-beat vibrator were

A ,'tttagnC'lic' Detertol'

(~r Elerll'iral

Wa\'es amI sO/ne

oI its Applicatiolls

89

used, e.g., plates with a short inductance, it was found that the deflection gradually increased with increase of length of the receiving wires, reaching a maximum which was unchanged by any further increase of length. The effect on the detector was found to be practically independent of the sectional area of the receiving wires. A thin wire and a thick rod of the same length had equal effects; a plate of metal, 6 eentims. wide, produeed the same deflection as a thin wire. I f two wires instead of one were used in parallel the effect was the same as one, though the wires were some distance apart. Any number of wires in parallel had the same effect as a single wire or plate. No difference could be detected whether the first half oscillation in the receiver tended to magnetize the needle or the reverse. Since the vibrator used was nearly dead-beat, this shows that the damping of the oscillations in the receiver is very sman. On introducing a short earbon rod in the circuit the deflection was greatly reduced. It was found impossible to magnetize soft iron or steel when placed in the reeeiving circuit on account of the slow decay of the amplitude of the oscillations. The detector needle may be kept in position for a succession of observations, provided the current in the receiving circuit is steadily increasing for each experiment, otherwise the detector should be remagnetized and placed in position again after each observation. The deflection was found to be very constant for aseries of experiments under the same conditions. The conneetion between the intensity of tbe electric force at the receiver and tbe deflection of tbe magnetometer needle can be easily determined by swinging the reeeiving wires through different angles. Wben tbe receiver is plaeed symmetricaIly with regard to the vibrator, the deflection was a maximum when tbe receiving wires were parallel to the axis ofthe vibrator, and tbe intensity ofthe eleetric force acting along the receiver varies as the eosine of the angle from the maximum position. With plate vibrators the deflection was found to be nearly independent of the degree of brightness of the spark terminals and remained sensibly constant for long intervals. In the case of tbe small cylindrieal vibrator used by HERTZ with tbe parabolic reflectors, the deftection eontinually varied with the state of the sparking terminals, and such small vibrators cannot be relied on for metrical experiments. Some experiments were made to see if tbe magnetic force in tbe wave front could be directly detected. A collection of thin wires, insulated from eacb other and magnetized to saturation, were used and placed in the direction of the magnetic displacement, but the values of the magnetic force were too small to be observed, except quite c10se to the vibrator.

Waves along Wires It was found that tbe use of a detector, composed of fine insulated wires, was quite delicate enough to investigate waves along wires when there was only

90

The Collected Papers

0/ Lord Ruther/ord

one turn of wire round the detector needle. Since the reduction of magnetic moment was nearly proportional to the amplitude of the current, the intensity of currents at various points could be very approximately compared.

A

A' C

Coil.

D

B

B'

E

F

Fig. 3

The ordinary Hertz arrangement (fig. 3) was set up for obtaining free vibrating circuits. A and B were two plates set up vertically. Beside them were two small plates A' and B', and long wires A'E, B'F were led from these plates. A fixed bridge was placed at EF, the ends of the wires, and adetector placed at the middle point of EF, with a small magnetometer fixed in position. A sliding bridge, CD, was then moved till the fall of the deftection of the detector needle was a maximum. This position of the bridge could be very accurately determined, for a movement of the bridge through 1 centim. altered the deflection considerably. The detector was then placed in various parts of the circuit CE, and the amplitude of the current at the different points determined. It was found that the current was a maximum at C and the middle point of EF. A well-defined node was found at the middle point ofCE. The length CEFD was thus half a wave-Iength. Since the use of the bolometer has been the only means of accurately investigating waves along wires, it was interesting to observe whether the magnetic detectors were of the same order of sensitiveness as the bolometer. One turn of wire was wound round each of two glass tubes, sliding along the wires CE and DF, as in REUBEN'S experiments. Instead of the fine bolometer wire, adetector needle, with several turns of wire around it, was placed in series with the two turns of wire. The charging and discharging of the small condensers, formed by the straight wires and the small coils around them, was quite sufficient to almost completely demagnetize the needle. By this method the movement of extremely small quantities of electricity could be detected and the sensitiveness was quite comparable with that of a delicate bolometer. No appreciable damping could be detected for the long wire circuits, showing that they were probably vibrating almost independently of the primary vibrator. If a metre or two of wire was fixed to the pole of a Wimshurst machine, on

A Magnetic Detectol' ofElectrical Waves and some ofits Applications

91

the passage of a spark, there was evidence of a rapidly-oscillating current set up in the wire. By using a sensitive detector, with a few turns round it, the variation of the current along the wire could readily be determined. If short lengths of wire were fixed to any portion of a Leyden jar circuit, on the passage of a discharge, there was always evidence of a rapid oscillation set up in the wire. Each of the short circuits had a tendency to vibrate in its own natural period, but the results were complicated by the oscillations of the main circuit. Damping of Oscillations

The use of magnetized needles offers a simple and ready means of determining the damping of oscillations in a discharge circuit. Let L be self-inductance of discharge circuit for rapid currents. " C = capacity of condenser. " R = resistance of leads and airbreak to the discharge. " V0 = potential to which condenser is charged. The current y at any instant is given by y =

CVo e-R/2L.t sin _1_. (LC)t (LC)

The exponential term only includes the case of frictional dissipation of energy, and does not take into account radiation into space. In the experiments at present considered, where the condenser is of the type of a Leyden jar, there can be but very small amount of dissipation of energy due to radiation. Assuming R to be constant, the amplitude of the current decays in geometrical progression. Consider two similar small oppositely wound solenoids A and B placed in series in the discharge circuit. Two magnetized needles are placed in A and B, the north poles facing in the same direction. After the passage of a discharge, it will be found that the reduction of magnetic moment is not the same in the two needles. Let (Xl (X2(X] ••• be the half-oscillations of the discharge in one direction. " ßIß2ß3 ... be the half-oscillations in the opposite direction. Suppose that the half-oscillation (Xl tends to magnetize the needle in the solenoid A still further. Since the needle is saturated no effect is produced. ßl demagnetizes the surface skin, (X2 tends to remove the effect of ß2. and so on. In the solenoid B (X 1 demagnetizes the needle, ß1 tends to remagnetize it in its original direction, and so on. Since the maximum value of the current of (Xl is greater than the maximum value of ßh the needle in B will be more demagnetized than in A. If, however, we increase the number of turns per centimetre on the solenoid A, until the effects on the two needles are exactly the same, then

The Collected Papers 0/ Lord Ruther/ord

92

assuming that the value of the current decreases in geometrical progression, the maximum value of the magnetic force due to the oscillation ßI acting on the needle A is equal to the maximum value due to the oscillation (Xl on ß. Let YIY2 be the maximum values of the current in the first and second half-oscillations respectively. Let nln2 be the number of turns per centimetre on solenoids A and B respectively. Then, since 47TIlIYl = 47Tn2Y2,

Y2!Yl = nt!1l2

the ratio of the second to the first half-oscillation is therefore known, and the damping is thus determined. The actual resistance in the circuit mayaIso be deduced. Now YI = pCVoe-R/2L.T/4 where T = period of complete oscillation and 1 p = (LC)t' Y2

= pCVoe- R!2L.3T/4.

Therefore e·-R/2L.T/2

=

Y2 Yl

=

Pb say.

Therefore R T log PI = - 2L . 2

(1).

Since Land T are known from the constants of the discharge circuit, and

plis detennined by experiment, R is known.

In practice, in order to avoid the necessity of determining the constants of the discharge circuit, an additional known resistance r is introduced into the circuit. If an electrolytie resistanee of zine sulphate with zine eleetrodes be used, the resistance will be found to be practically the same for steady as rapidly alternating currents, as the specific resistance is very great. Let P2 be the ratio of the amplitudes of the two half oscillations when R + r is in the circuit R+r T (2). log P2 = - ~ . -2 Dividing (2) by (1)

+

R r log P2 --=--, R log PI R is therefore determined in tenns of r, a known resistance. The method of two solenoids was not adopted in practice, but one theoretically equivalent employed.

A

Ma~ll('/i('

Detector 0/ Electrical Wal'es ami some

0/ its ApplicatiollS 93

A narrow piece of sheet zinc ABC was taken (fig. 4) and bent into abnost a complete cirele of 7 centims. diameter. This was fixed on a block of ebonite. At the centre of the cirele a thin glass tube OM was placed, which served as the axis of a metal arm LM, which pressed against the circumference of the circle and could be moved round it. The 'detector' consisted of about thirty very fine steel wires, 0·003 inch in diameter, arranged into a compound magnet about 1 centim. long. The wires were insulated from each other by shellac varnish, and the small needle was fixed inside a thin glass tube which could be easily slipped in and out of the central glass tube OM.

M

CA

B

o

M Fig. 4

A divided scale was placed round the circumference ABC, and the whole arrangement was fixed in position before a smaH mirror magnetometer. The magnetized 'detector' needle was placed in position by sliding it in the glass tube, and the deflection due to the needle was compensated by another magnet. The wires of the discharge circuit were conneeted to C and M, and when the arm ML was at C no effeet was produced on the needle. When the discharge passed round the cirele there was adefleetion due to the partial demagnetization of the deteetor. The detector was then quickly removed and magnetized to saturation in an adjacent solenoid and then replaced. It was found that, provided the deteetor was magnetized in a very strong field, on replacing it in position the zero remained unchanged, and the same deflection was obtained time after time for similar discharges. Since the magnetic field at the centre of a circle due to an arc of length 1 is given by H

=

l~, where

y is the current,

I

we see that the magnetie force acting on the needle is proportional to the length of the are traversed by the diseharge. Aseries of observations were made and it was found that the deflection due to the detector was approximately proportional to the magnetic force acting on the needle, provided the magnetic force was weH below the value required to completely demagnetize the steel.

The Collected Papers 0/ Lord Rutherford

94

Curve (2) represents the relation between the defiection of the magnetometer and the magnetic force acting on the needle. The curve is nearly a straight line except near the top part of the curve. To determine the damping of the oscillations a discharge was passed in one direction and the defiection noted. The detector was removed, magnetized and replaced. The direction of the discharge was reversed and the arm of the circ1e moved until the defiection was the same as before. When this is the case

Oel!ectid(,.

Cur e(2). Curre(2).

Cur e(2). Cur e(2). Curre(2). Curre(2). on Damping of Oscillations. Diseharge cireuit rectangular, 184 centims. by 90 centims.; Self inductanee of circuit, L = 7,400; Capacity, C = 2,000 eleetrostatic units; Frequency, 1·25 millions per seeond.

EXPERIMENTS

Length of spark gap centim.

0·06 0·12 0·24 0·37 0·49 0·61

Ratio of amplitudes of two first half oscillations

0·98 0·97 0·93 0·9 0·79 0·7

Resistance

ohms

-

1·1 2·6 3·7 8·4 12·4

A M agnetic Detector of Electrical Wal'es and some of its Applications

95

the ratio of the maximum values of the first and seeond half oseillation is given by the ratios of the ares traversed by the diseharge. In this way the rate of deeay of oseillations in ordinary diseharge cireuits was examined. With short air breaks and eopper wires for conneetion, it was found that the damping was hardly appreciable. As the length of the spark gap was increased, the absorption of energy in the air-break caused the oscillations to damp rapidly. If the copper wires of the discharge circuit were replaced by iron wires, there was in all cases a very rapid decay of the oscillations, whatever the length of the air-break. If an iron cylinder were placed in a solenoid, the absorption of energy by the cylinder caused a rapid damping, while a copper cylinder of the same diameter had no appreciable effect. In the third column the apparent resistance, corresponding to the absorption of energy in the conducting wires and spark gap is tabulated. The calculated value of the resistance of the wires of the discharge circuit was 0·4 ohm, so that the remainder of the resistance is due to the great absorption of energy in the air break.

Oel!ectid(,.

CurJke(3).

Cur e(2). Cur e(2). Curre(2). Curre(2). The above curve represents the relation between the length of the spark and the apparent resistance that the spark offers to the discharge. The ohmic resistance of the air break is probably very variable, depending on the intensity of the charge at any instant, but the absorption of energy is quite definite and may be expressed in terms of the non-inductive resistance whieh, when placed in the circuit, would absorb the same amount of energy. It will be observed that the damping of the oseillations increases rapidly with the length of the spark, and that the resistance of the air break increases very rapidly with its length. lt was also found that the damping depended on the capacity when the inductance and spark length were kept constant. The damping and also the resistance of the spark were found to increase with increase of capacity. For

96

The Collected Papers

0/ Lord Rutherford

example with an air break of O· 32 centim. the damping and resistance are given below. Capacity

Ratio of Oscillations

Resistance

1000 2000 4000

0·94 0·9 0'81

2·2 2·6 3·8

When the capacity of the circuit was small the damping was found to be very small. If iron wires were put into the place of the copper wires in the discharge circuit, the damping was found to be great for all capacities investigated. When the capacity of the circuit was only 130 electrostatic units, and inductance 2400, no appreciable damping was found for an air break O' 5 centim. : When the copper wire was replaced by an iron one of the same dimensions, the second half-oscillation was only 0·6 of the amplitude of the first. Resistance of Iron Wires for High Frequenc)' Discharges

The rapid decay of the oscillations when iron wires formed the discharge circuit has been already noted. This has been observed by TROWBRIDGE ('Phil. Mag.', December, 1891), who found, by photographing the spark, that there was evidence of much fewer oscillations when iron wires were used __ instead of copper. For very rapid oscillations the resistance R' is given by R' = V'tpp.1R (see Lord RAYLEIGH, 'On the Self-Induction and Resistance of Straight Conductors', 'Phil. Mag.', 1886), where R is the resistance of the wire for steady currents, I the length of the wire, p. the permeability, and p = 2'Tf'n, where n is the number of oscillations per second. Since the expression involves p., we should expect the resistance to be much greater for iron wires than for wires of the same conductivity, but nonmagnetic. To determine the resistance of iron wires a very simple method was used. The fall of deßection, due to the detector needle, arranged as in (fig. 2) was noted. The iron wire was then removed and a copper one of the same diameter substituted. Since the inductance of the circuit was practically unchanged, if the damping in the two circuits are equal, the resistances should be the same. A short piece of high-resistance platinoid wire was introduced into the circuit of the copper wire, and the length adjusted until the deßection was the same as in the first case. When this is so, the resistance of the platinoid wire, together with the resistance of the copper wire, is equal to the resistance of the iron wires for the frequency employed.

RUTHERFORD'S MAGNETIC DETECTOR FOR ELECTRIC WAVES, (Cavendish Laboratory). See 'A Magnetic Detector .. .' (p. 80)

1896

This page intentionally left blank

A Magnetic Deteetor vI Eleetrical Wal'es

Ulld

svme oI its Applieations

97

The resistance of the copper wires was calculated for the frequency used, but was, in general, small compared with the resistance introduced. The resistance of the platinoid wire was also calculated, but was found to be practically the same as for steady currents. The length of wire placed in the current was 265 centims.; spark length was O' 25 centim. Kind of wire

Diameter

R

R)

Rt/R

I

I

Soft iron

Pianoforte steel wire Nickel wire..

centims.

0·025 0·047 0 . 094 0·295 0·062 0·062

6· I 2·62 O' 57 0·051 I· 51 0·66

11 ·8 12·8 9 .2 4·2 0·66 5·9

I ·9 4·9 16 72 6·5 5·9

In the above Table* R is the resistance for steady currents, R 1 the resistance for a frequency I . 6 millions per second. The last column gives the ratio R I IR for the different wires. In the case of the wire O' 295 centirn. in diameter, the resistance is 82 times the resistance for steady currents, while in the case of the wire ofO'025 centim. diameter, it is only 1·9 times. If the value of fL, the permeabi1ity of the specimens of soft iron, be calculated from the formula R' = V!PfL1R, it will be found that the value varies with the diameter of the wire. Diameter centims.

0·025 0·042 0·094 0·295

I'

3'5 9·4 18 53

The curve below (Curve 4) shows the relation between the diameter of the wire and the permeability for the discharge. It will be seen that fL varies approximately as the radius of the wire. The very small value of the permeability for fine wires is to be expected when we consider the very large currents that pass through the wire, and the • Editor's Footnote: Correction: In the above table the last three values in column Rl!R should read: 82,7'5,5'6. D

98

The Collected Papers of Lord Rutherford

consequent large value of the magnetic force that acts at the surface of the iron. The maximum current of discharge, assuming the damping to be small, is given by y = pCVo, where C is capacity and Vo the potential to which condenser is charged.

Oel!ectid(,.

CurWe(4).

Cur e(2). Cur e(2). Curre(2). Curre(2). In the above experiments the air-break was about -110 inch, and the difference of potential about 10,000 volts, and since p = 5.106, C = 4000, the maximum current was about 222 amperes. The magnetic force at the surface of the wire of radius r, which conveys the current, is given by H

= 2')' = 3552 C.G.S. units, if r = 0·0125 centim. r

Ifwe assume the value ofB to be ab out 14,000, we see that the permeability for the extreme surface layer would be about 4. The value of the magnetizing force diminishes from the surface inwards, so that the mean permeability of the iron to the discharge should be greater than the value at the surface. These considerations show that the permeability of iron to these discharges is by no means constant, but depends on the diameter of the wire and the intensity of the discharge. The resistance of iron wires was found to vary with the length of the spark. Short air-breaks gave higher values of the resistance than long ones. When the length of the spark was so adjusted that the maximum current was approximately constant for different periods, the resistance was found to vary as the square root of the frequency, as we should expect from theory. Several specimens of pianoforte steel wire were examined to see whether the larger waste of energy due to hysteresis in steel materially affected the

A Magnetic Detector

0/ Eler/riral Waves and some 0/ its Applications

99

value of the resistance, but the increase of resistance was not so great as for soft iron wires of the same diameter, although the loss due to hysteresis in steel is much greater than in soft iron for slow cycles. Absorption

0/ Energy by Metal Cylinders

This subject has been treated mathematically and experimentally by J. J. THoMsoN ('Recent Researches' , pp. 321-326). It is there shown by observing the electrodeless discharge that a cylinder of iron placed in a solenoid absorbs considerably more energy than a copper one of the same dimensions. The method adopted here admitted of quantitative as weIl as qualitative results. An ordinary Leyden jar was discharged through a solenoid of about thirty turns and 14 centims. long. The metal cylinder was then placed in the solenoid and the damping of the oscillations observed. The cylinder was then removed and a non-inductive resistance added until the damping was the same as when the metal cylinder was in the solenoid. The absorption of energy in the cylinder was then equal to the absorption of energy in the added resistance, whose value was known. In the above we have taken no account of the change of inductance of the circuit due to the metal cylinder being placed in the solenoid. The change is small, and could be made negligible by making the inductance of the solenoid small compared with the rest of the circuit. (1) A test-tube was filled with finely laminated iron wire, 0·008 inch in diameter. The test-tube was filled with paraffin oil, to insure insulation from eddy currents. The absorption of energy in this case corresponded to an added resistance of 10·25 ohms to the circuit. (2) A test-tube filled with steel filings and insulated as in (1). Increase of resistance, 9 ohms. (3) A thin soft iron cylinder, 1· 9 centims. in diameter. Increase of resistance, 3· 9 ohms. (4) Solid iron rod. Increase of resistance, 3· 3 ohms. (5) A copper cylinder, a test-tube filled with a copper sulphate solution, and a platinum cylinder showed no appreciable absorption of energy. (6) A carbon rod absorbed a large amount ofenergy. Increase ofresistance, 3·3 ohms.

The frequency of the oscillations in the above experiment was two million per second. If the experimental value obtained for the increase of resistance due to the solid iron cylinder be compared with the theoretical value ('Recent Researches', p. 323), the value of the permeability will be found to be 172, which accounts for the much greater absorption by an iron cylinder than a copper one.

100

The Collected Papers of Lord Rutherford

TADLE of Absorption of Energy of various Conductors; Absorption of Energy expressed in Terms of the Increased Resistance of the Discharge Circuit. Substance

Increase of resistance

Laminated soft iron wires Solid soft iron cylinder Hollow iron cylinder Carbon cylinder .. Copper, platinum, zinc cylinders Steel filings

10·25 ohms 3·5 3·9 3·3 not appreciable 9 ohms

From the peculiar deadened sound of the spark, it could always be told when much energy was being absorbed in the discharge circuit. With copper wires for the discharge circuit, the spark was sharp and bright; when iron wires were substituted, the spark was weak; when an iron cylinder was put in the place of a copper one, the spark was neither so bright, nor so sharp in sound.

Determination 0/ the Period 0/ a Discharge Circuit It is often a difficult matter to obtain even an approximation of the period of

oscillation of a discharge circuit when the capacity of the condenser and the self-inductance of the circuit cannot be directly calculated. The following simple method was found to work very accurately in practice, and could be used for a fairly wide range of frequencies. Let ACB, ADB (fig. 5) be two branches of a discharge circuit in parallel, Rand L the resistance and inductance of the branch ACB, Sand N the resistance and inductance of branch ADB.

C C

C C Fig. 5.

Let M be the coefficient of mutual inductance between the two branches. Let x and y be currents in branches ACB, ADB respectively.

A Magnetic Detector of Elee/fical Wa~'es and some

0/ its App/icalions

101

It is shown ('Recent Researches', p. 513) that for a rapidly-alternating current of frequency n, where p = 27Tn, that

}! cos(pt+e)=Acos(pt+e),say }t + +

x= { - S2 + (N - M)2P 2 2 (L+ N -2M)2+(R+S) Y= {

tan

= e

•

tan

cos(pt

e /) = Bcos(pt

E'),

p{R(N - M) - S(L - M)} S(R + S) + (L + N - 2M)(N - M)p2'

e' _ _

-

+

(L - M)2 2 R2 2 P 2 (L + N - 2M) + (R + S)

R(R

p{R(N - M) - S(L - M)} (L + N - 2M)(L - M)p2'

+ S) +

A and Bare the maximum currents in the two branches ACB, ADB respectively, and ~ = /S2 + (N - M)2p2. B \f R 2 + (L _ M)2 p2 lf the circuits be so adjusted that A = B

R2 and

+ (L -

Mf p 2

=

S2

+ (N -

M)2 p 2,

, R2 - S2 p- = N2 - L2 - 2M (N - L)

The value of the impedance VR 2 + p2L 2 is nearly independent of R for rapid frequencies in ordinary copper wire circuits Suppose L = 104, 11 = 106 , p = 211' . 106 , and then p2L2 = 411'2 . 1020. lf the value of R for the particular period was 2 ohms say, then

R2 p2L2

10011'2

a very small quantity. We therefore see that, under ordinary circumstances, the resistances may be neglected in comparison with the inductances. lf the branch ACB of the divided circuit consist of a high resistance of short length, and consequently small inductance, and the other branch of an inductance N, and the maximum values ofthe currents in the two branches 2 2R 2' since S2 may be neglected in comparison with N -L R2, and supposing the value of M to be small compared with N.

are equal, then p2 =

102

The Collected Papers o[ Lord Ruther[ord

If the inductances are unchanged, N2 - L2 is practically a constant for all periods, and p is therefore proportional to R. In practice, one branch of the divided circuit consisted of a standard inductance, N, and the other branch of an electrolytic resistance, R. The equality of currents in the two circuits was obtained by altering the value of R until the effect on the detector needles was the same for both circuits. Since the effect on the needle was the same in both circuits, the maximum values of the current are the same, since each branch is traversed by an oscillation of the same period.

D

A :C

B

P Q

x

~ ~

101:

~I;;:

~

S

~

d

Fig.6 It was not found necessary to place the divided circuit in series with the discharge circuit, but the arrangement was more satisfactory when it was shunted off a portion, PQ, of the discharge circuit XPQS (fig. 6), whose period is to be determined. The addition of the shunt circuit had no appreciable effect on the period of the oscillation for the equivalent inductance of the two branches QP, QBAP, was slightly less than that of QP, and the length of QP was generally not a tenth part of the whole discharge circuit. ACB, ADB are the branches of the divided circuit. In C was placed a resistance consisting of zinc sulphate with zinc electrodes. The amount of resistance in the current could be varied by altering the length of electrolyte through which the current passed. In D was placed a standard inductance consisting of six turns of insulated wire wound on a bobbin 10 centims. in diameter. The self-inductance of this could be accurately determined by calculation, and was very approximately the same for steady as for rapidly changing fields. If the inductance L of the resistance branch was small compared with N, the value of pis given by p = R/N.

Even if L

=

~ the correction would only be 1 per cent. The resistance R

of the zinc sulphate solution was determined for steady currents, and it can be shown that the change of resistance due to the concentration of the currents on the surface is quite inappreciable for the periods investigated on account of the high specific resistance of the solution. This can experimentally be shown as folIows: a tube containing the solution to be tested is placed inside a solenoid of a few turns, and adetector needle placed in the solution. After the passage of a discharge it will be found

A Magnefic Defeefor o{ Eleetrical Wal'es and some of its Applications 103 that the effcct on the ne edle is the same as when the solution is removed, showing that there is no screening action on the needle due to the solution. Since the law of decrease of magnetic force from the surface inwards is thc same as for the decrease of amplitude of a current through the conductor, it follows that the amplitude of the current at the centre of the solution was the same as at the surface, and that there was no alteration of the resistance of the electrolyte due to concentration of the current on the surface. An air condenser of calculable capacity C was discharged through a circuit whose inductance L for rapid frequencies could be very approximately determined. The value of p

1 obtained from theory was found to agree to VLC within 3 per cent. of the experimentally-determined value, and, from the difficulty of accurately calculating the inductance, it is probable that the experimental determination is nearer the true value. From the dose agreement of theory and experiment, we have indirect1y proved that the resistance of an e1ectrolyte like zinc sulphate is the same for high frequencies as for low. As an example of the determination of the period of oscillation, the value of N, the standard inductance, was 6500 units. When the current was the same in both circuits, the value of R was 168 ohms. Therefore R 168.109 = 2.6 . 107. p = N = 6500 =

The frequency 11 = pJ21T = 4·1 . 106 • The value of the capacity and the inductance for rapid frequencies of the discharge circuit could also be determined. If a Leyden jar of unknown capacity C be replaced by an air condenser of known capacity C', the value of L remaining unaltered, and the value of the resistance necessary for equality of currents in the two circuits determined as be fore, then if p' and R' be the new values of p and R, 1

p=

vU:

and p'

R = pN and

=

1

VU:,.

R' = p'N.

Therefore C C'

R'2 Ij"2 - R2'

p'2

Since Rand R' have both been determined, C is known in terms of the standard capacity C'. Similarly, the value of L, the self-inductance for rapid frequencies, may

104

The Collected Papers of Lord Rutherford

also be found. If an additional standard inductance L' be introduced into the circuit, the value of the capacity remaining unaltered, p =

Therefore

V~C L

and p'

=

V(L

~ L')C'

+ L' _ L

p2 _ R2 - p'2 - R'2'

The value of L for rapid frequencies is thus determined in terms of L', a known inductance. In the experiments on the determination of periods, a detector consisting of twenty or more fine insulated steel wires, about 1 centim. long, was used, with one or two turns of wire round it through which the oscillatory current passed. This small detector coil was fixed before a magnetometer, and was so arranged that it could be switched either into the resistance or inductance branch of the divided circuit. The inductance of the detector coil was too small to appreciably alter the distribution of the current in either circuit. The equality of currents in the two circuits could thus be readily compared. The 'longitudinal' detector mayaIso be used for rough determinations; but it is not so sensitive to slight changes of current as the solenoidal detector of fine wire. It was found that the inductance of a current when the wire was of iron was nearly the same as when replaced by copper of the same diameter. Jt was difficult to determine the variation of inductance accurately in this case, in consequence of the oscillations being rapidly damped when iron wires were used. Since the capacities of condensers for very rapid alternations may be determined, it was interesting to observe whether the values of the specific inductive capacity of glass was the same for slow as for very rapidly varying fields. Some observers had found that glass had a much lower specific inductive capacity for rapid oscillations than for slow, while others again found values about the same in the two cases. The value of the specific inductive capacity found for plate-glass was about 6· 5 for periods of about 3 million per second. This is considerably higher than the value obtained for plate-glass by J. J. THOMSON and BLONDLOT, who found values of 2·7 and 2· 3 respectively for periods of about 20 million per second. The value of the specific inductive capacity of ebonite, tested by the same method, was found to be about the same as for slow alternations. These experiments were performed in the Cavendish Laboratory, Cambridge.

On the Passage of Electricity through Gases Exposed to Röntgen Rays by J.

Cavendish Professor of Experimental Physics, Cambridge, and E. R UTHERFORD, M.A., Trinity College, Cambridge, 1851 Exhibition Scholar, New Zealand University

J. THOMSON, M.A., F.R.S.,

From the Philosophical Magazine for November 1896, sero 5, xlii, pp. 392-407 (Read before Section A of the British Association, 1896)

facility with which agas, by the application and removal of Röntgen rays, can be changed from a conductor to an insulator makes the use of these rays a valuable means of studying the conduction of electricity through gases, and the study of the properties of gases when in the state into which they are thrown by the rays promises to lead to results of value in connection with this subject. We have during the past few months made aseries of experiments on the passage of electricity through gases exposed to the rays, the results of these experiments are contained in the following paper. Agas retains its conducting property for a short time after the rays have ceased to pass through it. This can readily be shown by having a charged electrode shielded from the direct infiuence of these rays, which pass from the vacuum tube through an aluminium window in a box covered with sheet lead; then, though there is no leak when the air in the neighbourhood of the electrode is still, yet on blowing across the space over the aluminium window on to the electrode the latter immediately begins to leak. To make a more detailed exarnination of this point, we used the following apparatus. A c10sed aluminium vessel is placed in front of the window through which the rays pass. A tube through which air can be blown by a pair of bellows leads into this vessel: the rate at which the air passed through this tube was measured by agas-meter placed in series with the tube; a plug of glass wool was placed in the tube leading to the vessel to keep out the dust. The air left the aluminium vessel through another tube, at the end of which was placed the arrangement for measuring the rate of leakage of electricity (usually a wire charged to a high potential placed in the axis of an earthconnected metal tube through which the stream of gas passed, the wire being connected with one pair of quadrants of an electrometer). This arrangement was carefuUy shielded from the direct effect of the rays, and

THE

D*

106

The Collected Papers 01 Lord Rutherford

there was no leak unless a current of air was passing through the apparatus; when, however, the current of air was flowing there was a considerable leak, showing that the air after exposure to the rays retained its conductive properties for the time (about l sec.) it took to pass from the aluminium vessel to the charged electrode. We tried whether the conductivity of the gas would be destroyed by heating the gas during its passage from the place where it was exposed to the rays to the place where its conductivity was tested. To do this we inserted a piece of porcelain tubing which was raised to a white heat; the gas after coming through this tube was so hot that it could hardly be borne by the hand; the conductivity, however, did not seem to be at all impaired. If, however, the gas is made to bubble through water, every trace of conductivity seems to disappear. The gas also lost its conductivity when forced through a plug of glass wool, though the rate of flow was kept the same as in an experiment which gave a rapid leak; if the same plug was inserted in the system of tubes before the gas reached the vessel where it was exposed to the Röntgen rays, in this case the conductivity was not diminished. This experiment seems to show that the structure in virtue of which the gas conducts is of such a coarse character that it is not able to survive the passage through the fine pores in a plug of glass wool. A diaphragm of fine wire gauze or muslin does not seem to affect the conductivity. A very suggestive result is the effect of passing a current of electricity through the gas on its way from the aluminium vessel where it is exposed to the Röntgen rays to the place where its conductivity is examined. We tested this by inserting a metal tube in the circuit, along the axis of which an insulated wire was fixed connected with one terminal of a battery of small storage-cells, the other terminal of this battery was connected with the metal tube; thus as the gas passed through the tube a current of electricity was sent through it. The passage of a current from a few cells was sufficient to greatly diminish the conductivity of the gas passing through the tube, and by increasing the number of cells the conductivity of the gas could be entirely destroyed. Thus the peculiar state into which agas is thrown by the Röntgen rays is destroyed when a current of electricity passes through it. It is the current which destroys this state, not the electric field; for if the central wire is enc10sed in a glass tube so as to stop the current but maintain the electric field, the gas passes through with its conductivity un~ impaired. The current produces the same effect on the gas as it would produce on a very weak solution of an electrolyte. For imagine such a solution to pass through the tubes instead of the gas; then if enough e1ectricity passed through the solution to decompose a1l the electrolyte the solution when it emerged would be a non~conductor; and this is precisely what happens in the case of the gas. We shall find that the analogy between a dilute solution of an electrolyte and gas exposed to the Röntgen rays holds through a wide range of phenomena, and we have found it of great use in explaining many of the characteristic properties of conduction through gases.

On the Passage

0/ Electricity throllgh Gases Exposed to Röntgen Rays

107

Thus Röntgen rays supply a meansof communicating acharge of electricity to agas. To do this, take an insulated wire charged up to a high potential and surrounded by a tube made of a non-conducting substance: let this tube lead into a large insulated metallic vessel connected with an electrometer. If now air which has been exposed to Röntgen rays is blown through the tube into this vessel the electrometer will be defiected. This proves that the gas inside the vessel is charged with electricity. If the Röntgen rays are stopped and the gas blown out of the vessel the charge disappears. In these experiments we took precautions against dust. The fact that the passage of a current of electricity through agas destroys its conductivity explains a very characteristic property of the leakage of electricity through gases exposed to Röntgen rays; that is, for a given intensity of radiation the current through the gas does not exceed a certain maximum value whatever the electromotive force may be, the current gets, as it were, 'saturated'. The relation between the electromotive force and the current is shown in the following curve, where the ordinates represent the current and

Fig.l the abscissae the electromotive force. It is evident that this saturation must occur if the current destroys the conducting power of the gas, and that the maximum current will be the current which destroys the conductivity at the same rate as this property is produced by the Röntgen rays. If we regard the gas as an electrolyte, then the passage of a quantity e of electricity will destroy eIE of the conducting particles, where E is the charge carried by one of these particles. Let n be the number of conducting particles in unit volume of the gas, q the rate at which these are produced by the rays, a.n 2 the rate at which these disappear independently of the passage of the current, t the current through unit area of the gas, I the distance between the electrodes. Then we have dn dt = q - a.n 2 - ~ . (1) lE'• . so that when the state of the gas is steady, 2

_I.

O=q-a.n -/e

. (2)

108

The Collected Papers of Lord Rutherford

When the current is small this equation gives n2 = q/rx.;

and as the number of conducting particles is independent of the current, the current will be proportional to the E.M.F. This corresponds to the straight part of the curve. In the general case the current is proportional to the product of n, the number of conducting molecules, and the potential gradient. If E is the difference of potential between the plates, U the sum of the velocities of the positively and negatively electrified particles when the potential gradient is unity, we have n€UE h or '=-/n = €UE. Substituting this value of n in equation (2), we get

o= q -

aJ2,2 €2U2E2 -

,

h.

• (3)

We see from this that , approaches the limit q€/. Thus the limiting current is proportional to the distance between the electrodes; so that when we approach saturation the current will increase as the distance between the electrodes increases, and we get what is at first sight the paradoxical result that a thin layer of air offers a greater resistance to the passage of a current than a thicker one. This is, however, easily accounted for if we remember that the current destroys the conducting power, and that as in a thicker layer there are more conducting partic1es than in a thinner one the current required to destroy them all will be greater. The experiments show that the effect of the distance between the electrodes (two parallel plates) on the current is very marked. The following tables show the result of some experiments on this point. POTENTIAL-DIFFERENCE BETWEEN ELECTRODES

Distance between electrodes, in millimetres

VOLTS

Current (arbitrary scale)

0·1 0·12 0·25 0'5

15 21

1· 5 3 8

62 91 110

1

60

9

37

50

With this large potential difference the current was saturated in all the experiments.

Oll fhe Passage

0/ Electricity throug/z Gases Exposed to Röntgen Rays

109

The next table contains measurements with a small potential difference. POTENTIAL-DIFFERENCE BETWEEN ELECTRODES

Distance between electrodes, in millimetres

0·25 0·75

scale)

10

32

2

48

18

40

3 8

1· 3 VOLT

Current (arbitrary

53 53

In this case the effect of distance is not so weIl marked as in the previous one, where the E.M.F. was sufficient to saturate the current at a11 distances. The measurement of the rate of leak when the current is saturated enables us to form an estimate of the number of conducting partic1es present in the gas; as in this case the number of conducting partic1es produced in unit time by the rays is equal to the quantity of the electrolyte destroyed by the current in the same time. Let us take the case of hydrogen; when the current was saturated, the rate of leak between two plates each about 10 sq. cm. in area and 1 cm. apart was about 1 volt per second when a capacity of about 30 cm. was in connection with the electrometer. Thus the quantity of electricity passing between the plates in 1 second was about 10-1 electrostatic units, or 1/3 X 101I electromagnetic units, and this quantity is sufficient to electrolyse all the electrolytic gas produced by the Röntgen rays. Now 1 electromagnetic unit of electricity sets free 10-4 grammes of hydrogen, or about 1 c.c. at atmospheric temperature and pressure. Hence 1/3 X 1011 electromagnetic units correspond to about the same number of cubic centimetres of hydrogen; the volume of the space between the electrodes was about 10 C.C., so in this experiment the fraction of the gas electrolysed was onIy 1/3 x 1012, i.e. one three-billionth of the whole amount of the gas. It is not surprising that some experiments we made to see if any alteration in pressure was produced when a gas was transmitting Röntgen rays should have given negative results. The preceding estimate gives the average number of conducting particles; if the conducting state is intermittent there may at certain times be a much Iarger number of these moleeules present. It is probable that, at all events, when the current is saturated the conducting power is intermittent. The action of the coil used to send the discharge through the vacuum tube is intermittent; thus, if between the passage of two sparks the conductivity has time to vanish (and when any current is passing through the gas the rate at which it vanishes is very rapid) the gas will be alternately an insulator and then a conductor. The following experiment is explained by the intermittent character of the discharge. The gas exposed to the Röntgen rays was in a piece of lead tubing open at both ends; this was connected with one terminal of a battery; the other terminal of which was connected with a wire running down the

110

The Collected Papers of Lord Rutherford

axis of the tube. Ablast of air was blown through this tube, and it was found that when the current between the wire and the tube was small, the blast diminished the current to a large extent, though a current approaching saturation was hardly affected by the blast. When the current was affected the gas blown out of the tube was conducting; when the current was not affected the gas did not conduct. If the gas were exposed to steady radiation it would not be affected by blowing unless the time taken by the gas to acquire the conducting state under the infiuence of the rays was comparable with the time taken by the gas to pass through the tube; this is inconsistent with what we know from other experiments as to the rapidity of action of the rays. If, however, the state of the gas is intermittent, then, since the blast continues when the rays are not acting, it blows out conducting gas and so diminishes its average conductivity. To return to equation (3), if I is the value of ~ when E is infinite, we may write the equation in the form L2

. (4)

I - ~ = C E2'

rxP

where

C = eU 2 '

and is independent of both E and L. We have observed the relation between the current and the electromotive force for several gases, and for different intensities of the Röntgen rays. The comparison of the results of these experiments with equation (4) is given in the following tables. LEAKAGE THROUGH CHLORINE GAS

Electromotive force

Current observed

*9 18 35 *70 140

65 124 200 245 270

Current calculated by equation (4)

116 180 275

The observations marked with the asterisks were used to calculate the constants. LEAKAGE THROUGH AIR

*9 18 35 *70 140

22 39 67 83 90

38 67 86

The observations marked with the asterisks were used to calculate the value of the constants in equation (4).

On fhe Passage of EleCfricity fhrough Gases Exposed to Röntgen Ra)'s

111

LEAKAGE THROUGH HYDROGEN

Electromotive force

Current observed

5 *9 18 35 *70

18 31 53 63 65

Current calculated by equation (4)

19 48 58

The observations marked with the asterisks were used to ca1cu1ate the constants in equation (4). LEAKAGE THROUGH CHLORINE (STRONG RADIATION)

5 *10 21 35 *70 140

53 100 189 275 355 380

53·4 183 255 405

LEAKAGE THROUGH CHLORINE (WEAK RADIATION)

*5 8'5 17 *35 105

10 16 26 32 34

15 23 37

COAL GAS (1)

1·4 2·8 4·2 8·4 16·8 35 110

10

17·3 22 32·3 38·3 43 45

9·8 23 33 40 44

COAL GAS (2) (WEAK RADIATION)

1·4 2·8 4·2 5·6 8·4 12·6 16·8

3·6 8 11 14·7 21·7 32

38

4·2 11·2 15·2 21·9 30·4

112

The Collected Papers of Lord Rutherford HYDROGEN

Electromotive force

Current observed

3·4 5·1 8·5 15·6 34 68

5 7·5 10 15 16·6 16·6

Current caIcuIated by equation (4)

6·9 10'1 13·4 17

SULPHURETTED HYDROGEN (STRONG RADIATION)

15·6 34 68 126

8·7 18 30·8 40

17·1 28·5

SULPHURETTED HYDROGEN (WEAK RADIATION)

15·6 34 68 136

3·8 6·3 8 8·7

6·2 8

MERCURY VAPOUR

5·1 8·5 15·6 34 68 136

14·2 23 35 55 75 75

14·6 36·9 59 8·2

As these measurements require the intensity of the radiation to be maintained constant during each series of observations, a condition which it is very difficult to fulfil, we think the agreement between theory and observation is as dose as could be expected. We have seen how from the measurement ofthe limiting current we could form an estimate of the proportion which the conducting partic1es bear to the rest of the molecules of the gas. We can, in addition, get from the curve representing the relation between the current and the electromotive force an estimate of the velocity with which these partides move. Taking equation (3) rx,[l,2 q - E2U2E'l -

,

k

= 0,

On the Passage of Electl'icity tlzrough Gases Exposed to Röntgen Rays 113 we shall endeavour to express the coefficients in terms of quantities which our experiments enable us to estimate. Let I be the limiting current when the electromotive force is infinite, then

I=qh. Let T be the time which elapses after the rays have been stopped for the number of conducting partic1es to fall to one half the number just before the rays ceased, no current passing through the gas. Then, just before the rays cease to fall on the gas, we have from equation (2),

N={~r where N represents the number of conducting partic1es at this stage; after the rays have ceased, we have

d"

dt = I

or

11

I

-lXn 2,

N=

IXt,

if t is the time which has elapsed after the rays have stopped, when t = T, n = ~N, hence

N= I

IXT;

substituting for N its value, we get 1 T2 - IXq

or

IX

Substituting for q and

IX

=

I h T2q = T2I .

the values just found, equation (4) becomes T-

t

Pt 2€

~ =

or

.

I(I-t)

IT2€2E'l

a 2'

[4 t 2

. (5)

T 2E2U2' .

Thus, in the straight part of the curve, where we have approximately EUT

/=7'

t

is small compared with I, .

(6)

Now, EU/l is the sum of the velocities of the positively and negatively charged particles in the electric field. Hence, equation (6) shows that the

114

The Collected Papers of Lord Rutherford

current bears to the maximum current the same ratio as the space described by the charged particles in time T bears to the distance between the electrodes. In an experiment where I was about 1 cm., the rate of leak through air for a potential difference of 1 volt was about 310 of the maximum rate of leak, hence the charged particles must in the time T have moved through about 31Ö of a centimetre. The time T will depend upon the intensity of the radiation; it could be determined by measuring the rate of leak at different points on the tube through which the conducting gas was blown in the experiment mentioned at the beginning of this paper. We hope to make such experiments and obtain exact values for T; in the meantime, from the rough experiments already made, we think we may conclude that with the intensity of radiation we generally employed, T was of the order of 110 of a second. This would make the velocities of the charged particles in the air about O' 33 cm.jsec. for a gradient of one volt per cm. This velocity is very large compared with the velocity of ions through an electrolyte; it is, however, small compared with the velocity with which an atom carrying an atomic charge would move through agas at atmospheric pressure ; if we calculate by the kinetic theory of gases this velocity, we find that for air it is of the order of 50 cm.jsec.; this result seems to imply that the charged particles in the gas exposed to the Röntgen rays are the centres of an aggregation of a considerable number of molecules. The relation between the current and electromotive force given by equation (4) corresponds to that obtained by experiment for a number of gases; it does not, however, exhibit a peculiarity which we have sometimes observed, especially when the radiation was strong, i.e. the existence of apart of the curve where the current increases faster than would be the case if Ohm's law were true; this is shown by the portion EF of the curve in Fig. 2, which

Fig.2 represents the relation between the current and electromotive force through sulphuretted hydrogen. When the intensity of the Röntgen rays is altered, the alteration in the current is not the same at different points in the curve.

On the Passage 01 Electricity through Gases Exposed 10 Röntgen Rays 115

When the intensity of these rays is diminished, the saturation current is diminished in a larger proportion than the current for small electromotive forces. This is shown by the following diagram, which represents

Fig. 3 the Land E curves through chlorine gas for different intensities of the Röntgen rays; the weak radiation was got by interposing a thick aluminium plate. In this diagram the ordinates for the weak radiation have been in.. creased so as to make the ordinate for the saturation current of the weak radiation the same as that of the strong. When this is done, the rest of the 'weak' curve is above the strong, showing that the diminution in the radiation has affected the saturation current to a greater extent than the weaker currents. The saturation current depends only on the number of conducting particles produced by the rays; for the smaller currents the diminution in the number of molecules is, to some extent, compensated for by the increase in the time taken for these to recombine; thus T is increased when the intensity of the rays is diminished, so that, as we see from equation (6), the proportion between a small current and the saturation current is increased when the intensity of the rays is diminished. Whatever is the magnitude of the electromotive force, a diminution in the intensity of the rays is accompanied by a diminution in the current, so that the land E curves for two intensities of radiation would not intersect if both were drawn on the same scale. If, however, instead of keeping the gas the same and altering the intensity of the radiation, we alter the gas and keep the intensity of the rays constant, then the land E curves for two different gases may intersect. This effect is shown in the following diagram, which represents the land E curves for hydrogen and air. We see that for small electromotive forces the current is greater in hydrogen than in air, while the saturation current is much greater in air than in hydrogen. The saturation current depends merely on the

116

The Collected Papers

0/ Lord Ruthelford

Fig.4 number of conducting partic1es produced by the rays, while the current in the earlier part of the curve depends on the space described by the conducting particles in the time T [see equation (6)], and we infer that more conducting particles are produced by the rays in air than in hydrogen, but that the product of U, the velocity of these particles, and T, a time which is proportional to the time these partic1es linger after the rays are cut off, is greater for hydrogen than it is for air. In Fig. 5 we give the curves for air, chlorine, sulphuretted hydrogen, and mercury vapour, the curves being drawn on such scales that the ordinate representing the saturation current is the same in all these cases. lt will be noticed that the curves for air, for sulphuretted hydrogen, and for chlorine coincide, mercury vapour falls below, while the hydrogen curve would be

Fig.5

On the Passage oI ElectricilJ' through Gases Exposed to Röntgen Rays 117 above. This shows that, using the notation of equation (6), ur is the same for air, chlorine, and sulphuretted hydrogen, and that its value for these gases is smaller than for hydrogen and greater than for mercury vapour. It is remarkable that the shapes of the curves for air, sulphuretted hydrogen, and chlorine should agree so closely, for the absolute values of the current in these gases are very different, the saturation current in sulphuretted hydrogen being in some cases three or four times that of air, while that of chlorine is in some cases as much as ten times that of air. The value of the saturation current varies greatly in different gases; of the cases we have tried it is least in hydrogen, greatest in mercury vapour, the saturation current in mercury vapour being about 20 times that for air. It does not seem to depend entirely on the density of the gas, as in sulphuretted hydrogen it is three or four times what it is in air, though the densities are nearly equal, while, though the density of the vapour of CH2I 2 is greater than that of mercury vapour, the saturation current in the former gas is only a small fraction of its value for the latter. The gases which have large saturation currents are those which contain the elements which have an abnormally large specific inductive capacity in comparison with their valency. We have made a large number of experiments with the view to seeing whether there is any polarization when a current of electricity passes through the gas; we have not, however, been able to satisfy ourselves ofthe existence of this effect. The absence of polarization implies, however, that the ions are able to give up their charges to the metal e1ectrodes. Experiments on electrified gases show, however, that it is very difficult to get acharge of electricity from agas to a metal unless the metal is exposed to radiation, either by the metal being sufficiently hot to be luminous, or when it is exposed to ultra-violet light. But in the case of the passage of electricity through agas which has been exposed to Röntgen rays the conduction takes place even when the system is not exposed to the direct radiation from the exhausted tube; we think it probable, therefore, that the gas itself radiates after being exposed to the Röntgen rays. To test this we tried the following experiment. AB, CD are two concentric cylinders made of thick lead tubing, the base of the inner one was cardboard, so as to allow Röntgen rays to pass through the gas in the inner cylinder. A metal ring was placed between the two cylinders and connected with one pair of quadrants of an electrometer so as to allow the leak from it when raised to a high potential to be measured. A slit was cut in the inner cylinder in such a place and of such a size that no rays could pass through it directly from the bulb. The apparatus was fi1led with chlorine, as this gas is one which gives a very rapid rate of leak. When the slit was left open there was a rapid leak due to the diffusion from the inner cylinder of gas which had been exposed to Röntgen rays. When, however, the slit was covered up with a strip of paper, the leak wholly disappeared, though the ring connected with the electrometer was placed at the same level as the

118

The Collected Papers of Lord Rutherford

A

C

0

B

Fig.6 stit and, therefore, exposed to any radiation that might come from the gas. This radiation, if it exists, must therefore either be of very feeble intensity or else it must differ from the Röntgen rays in not making a gas through which it passes a conductor of electricity. We are inc1ined to think that when Röntgen rays are incident on a metallic surface the 'diffusely refiected' rays are not of the same character as the incident ones, and have not nearly the same power of rendering agas through which they pass a conductor of electricity. We base this opinion on the experiments we have made to detect the existence of electrical effects due to the 'refiected' rays; though we have made many attempts we have never been able to detect the existence of any electrical effects from the reflected rays. Thus we introduced in the apparatus in Fig. 6 a lead plate inclined at an angle of 45° to the axis of the cylinder, and so placed as to reflect the rays through the slit, which was covered with a strip of paper; the arrangement was so sensitive that if the plate had reflected anything like one per cent of the rays incident upon it, the leak from the metal ring would have been easily detected; there was, however, no trace of a leak. The results of experiments on the photographic effects produced by rays diffusely refiected from metallic plates seem to show that these rays are fairly abundant. Taking this result in connection with the absence of any noticeable electrical effect produced by these diffusely reflected rays, we think that the latter differ in character from the incident rays. We have not been able to detect any effect produced by a magnetic field on the rate of leak; we tried with the lines of magnetic force parallel and also at right angles to the current, and with both small and saturated currents. The rate of leak through air that had been dried by standing for three days in the presence of phosphorus pentoxide did not differ appreciably from the damp air of the room. In conclusion, we desire to thank Mr E. Everett for the assistance he has given us in these experiments. The period during which a bulb gives out Röntgen rays at a uniform rate is not a long one, and as most of our experiments required the rate of emission to be constant, they have entailed the use of a very large number of bulbs, all of which have been made by Mr Everett.

On the Electrification of Gases exposed to Röntgen Rays, and the Absorption of Röntgen Radiation by Gases and Vapours by E. R UTHERFORD, M.A., 1851 Exhibition Science Scholar, University 0/ New Zealand,' Trinity College, Cambridge

From the Philosophical Magazine for April 1897, sero 5, xliii, pp. 241-255 Communicated by Professor J. J. Thomson, F.R.S.

IN arecent paper by Professor J. J. Thomson and myself, 'On the Passage ofElectricity through Gases exposed to Röntgen Rays' (Phi!. Mag., November, 1896), a method of obtaining electrified air by means of the Röntgen rays was very briefly explained. The present paper deals with further experiments which have been made to investigate more fully the way in which electrified gases can be obtained by means of the Röntgen rays, and also to examine the properties of the charged gas. The opacity of gases for Röntgen radiation has also been examined. Agas becomes a temporary conductor under the influence of the Röntgen rays, and preserves its power of conducting some short time after the rays have ceased to act; since the conduction in the gas is probably due to the convection of charged particles which travel through the gas with a velocity of the order of 1 cm. a second for a potential gradient of one volt per em., it is not surprising that we can separate the positive from the negative conducting particles before they give up their charges to the electrodes. The method of separation used was to direct a rapid eurrent of air or other gas along the surface of the charged electrode of avessei exposed to the Röntgen rays; a large meta! cylinder was taken, either of thin metal to allow the rays to readily pass through the side, or a piece was cut out and a sheet of very thin metal substituted to serve the same purpose. A (Fig. I) was the metal cylinder, BC a glass tube fixed centrally inside the cylinder. The wire DE was fixed in the glass tube BC and supported by thin metal spikes in the cent re of the tube. Several inches of wire, BD, projected from the glass tube. A current of gas was sent from a pair of bellows or a gas reservoir along the tube CB, and then along a metal tube into an insulated conductor connected with one pair of quadrants of an electrometer, the other pair being connected to earth. The wire DE was connected to one pole of a battery 0 f small lead cells, the other pole being to earth.

120

The Collected Papers

0/ Lord Rutherford

The outside of the cylinder was connected to earth, and the bulb and Ruhmkorff coll were placed inside a metal tank, so as to completely screen the outside apparatus from electrostatic disturbances. A hole was cut in the tank, and the bulb arranged so as to allow the rays to fall on the part BD of the charged wire. When the bulb was not working, however rapid a current of air was sent along the charged electrode, no electrification was obtained in the inductor, but the moment the rays were turned on the inductor became charged opposite in sign to the charged wire. The defiection of the electrometer continuously increased as long as the rays and blast of air were acting. A

Inductor

F

o

o

o

o

7i Ihttery

Fig.l The inductor was generally placed some feet from the generating vessel A, the air passing to the inductor through a metal tube of 3 cm. diameter. Since the electrification of the inductor is opposite in sign to the charged wire, the effect can in no way be due to conduction through the Röntgenized air from the charged electrode to the inductor, since the inductor would then be charged to the same sign as the electrode. A small plug of glass wool placed in the metal tube between the generator and the inductor completely stopped all electrification, and the inductor received no charge, however rapid ablast of air was sent through the apparatus. On account of the large quantity of air blown through the inductor in order to obtain a convenient defiection on the electrometer (an amount sufficient to fill the inductor many times over), only a small proportion of the charge could be blown out. If, however, a gentle current of electrified air was blown into the inductor for two or three seconds, and the rays then stopped, it was found possible to blow out most of the charge again, after a short interval, provided there was a fairly wide opening i!). the inductor. If the opening was stopped with a plug of glass wool, it was found impossible to blow out the slightest amount, since the electrified particles gave up their charge freely to the glass wool.

On file EleCIJ'ijicatioll 0/ Gases Exposed 10 Röntgen Rays

121

Since the glass wool has the power of completely discharging the electrification both positive and negative, a short wide metal cylinder lightly packed with glass wool was used instead of the inductor for testing the amount of the electrification in most of the experiments that follow. The amount of electrification obtained varied with the potential of the charged wire and the velocity of the current of air. The relation between the amount of electrification and the E.M.F. of the charged wire is shown in the table below: E.M.F. in volts

17 35 70 200

Amount of Electrification in scale divisions

60

100 130

82

The amount of electrification increases up to a certain point and then diminishes. The maximum amount of electrification is closely connected with the value of the E.M.F. which is just sufficient to give the saturation value of the eurrent through the gas. In the above table the saturation value of the E.M.F. was about 70 volts, and this corresponds to the maximum amount of electrification. Since the velocity of the conducting particles increases with the E.M.F. but the current through the gas remains constant, when the E.M.F. is raised above its saturation value it is to be expected that a greater proportion of tbe conducting particles would reach the electrode. This agrees witb experiment, for as the E.M.F. is increased above a certain value the amount of electrification obtained steadily diminisbes. The amount of electrification obtained for a given E.M.F. increases at first with the velocity of the blast, and then tends to a maximum value, which cannot be increased, however rapid ablast is sent along the wire. An experiment proving conc1usively that the amount of electrification is intimately connected with the conduction of electricity through the gas is as follows: The electrode along which the air was blown was carefully insulated and connected to one pair of quadrants of the electrometer. The two pairs of quadrants were charged up to the same potential and then insulated from each other, and the rate of leak of the charged wire determined. The rate of leak steadily diminished with increase of velocity of the blast; when the air issuing from the glass tube had a velocity of about 1,000 em. per second the rate of leak was only one-fourth of its value when the air was still, and the amount of eIectrification in the air passing from the wire, as tested by the glass wool cylinder, was nearIy equal to the quantity of electricity corresponding to the difference between the two rates of leak. We should not expect them to be exactly equal, since some of the Röntgenized air containing both positively and negatively charged particles is also blown out.

122

The Collected Papers o[ Lord Ruther[ord

The eharged gas obtained in this way is thus due to an exeess of the positive or negative eondueting particles, whatever they may be, to whieh eonduetion in gas under the Röntgen rays is due. In an these experiments preeautions were taken against dust. It was found that the amount of eleetrifieation obtained was independent of the quantity of dust in the air provided the velocity of the issuing blast was kept eonstant. The air in one ease was sent through a long tube filled with glass wool into the gas reservoir, whieh was then allowed to stand for a eouple of days without being disturbed. The air in passing from the reservoir to the generator was again passed through glass wooI, but the effeet obtained was exaet1y the same as if the dust-eharged air from the room were sent direet1y through the apparatus.

Electrification from Charged Insulators If the central eleetrode through whieh the air was blown was eoated with

paraffin or sealing-wax, it was found that the amount of eleetrifieation obtained was at first about equal to the amount with the bare eleetrode. If the bulb was kept working the amount of eleetrifieation diminished after a time. The central eleetrode was then eonneeted to earth, and when the x-rays were aeting eleetrifieation eould still be obtained, but of opposite sign to that obtained before. If the wire with the eoating of dielectric on its surfaee was kept charged to a high potential and the rays continued for some time, on applying a smaller E.M.F. to the wire in the same direction the sign of the electrification is generally changed. The explanation of these and similar phenomena is simple if we consider that the condueting partic1es of the gas either give up their charge to the surface of the insulator, or adhere to the surface which becomes charged opposite in sign to the wire itself. If the bulb is kept working, the electromotive intensity aeting on the gas is diminished, owing to the effect of the oppositely eharged insulator. The amount of eleetrification obtained therefore diminishes if the E.M.F. is not wen above the saturation value. If the eentral eleetrode be then conneeted to earth, the charged insulator causes a eurrent through the gas in the opposite direction, and thus ehanges the sign of the charge in the gas blown out. If the charge on the insulator is large, as is the case if the eentral wire has been raised, for example, to a potential of 200 volts and exposed to the rays for some time, on applying an E.M.F. of 30 volts, say, in the same direction the electrification changes sign. In this case, the eleetromotive intensity due to the eharged insulator is greater and opposite in sign to that due to the 30 volts, and so the current through the gas is reversed. The sign of the eleetrifieation obtained when the wire is covered with insulating material is thus dependent on the amount and sign of the charge on the surfaee of the dieleetric. From a charged wire coated with paraffin or sealing-wax, which had

On the Electrijication of Gases Exposed to Röntgen Rays

123

been exposed to the Röntgen rays for several minutes, it was found possible to obtain electrified air, by directing a current of air along its surface several hours after the central electrode had been connected to earth. Properties 0/ the Charged Gas

Since the charged gas obtained is due to the separation of the oppositely charged conducting particles to which conduction in agas under the x-rays is probably due, we should expect the positively and negatively electrified gas to c10sely resemble Röntgenized air in its properties, and such is found to be the case. The gas completely loses its charge in its passage through the pores in a plug of glass wool; while Röntgenized air, after being forced through glass wool, loses all trace of conductivity. The gas readily gives up its charge to any conducting or insulating surface against which it impinges. The greater amount of electrification is discharged in the passage of the gas down a long tube. If the e1ectrified air is allowed to impinge against the surface of an insulated metal plate, it gives up a portion of its charge to the meta!. The facility with which the gas is discharged is to be expected, since no evidence of polarization has been found in the conduction of the gas exposed to the Röntgen rays when metal electrodes are used. A remarkable property of the electrified gas is that positive and negative electrification are not discharged with equal facility by all metals. When the charged gas was passed through a long zinc tube, the amount of negative electrification on the issuing gas was always less than the amount of positive for the same velocity of the blast. By insulating the zinc tube it was found that it received a greater charge of negative than of positive. In order to test this difference, cylinders of zinc, tin, and copper were made of the same size, and the charged gas forced through them. It was found that zinc and tin discharged negative electrification more rapidly than positive, the difference, in general, amounting to about 20 per cent. Copper apparently discharged the positive and negative with about equal facility, but many experiments seemed to point to the conc1usion that even in the case of copper negative was slightly more readily discharged than positive. If the electrified gas impinged against insulated plates of different metals, the same general results were obtained. Not only was there a difference in the discharging powers of positive and negative electrification for any particular metal, but a copper plate, for example, discharged positive more readily than a zinc plate placed exactly in the same position, while the zinc plate discharged more negative than the copper. Other metals, like aluminium, lead, were tested, and in all cases negative electricity was discharged with slightly more facility than positive. The variable discharging power of the different metals agrees in some

124

The Collected Papers of Lord Rutherford

respects with the results obtained by Minchin (Electrician, March 27, 1896), who found that under the influence of the Röntgen rays insulated metal plates were all charged up to a small potential. According to his results copper was charged positively and zinc negatively, while sodium was highly negative. He also found that the potential to which some of the metals could be raised depended on the degree of polish of the exposed surface. In the experiments on the discharging power of the metals, the results were dependent to some extent on the brightness of the surface, especially in the case of tin and zinc. The amount of electrification discharged by a metal tube one inch in diameter and a foot long is very large, amounting in some cases to over onefourth of the whole charge on the gas. It must be remembered, however, that the current of air conveying the charged gas is travelling at a high velocity, and is, in consequence, in astate of violent eddying motion, so that probably a large proportion of the gas approaches near the surface in its passage along the tube. The charge is taken from the gas not only when it passes through metal tubes, but also when it passes through tubes coated with an insulator . A metal tube was taken and coated with a thin layer of paraffin, and it was found that the charge on the cylinder was about the same as with the clean metal. It was difficult to determine with certainty whether insulators exhibited similar properties to metals in regard to discharging power. The amounts of positive and of negative electrification discharged were approximately the same, but the differences were too small to make certain of. The conductivity of the charged gas was tested by placing an insulated wire kept at a constant potential inside a metal vessel through which the electrified gas was blown. It was found that when the electrification was of the same sign as the charged wire, the gas gave up its charge to the outside vessel, and when of the opposite sign, to the charged wire. The current through the gas was only temporary, and ceased as soon as the current of electrified gas was stopped.

Electrification of Different Gases All the gases which were experimented with could be eIectrified in the same way as air. A gas-bag was fi1Ied with the gas to be tested, and then forced along the electrode as in the case of air, care being taken to allow the gas to run through some time before the rays were turned on, in order to remove the air as far as possible from the generating vesse1. The amounts of eIectrification obtained for a given velocity of the gas and intensity of the rays varied with the conductivity of the gas under the x-rays. Gases that have a greater conductivity than air gave more electrification than air. Oxygen and coal-gas gave slightly less electrification than air, while carbon dioxide gave slightly more-the amounts being sensibly proportional to their conductivities.

0" fhe Electrifieation

0/ Gases Exposed to

Röntgen Rays

125

The vapour of methyl iodide was tried, which has a very high conductivity -over twenty times that due to air. Only a partial test could be made of it, as sufficient quantity of the vapour was not obtainable. Some of the liquid was placed in the generator (Fig. 1), and gently heated to its bolling point, till the vessel was filled with vapour. The rays were then turned on, and a rapid current of air sent for 3 or 4 seconds along the central electrode. The amount of electrification obtained was over five times the amount from air in the same time. After a few seconds the highly conducting vapour was blown out and the electrification became sensibly that due to air alone. If a current of the vapour could have been sent instead of a current of air, it is probable that the amount of electrification obtained would have been over twenty times that of air in the same time. The experiments on hydrogen were interesting as bearing on the question of the relative velocity of the conducting particles of hydrogen and air. For a given small weight on the gas-bag the amount of electrification from air was 2·5 times that due to hydrogen in the same time. As the weight was increased the ratio fell to 1·5. This is as we should expect if the velocity of the hydrogen ions was greater than those of air. For small velocities of the blast a much smaller proportion of the hydrogen than the air ions escape. As the velocity is increased the amount of electrification from the air increases slowly, as nearly all of the ions are blown out, while the amount from hydrogen increases rapidly as the veloeity inereases. In a previous paper (loe. eit.) it was shown that hydrogen was saturated for a mueh lower value of the E.M.F. than air, while the velocity of the hydrogen ion was much greater than that of air. The experiments on the amount of eleetrification with variation of E.M.F. and velocity of the blast eonfirm the previous results whieh were obtained in an entirely different way. Veloeity 01 the [ons.-An approximate determination of the veloeity of the eonducting particles for air ean be made by determining the rate of leak of a charged wire eonneeted to an eleetrometer when air is blown at varying velocities from a tube of known diameter along the charged wire. If we assume the current of air of high veloeity from the tube to be confined within narrow limits for a short distance from the orifice, the velocity of the ions in order that a known proportion of the ions should reach the electrode can easily be dedueed. The velocity of the blast issuing from a tube 0·8 em. in diameter was 800 em. per seeond, and the wire was eharged to a potential of 35 volts. With this veloeity the rate of leak of the wire was only one third of the natural leak; so that two thirds of the eondueting particles of one sign were blown out. The length of the exposed wire BD (Fig. 1) was 6·3 cm., and knowing the diameter of the wire BD and of the cylinder, it can be shown that the veloeity of the condueting particles for air is about 1 em. per second for a potential gradient of one volt. This is of the same order as the rough determim.tion made in the previous paper by Professor Thomson aod myself.

126

The Collected Papers o[ Lord Ruther[ord

The positive and negative conducting particles of air travel with the same velocity , for when the sign of the charged wire is reversed the rate of leak is the same as before with the same velocity of the blast. When different amounts of positive and negative electrification were obtained, it was at first thought that part of the difference might be due to inequality in the velocity of the ions, but later experiments showed that it was entirely due to the greater facility with which metals discharged the negative electrification.

Volume Density 0/ Electrification

0/ the Charged Gas

Only a very minute portion of the gas conveys the charge in the cases we have been considering. In the paper previously referred to it has been shown that assuming that conducting partic1es convey an atomic charge, only about one billionth of the gas is required to be split up to give the conductivity observed. In the previous experiments the conducting air is still further diluted by the blast along the electrode which conveys the charged partic1es with it. From data of the capacity of the electrometer and velocity of the blast it can be shown that the amount of charge per c.c. of air was about 10-4 electrostatic units. In the case ofthe better conducting gases and vapours the volume density is greater. For the vapour of methyl iodide the volume density would be over twenty times as great. The facility with wh ich the electrified gas is discharged by metals and insulators may at first sight lead to the conclusion that we are dealing with electrified dust, which, as is well known, is compIeteIy discharged by glass wooI, and also readily gives up its charge to whatever it comes in contact with. It has been shown, however, that the amount of electrification is quite independent of the amount of dust in the air, and that, therefore, the electrification can be in no way due to electrified dust. The theory has been advanced that the discharge of electricity from the surface of a metal under the action of ultra-violet light is due to the disintegration of metallic partic1es or vapour from the surface, and that these carry off the charge. The discharge of electrification by the Röntgen rays might possibly be due to a similar cause, and this was fully investigated. In the first place, there are many experiments which negative this view. It has been shown in a previous paper that the current through agas conducting under the x-rays increases with the distance between the electrodes although the surface exposed to the gas is unaltered. The amount of electrification obtained from a gas was found to be quite independent of the nature of the electrodes. The inside electrode (Fig. I) of the cylinder was coated with paraffin or wax, and provided we do not allow the charge to collect on the surface of the insulator, the amount of eIectrification was unaltered. Similarly, if the inside of the cylinder was coated with an insulator, no difference in amount could be detected immediately after the rays were turned on; but after the rays had been acting for a short time the amount decreased, owing to the charging of the surfaces of the insulators. These

On fhe Electrijicafion of Gases Exposed fo Röntgen Rays

127

conc1usions show that the conductivity in agas is independent of the nature of the surface of the electrodes, for it is extremely improbable that the same amount of dust would be dislodged by the rays from the surface of all metals and insulators. The most conclusive experiments on this subject are some which I recently made on the diminution of the intensity of the Röntgen rays due to the absorption in their passage through gases and vapours. Absorption

0/ Energy by Gases and Vapours

Since gases all conduct under the influence of the x-rays, it was interesting to investigate the relative absorption in order to make the gases conductors, and whether the absorption was in any way related to the constitution or conductivity of the gas. The absorption of energy in gases like air, hydrogen, oxygen is small and is not easily detected unless a delicate null method is used. Two equal and similar conical-shaped vessels, ABC, A'B'C' (Fig. 2), much larger in diameter at the top than the bottom, were placed in such

o

D'

/I'

/I' /I'

/I'

/I'

/I' /I'

C',

Fig.2 positions that the axis of each cone passed as nearly as possible through the anode of the focus bulb. From experiments it was found that the x-rays appeared to emanate in all directions from the anode. The upper parts of each vessel, AB, A'B', were made of lead, and were separated from the lower portions, Be, B' C', which were made of glass, by thin ebonite plates. Thin ebonite plates also covered the ends of the glass cylinders at C and

128

The Collected Papers

0/ Lord Ruther/ord

C', so that the vessels BC, B'e' were airtight, and could be exhausted when required. The lead cylinders AB, A'B' were used to compare the rate of leak after the rays had passed through the glass cylinders. Insulated wires DE, D'E', formed the electrodes, and these were connected to opposite pairs of quadrants of the electrometer, and both quadrants were at first charged up to the same potential which, in practice, was generally 200 volts. The outsides of the vessels AB, A'B' were connected to earth. The position of the bulb was so adjusted that the rate of leak in each cylinder was exactly the same, so that since the potential of each pair of quadrants fell at the same rate, the needle of the electrometer remained at rest while the rays were kept acting. If another gas was introduced into one of the glass vessels BC, it was found that the balance was disturbed, owing to the variation of intensity of the rays in the vessel AB, which was caused by the less or greater absorption of the rays in their passage through the gas. In the experiments the only rays which caused conduction in the lead cylinders had to pass through the gas, and all stray radiation was carefully screened off. If we assume that the absorption of energy in passing through a thin Iayer of gas of thickness dl is proportional to the intensity of the rays I at that point and to the length of the gas dl traversed, the decrease of intensity of the rays due to absorption 0/ energy in the gas is equal to Mdl, where A is a constant for any particular gas but varies for different gases, and may be called the coefficient of absorption of the gas. Experimentally it was found that the rate o/leak of agas is proportional to the intensity of radiation at any point. From these considerations it can readily be shown that the ratio of the rate of leak when the rays pass through a length I of the gas, to the rate of leak when the gas is removed and a vacuum substituted is e-Al where e = 2· 7, and this result is independent of any metal or insulators which the rays pass through in both cases berore reaching the testing vessels. The ratio of the rates of leak can be readily deduced from the movement of the electrometer needle, and since the length of the gas traversed is known, the coefficient A is thus determined. Experiments were first made to see whether air absorbed any appreciable amount of energy of the radiation. The balance was obtained and then one of the glass vessels was exhausted by an air pump; the electrometer slowly moved in one direction while the rays were kept acting. If the other vessel was also exhausted, the balance was again restored, and if air was then let into the vessel first exhausted, the electrometer needle moved in the opposite direction. The variation of the rate of leak after passing through 10 cm. of air was about one per cent, but it was a difficult matter to determine such a small variation with accuracy. It will be at once seen that the approximate value of >..L is 10-2, and therefore for air A = 10-3, since 1= 10 cm. If we suppose x-radiation to be emitted by the sun, assuming the radiation

PHYSICS RESEARCH GROUP, CAVENDISH LABORATORY, JUNE 1898 Top Row: O. W. Richardson; J. Henry Middle Row : E. B. H. Wade; G. A. Shakespear; C. T. R. Wilson; E. Rutherford; W. Craig Henderson; J. H. Vincent; G. B. Bryan Bot/om Row: J. C. McClelland; C. Child; P. Langevin; Professor J. J. Thomson; J. Zeleny; R. S. Willows; H. A. Wilson; J. Townsend

This page intentionally left blank

Oll Ille UC('lrijical icm (~r Gases E-Yl'osecl tu Röntgen Rays

129

would havc to pass through four miles of homogeneous atmosphere, the intensity at the surfaee of the earth would be approximately 10-260 of the intensity of the radiation before it reached the earth's atmosphere. This is an excessively minute proportion, and it is not surprising, therefore, that experiments, made even on the highest mountains, to detect any Röntgen rays in solar radiation (Cajori, Phi!. Mag., November, 1896) should have been unsuceessful, even if the intensity of the x-radiation at the limits of our atmosphere were greater than could be produced at the surface of a Crookes tube. Gases like oxygen, coal-gas, carbon dioxide, whose leakage rates are about the same as that of air, absorb about the same amount of energy. Sulphuretted hydrogen, which has a conductivity six times as great as air, diminishes the intensity of the radiation by about 4 per cent in passing through 10 cm. of the gas. Chlorine, whose conductivity is eighteen times that of air, diminishes the intensity about 12 per cent for the same distance. The absorption of energy in these cases is not necessarily selective, for the same results were obtained whatever gas was used in the testing vessels. After the radiation had passed through sulphuretted hydrogen, the same diminution in intensity was obtained whether air or sulphuretted hydrogen was used in the testing vessels. Mercury vapour, which is one of the best conductors of electricity under the x-rays, is also one of the best absorbers of the radiation. In the ease of mercury is was not necessary to use a null method. A glass bulb 7 em. in diameter was taken and a small amount of mercury introduced. The bulb was slowly heated till it was filled with boiling mercury vapour, eare being taken that no mercury was allowed to condense on the sides of the glass through which the rays passed. The rate of leak was then taken in a eonducting vessel for the radiation which had passed through the mercury vapour. The mereury was then removed and the bulb now filled with air was heated to the same temperature. The rate of leak in the testing vessel with air in the bulb was found to be twice as great as when the bulb was filled with mercury vapour at the temperature of boiling mercury. In the passage therefore through 7 cm. of mercury vapour sufficient energy is absorbed to reduce the intensity of radiation by one half. The vapour of methyl iodide, which is even a better conductor than mercury vapour, is also a powerful absorber of the radiation. When the temperature of the vapour was raised above the boiling-point of the liquid, the intensity of the radiation after passing through 13 cm. of the vapour was only 0·4 of the intensity when the vapour was removed. From experiments on the absorption of energy of different lengths of the vapour of methyl iodide, and also of the gases sulphuretted hydrogen and chlorine, the ratio of the intensity of radiation after passage through the vapour or gas to the intensity of the radiation when the gas was removed was found to be in agreement with the theoretical ratio e-N. For short lengths of the gas the absorption is proportional to the length. E

130

The Collected Papers of Lord Rutherford

It has been shown that for the vapour of methyl iodide rAt = 0·4 when 1= 13 cm.; therefore, >. = 0·07. The intensity of the radiation after passing through a length of 1 metre of the vapour is only one thousandth part of its value when the vapour is removed. The absorption of energy varies with the pressure of the gas. The vapour of methyl iodide was used, and it was found that the values of >. were roughly proportional to the pressures down to apressure of one quarter of an atmosphere. Results of this kind, however, are difficult to determine with accuracy on account of the variation of the Crookes tube during aseries of observations. The following table gives the values of >. and relative conductivities of some of the gases: Gas

A (small) 0·001

Hydrogen Air Oxygen Nitrogen about 0'001 Coal-gas Carbon dioxide Sulphur dioxide 0·0025 Sulphuretted hydrogen 0·0037 Hydrochloric acid 0.0065 Chlorine 0.0095

Conductivities

0·5 1 1·2

0·9 0·8

1·2 4 6 11 18

These experiments show that good conductors under the x-rays are good absorbers of the radiation. The absorbing powers for the gases examined had the same relative order as their conductivities. Tbe absorption does not seem to depend to any great extent on the molecular weight of the gas. Hydrochloric acid is nearly twice as good an absorber as sulphuretted hydrogen, although their densities are nearly equal; while it is more than ten times as good an absorber as carbon dioxide, agas of greater density. It is interesting to observe that vapours like mercury and methyl iodide which allow light to pass through freely are very opaque to Röntgen radiation. Since the absorption of energy of the radiation varies with the length of the gas traversed and with the conductivity of the gas, it is very strong evidence that the discharge of electrification by the Röntgen rays is due to a process going on throughout the volume of the gas, and is not due to the disintegration of charged dust from the electrodes. Cavendish Laboratory December 28, 1896

On fhe Electl'ijicatioll o.f Gases Exposed to Röntgen Rays

131

NOTE ON THE PRECEDING PAPER

by J. J.

THOMSON

The connection obtained by Mr Rutherford between the coefficient of absorption and the saturation current through the gas admits of an interesting method of expression on the theory that the Röntgen rays so far resemble light as to be of the nature of an electromagnetic wave or impulse. We may regard such a wave or impulse as consisting of groups or a group of Faraday tubes travelling outwards through space. These tubes are of equal strength, the strength of each corresponding to the atomic charge carried by a univalent atom. If a molecule of the gas through which the rays are passing gets dissociated into ions by the electric field produced by the tubes, one and only one of the tubes will get detached from the group and will be anchored by having its ends attached to the ions into which the molecule is dissociated. The dissociation of one moleeule, or the production of one positive and one negative ion, will withdraw just one tube from those in the group forming the Röntgen rays. Now Mr Rutherford's result, if we can extend it to all gases, shows that the production of each ion corresponds to a weakening of the Röntgen rays (by the same amount) whatever may be the gas from which the ion is formed. The intensity of the rays is supposed to be measured by the conductivity they produce in a standard gas at standard temperature and pressure. Thus Mr Rutherford's result may be expressed by saying that the weakening of the rays is proportional to the number of Faraday tubes stopped; and hence that the intensity of the rays is proportional to the number of Faraday tubes.

The Velocity and Rate of Recombination of the Ions of Gases exposed to Röntgen Radiation by E. RUTHERFORD, M.A., D.se. 1851 Exhibition Science Scholar, New Zealand University, Trinity College, Cambridge

From the Philosophical Magazine for November 1897, sero 5, xliv, pp. 422-440

AIR whieh has been exposed to Röntgen radiation preserves the power of diseharging positive and negative eleetrifieation a short time after the rays have ceased. It has been shown (J. J. Thomson and MeClelland, Proc. Roy. Soc., lix, 1896) that a plate ean be diseharged some distanee from the direct line of radiation from the Crookes' tube by blowing the Röntgenized air towards the plate. In this way it is possible to diseharge eleetrification after the air whieh has been exposed to the rays has passed through a tube several yards in length (J. J. Thomson and E. Rutherford, Phi!. Mag., November 1896). In the following paper the duration of the after-eonduetivity of air and other gases has been investigated, and from the data thus obtained the veloeity of the ions through various gases has been determined. Two distinet methods of determining the duration of the eonduetivity were used, both of whieh gave consistent results, viz.:(1) By blowing air at a known velocity along a tube, and testing the eonduetivity at different distanees from the point of action of the rays. (2) By applying an electromotive force to the gas at definite intervals after the rays have ceased, and measuring the quantity of electricity that passed through the gas. In the first method an aluminium cylinder was connected in series with a long brass tube 3 cm. in diameter and a metre long. The aluminium cylinder was placed over an opening in the metal tank in which the bulb and eoil were placed in order to eompletely screen off aU external electrostatie disturbances. All radiation to external points was carefully screened off by a thick lead covering over the aluminium cylinder. The aluminium tube was connected in series with a gasometer filled with air, the pressure of whieh could be regulated by weights placed on top. The air, before reaching the aluminium tube, passed through a tube filled with glass wool to remove the dust from the air. Three equal and similar insulated eleetrodes were

Velocity and Rate of Recombination of fhe Ions of Gases

133

placed at known distances apart along the brass tube. The brass tube was connected to earth, and one of the electrodes was connected to the electrometer. Both quadrants were connected together, and the whole was charged up to a high potential (in practice about 100 volts). The quadrants were then insulated from each other. When the rays were turned on, there was no effect on the electrometer until a current of air from the gasometer passed along the tube, when the movement of the electrometer needle showed that the air passing the electrodes was conducting. The rate of leak for each of the electrodes was tested in turn, and since the capacity in the electrometer circuit was nearly equal in the three cases, the rates of 1eak were proportional to the conductivity of the air at the electrodes. By noting the volume of air which passed from the gasometer in a given time, and knowing the diameter of the metal tube, the me an velocity of the current of air could be readily calculated, and therefore also the intervals of time taken by the current of air to pass from one electrode to the other. By altering the weights on the gasometer the velocity of the blast of air could be varied at will. The following table shows the way in which the conductivity of the air varies with the length of time after exposure to the rays. The first column gives the time taken by the current of air to pass from one electrode to the other, and the second column gives the ratio of the rates of leak of the electrodes. Time in seconds

Ratio of the rates of leak

0·13

0·75 0·61 0·57 0·39

0·22 0·28 0·65 2·4

0·11

lt will be seen that the conductivity falls off rapidly with the time, and after 2·4 sec. is only t of the original value. It is probable that some of the conducting particles give up their charge to the sides on their passage down the tube, but the correction for this is probably very small as there is no force acting on the charged gas tending to repel itself to the sides as is the case when free electrification is present. In a previous paper (1. J. Thomson and E. Rutherford, Phi!. Mag., November 1896) it has been shown that when agas is acted on by the Röntgen rays a steady state is reached when the rate of production of the ions by the rays is equal to their rate of recombination. If q be the number of conducting particles per c.c. produced per second by the rays, and N the final number, then when a steady state is reached

q=

where

IX

IXN2,

is a constant for any particular gas but varies for different gases.

134

The Collected Papers

0/ Lord Rutherford

When the rays are stopped the rate of diminution of the number of conducting particles is given by

dn

=

dt

-rxn2,

or if N is the maximum number and n the number after an interval t, 1 n

1 N

- - - = cx.1.

The time T for the number of conducting partic1es to fall to half their total number is given by 1 N= /XT.

Now the rates of leak at the electrodes in the tube are proportional to n since an electromotive force is applied to the gas sufficient to completely saturate the gas. From the experimental data of the rates of leak at different intervals we may compare the experimental results with those obtained from the formula 1 1 n N= /Xt. This can be best shown by plotting curves whose ordinates represent conductivities and abscissae time intervals. Fig. 1 shows such curves where the continuous curve represents the

Fig. 1

~ '('lodl)'

ami Rate

0/ Reromhination 0/ the ions 0/ Gases

135

relation between the conductivity and time as determined by experiment. and the dotted curve the theoretical relation which Is deduced. In order to compare the curves, one point on the experimental curve is chosen also as a point on the theoretical curve. In Fig. 1, the common point is P, where the conductivity has fallen to half its value. It will be observed that the two curves are in dose agreement, and the differences between them are weIl within the limits of experimental error. This shows that the formula dn dt = -a.n2

represents very closely the law of the rate of recombination of the ions. This agreement has been tested for a large number of experiments in which the intensity of the radiation varied widely, but in all cases the rate of decay was found to be in elose agreement with theory. Second M ethod

The method just described of determining the duration of the conductivity by blowing air along a tube could only be used for air on account of the large volurne of gas required for aseries of observations. The following method could be used for testing the rate of recornbination of the ions for different gases, and for widely different values of intensity of radiation. A glass bell jar was taken, the bottorn of which was covered with a plate of thin ebonite. A central electrode which reached nearly to the bottorn of the bell jar was insulated by passing through a paraffin stopper in the rnouth of the jar. The outside of the bell jar was coated with tinfoil wh ich had a metallic connection with the inside. The bell jar was placed on insulating blocks over a hole in the metal tank, which was covered with aluminium, and the Crookes' tube was placed in position beneath. In order to determine the after-conductivity the coil was turned on for a few seconds and then turned off. At definite intervals after the cessation of the radiation a large E.M.F. was applied to the outside coating of the bell jar and the quantity of electricity wh ich was given up to the central electrode was determined. A pendulum interrupter was used to break the battery circuit and to apply the E.M.F. at varying intervals. A heavy iron ball was suspended by a wire 315 cm. long, and the contacts were broken by a rod fixed to the bottom of the ball. Fig. 2 represents the arrangement of the experiment. The induction-coil primary circuit was completed through the brass lever AB, which pressed against a copper support C. When the pendulum struck the lever AB was knocked away from C and the eurrent broken. In order to apply a high E.M.F. at adefinite

136

Tlte Collected Papers oJ Lord RutlterJord

instant, recourse was had to a shunt method which worked very well in practice. A battery of one hundred small accumulators had one pole connected to earth and the other pole through a high carbon resistance R to the copper support E. A wire passed from E to the tinfoil on the outside of the conducting vessel H. The lever DF, which was kept pressed against E by a spring, was connected to a good earth. When the lever was in position, therefore, the vessel H was very nearly at zero potential, for the resistance of the lever DF and the earth connections was very small compared with the carbon resistance R. When the pendulum struck the lever DF the earth connection was broken and the vessel H was immediately charged up to the potential of the battery. By altering the distance between the two levers the time between breaking the battery current and applying the E.M.F. could be varied within limits. A

foi!

C B

o

&tter,yOf Cells

F

Pathof

Pelldv/um

L.

H

Earlh

Earlh

E~rfh

N

Fig.2 Immediately after the passage of the pendulum the lever DF was replaced, thus reducing H to zero potential again. In order to prevent a sudden deflection of the electrometer when the E.M.F. was applied to and removed from H, the electrometer was not directly connected to L until the lever had been replaced. The capacity M was introduced in order to prevent the potential of the circuit LM rising to any considerable extent when the E.M.F. was applied to H. It was found that applying and removing the E.M.F. alone produced no effect on the electrometer; but when the rays had acted on the gas inside the vessel L, there was always a deflection in the electrometer, showing that the electrode L had received acharge of the same sign as the pole of the battery. Since the E.M.F. applied was usually over 200 volts, a value sufficient to completely saturate the gas, the quantity of electricity that passed through the gas was proportional to n, the number of ions present in the gas at the instant the E.M.F. was applied, while an immediate application of the E.M.F. after the rays had ceased gave the total number N.

Ve/oeityaml Rate

0/ Recombination 0/ the Ions 0/ Gases

137

The rate of recornbination of the ions could thus be completely determined, as the times taken by the pendulum to pass from one lever to the other could be very approximately calculated. The following table shows the way in which the deflection of the electrometer varied with the time after the rays had ceased, when the bell jar was filled with dust-free air. Time in seconds

Deflection of electrometer

0·004 0·08 0·45 2

184 183 106 37 19

4

The value of the deflection is practically constant for nearly ri; sec. after the rays have ceased. After an interval of 4 sec. the air still possesses appreciable conductivity. Since the gases have widely different conductivities under the X-rays, it is to be expected that the rate of recombination is different for the various gases. The following table gives the times T for the number of ions to fall to half their original number. The intensity of the radiation was sensibly constant for all the gases. Gas

Time in seconds

Hydrogen Air Hydrochloric Acid Gas Carbon Dioxide Sulphur Dioxide Chlorine

0·65 0·3 O· 35 O· 51 0·45 O· 18

Conductivity

compared with air= I

0·5 1 11

1·2 4 18

There seems to be no dose connection between the values of T and the conductivities although, as a general rule, it may be taken that the value of T diminishes with increase of conductivity. It was found, however, that the rate of recombination was not always the same for the same gas with the same intensity of radiation, but depended largely on the amount of dust suspended in the gas as will be shown later in the paper. E*

138

The Collected Papers of Lord Rutherford Effect o/Intensity 0/ Radiation on the value

0/ T

The value of T for the same gas was found to depend largely on the intensity of the radiation. This is to be expected from theoretical considerations, for q= rx.N 2, and 1 N= rx.T,

therefore

1

yq;. .

T=-

If, then, rx. is a constant for the same gas in the same state, T varies inversely as Vfi. This relation was found to hold experimentally, for q is proportional to the intensity of the radiation, which varies inversely as the square of the distance from the bulb. F or example, for adefinite intensity of radiation the value of T was O· 25 sec. On placing a thick aluminium sheet below the conducting vessel, which cut down the intensity of the radiation to i- of its former value, the value of T rose to 0·6 sec. For weak radiation the values of T are much greater than for strong radiation. Time in seconds

Deftections

0·004

174

2 4 8

107 54

0·45

16

139

30 16

The above table shows the variation of the after-conductivity with time for very weak radiation. The value of T is about 3 seconds, and even after 16 seconds -h of the original number of ions are still uncombined. Air distant about a metre from an ordinary Crookes' tube possesses quite a measurable proportion of its conductivity for over a minute after the rays have ceased. Effect 0/ Finely Suspended Particles in a Gas on its Rate 01 Loss 01 Conductivity

It was found that the value of T varied greatly for the same gas for the same intensity of radiation. When, for example, chlorine was first passed into the testing vessel the value of T was 0·19 sec.; after standing for an hour the value of T rose to o· 3 sec., although the conductivity of the gas as

Velociry and Rate 0/ Recombination 0/ the Ions

0/ Gases

139

tested by the usual method was found to be unaltered. Freshly made gases were found, In all cases, to lose their conduetivity more rapidly than when they had stood undisturbed for some time. The eause of this effect was not at first clear, but later experiments on the influence of dust in the air led to the eonc1usion that it was due to the presence of finely divided matter, liquid or solid, in the freshly prepared gas. The value of T was found to be greater for agas that passed through a long tube filled with eotton-wool than if the cotton-wool were removed. This is probably due to the fact that the eotton-wool would not allow the small particles to pass through its pores. The presence of dust in the air was found to very greatly affeet the duration of the after-conductivity. As an example of the eIreet of dust we may give the following experiment: The cylinder was filled with air which had passed through a plug of glasswool, and then allowed to stand all night, and the value of the afterconductivity was taken next day without disturbing the gas. The quantity of electricity that passed through the gas after the rays had ceased gave a deflection in the eleetrometer of 70 divisions, and the value of T was 1 sec. Ablast of dusty air from a bellows was then sent into the cylinder and the deflection due to the after-effect fell immediately to 15 divisions, with a value of T of about 0·15 sec. When the air was allowed to stand, the after-effect gradually increased again to 35 divisions, with a value of T of about O· 5 sec. after an interval of about 10 min. Several hours elapsed before the after-effect rose to 60 divisions. This experiment shows what a variable quantity T is for the same gas, depending as it does on the amount of suspended matter in the gas. The effects observed in air and other gases seem to point to the conclusion that freely suspended particles greatly assist agas to lose its conducting property after the rays have ceased. Since the dust partic1es are very large compared with the ions, an ion is more likely to strike against a dust partic1e, and give up its charge to it or to adhere to the surface, than to collide with an ion of opposite sign. A positive ion striking a dust partic1e gives it a positive charge, and this is neutralized by acharge from a negative ion, and in this way the rate of loss of conductivity is much more rapid than if the loss of conductivity were due to collisions between the ions themselves. It seems probable that if agas could be obtained completely dust-free the rate of recombination which would be due entirely to molecular collisions would be very much slower than for ordinary air. When the rays act upon agas the number of ions per c.c. increases until a definite stage is reached, when the rate of production is equal to the rate of recombination. It is of interest to find the time that elapses after the radiation has commenced before tbis maximum is reached. In most of the experiments there were generally 50 breaks per second in the induction coil, so that, for the sake of calculation, we may very approximately suppose that

140

The Collected Papers o[ Lord Rutherford

the bulb was giving out rays uniformly, corresponding to the production of q ions per c.c. per second. The rate of increase of n is given by dn

dt = q - rx.n 2•

Solving tbis equation it is easily seen that the time t required for the production of n ions per c.c. in the gas is given by

1. t= ~ I-loge 2-yqrx.

(J~+n) rx.

•

'1_ n rx.

When the maximum number N is reached dn

and

dt =0,

q = rx.N2.

1 1 +r t = 2rx.N 10~ 1 - r'

Therefore .f

1

n

r= N.

Now, if T be the time taken for the number of ions to fall to half their number when the rays have ceased,

1 = rx.NT, T 2

therefore

1=-10 0

l+r

-Ot!l-r'

21

or

ii - 1

r--~+ 1

-

eT

The following table gives the values of r deduced for different values of t from the above equation:t T

0·125 0·25 0·5 1 2 4

r

0·123

0·245 0·462 0·736 0·878 0·998

Velocity and Rate of Recombination of the Ions oj Gases

141

For example, in the case of air elose to the Crookes' tube, T = O' 3 sec. Therefore, in 0·15 sec. the number of ions is nearly one-half of their final value. After an interval of one second the number of ions has practically reached the maximum value. For air distant about a metre from the source of radiation the value of T is much larger, and several seconds will elapse be fore the number approximates to the final value. Velocity of the ions It is a question of considerable interest to determine the velocity with which

the ions travel through agas under the influence of an electromotive force, as it indirectly gives us some information in regard to the nature and size of the carrier of electricity in conduction under the Röntgen rays. The method of determining the velocities was based on an investigation given in a previous paper (J. J. Thomson and E. Rutherford, Phi!. Mag., November 1896), where it is shown that

P

EUT

VI(I-t) =

P

where t is the current between two parallel plates, I cm. apart, when there is a potential-difference E acting between the plates. 1= the maximum current through the gas, when a saturating E.M.F. is applied. U = sum of the velocities of the positive and negative ions for a potential gradient of one volt per centimetre. T = time taken for the number of ions to fall to half its original value after the rays have ceased.

If E is chosen so that

t

J is small this equation very approximately reduces to

P

EUT

J=P'

Now, all the quantities in this equation can be measured, so that the velocity is readily calculated. In practice the value of E was taken of such an order that the value of the current t corresponding to it was about lo of the maximum current. t

The ratio I is not constant for the same gas for the same potential gradient, as it depends on the intensity of the radiation, and also on the rate of recombination of the ions, which is in turn largely dependent, for the same intensity of radiation, on the amount of dust and other solid matter in the gas. Although these quantities are variable, the velocity U is a constant, for

142

The Collected Papers 0/ Lord Ruther/ord

its value is quite independent of the value of T or of the intensity of the radiation. The steepness of the curves showing the relation between the current and the E.M.F. for air and other gases, which are given in a previous paper (Ioc. eit.) is very variable, depending as it does on whether the gas is freshly made or has stood undisturbed for some time. For adetermination of the velocity of the ions the value of T requires to be known with accuracy; but it is a difficult matter to determine T accurately with a pendulum interrupter, and moreover the intensity of the radiation from the Crookes' tube, on which the value of T largely depends, is very liable to change over a long range of experiments. Recourse was therefore had to a simpler method of determining T, which was found to give very consistent and reliable results. It has been shown that the final number of ions N is given by q= rx.N2, and also I N=rx.T, therefore

N T- - q"

Now if a very large E.M.F. is applied to the testing vessel the instant after the rays have ceased, the defiection of the electrometer is proportional to N, and q is proportional to the rate of leak per second for a saturating E.M.F. The ratio of these two quantities is thus readily determined and the value of T known.

Arrangement 0/ the Experiment Tbe testing vessel consisted of two parallel plates, the lower being of aluminium and the upper of sheet lead. These plates were separated by blocks of paraffin, which were melted together and formed the sides of the testing vessel, and at the same time insulated the top from the bottom plate. In order to measure the after-conductivity with accuracy fairly large plates were required. In the apparatus used the plates were 22 cm. square and 4·7 cm. apart. The rays passed through the lower aluminium plate and made the gas inside the vessel a conductor, but were completely stopped by the top lead plate. eare was taken that the radiation fell only on the centra] portion of the plate where the electrostatic field was sensibly uniform. The arrangement of the apparatus was the same as that given in Fig. 2, with the exception that the bell jar was replaced by the vessel with parallel plates. One pole of a battery of small accumulators was connected to the lower plate, and the electrometer to the upper. To determine N an E.M.F. of 200 volts was applied to the lower plate the instant after the rays had ceased. The two shunt levers were elose

Velocity ami Rate 0/ Recombillatioll

0/ fhe Ions 0/ Gases

143

together, and the method of breaking the current and applying the E.M.F. has been explained in the earlier part of the paper (p. 136). The value of q was determined by noting the leak per second for a saturating E.M.F. of 200 volts. The value of t was obtained by applying an electromotive force of two or more volts to the lower plate, and determining the rate of leak. The following table shows the values of T,

1-,

U obtained for the various

gases. T is expressed in seconds, U in centimetres per second. The values of

l are given for a potential gradient of -h of a volt per centimetre

between the plates. The values of

j are only approximate, and are deduced

from the observed values on the assumption that Ohm's law holds for electro-

,

motive forces small compared with the saturating values. The ratios of I were determined for different electromotive forces in the different gases. For

,

.

example, the value of i for hydrogen was 0·32 for 1·4 volts actmg between

,

the plates, while the value of I for sulphur dioxide was 0·066 for 9·3 volts acting between the plates. TADLE OF VELOCITIES

Gas

T

I

Velocity, U

Hydrogen Oxygen Nitrogen Air Carbonic Acid Gas Sulphur Dioxide Chlorine Hydrochloric Acid Gas

0·4 0·4 0·31

0·021

0·108

10·4 2·8 3·2 3·2 2·15 0·99 2 2·55

0·29

0·34 0·17

0·21

0·18

0·019 0·019 0·015

0·0033 0·0085 0·01

It will be seen from the above table that the velocities of the ions follow the inverse order of the densities with the exception of chlorine gas. The velocity of the hydrogen ion through hydrogen is nearly four times as fast as the velocity of the oxygen ion in oxygen. The ions of sulphur dioxide gave the slowest velocity, being only about /0 of that of hydrogen. The velocity of the ions through a gas was found to be independent of the amount of ionization of the gas. The velocities deduced from the two different sets of experiments, when the intensity of radiation in one case was six times that of the other, were found to be the same. In the one

144

The Collected Papers 0/ Lord Ruther/ord

case, therefore, six times as many ions per c.c. were present as in the other, but the velocity remained unaltered. There is thus no correction to be applied for the velocity of the ions in agas like chlorine, whose ionization is large compared with that of air. The fact that the velocity under a given small electromotive force is independent of the number of ions per c.c. in the gas, shows that the movement of the positive and negative ions does not produce any resultant electrostatic field between the plates. In the method which has been used for determining the velocity, it has been assumed that the conductivity of the gas is purely due to volume ionization of the gas. Perrin (Comptes Rendus, March 1, 1897) has, however, recently shown that the rate of leak between two plates is made up of two parts, one due to the volume ionization of the gas, and the other due to surface ionization at the surface of separation of the metal plate, on which the radiation impinges, and the surrounding gas. The rate of leak due to the surface action is quite comparable with that due to the volume ionization, when the plates are 1 cm. apart, especially in the case when the electrodes are of silver, gold, or zinc. It seems probable that the gas elose to the surface of the plate on which the radiation falls has a much greater density of ionization than the gas between the plates, and since this increase is confined to a very thin layer elose to the electrode, that the rate of combination for the surface ions is far more rapid than for the Röntgenized air some distance from the surface. The existence of this effect would tend to increase the rate of leak q; while the number of ions, N, as determined by the application of an E.M.F. a short interval after the rays had ceased, would not be appreciably affected, since the ions near the surface probably recombine with great rapidity. The value of T which is obtained from the equation T

=

N would thus be too q

small. The correction for the surface ionization is probably, however, very small, for the lower plate of the testing vessel was of aluminium-a metal which does not appreciably exhibit the phenomena of surface ionizationand the upper of lead, in which the effect is slight. In addition to this the plates were nearly 5 cm. apart, so that the volume effect was very large compared with that due to the surface. The velocities which have been determined are the sum of the velocities of the positive and negative ions, but we have so far given no direct experimental evidence to show whether the velocities of the positive and negative ions are the same. In the case of air, an experiment which will now be described seems to show that the velocities of the two ions are equal or very approximately so. In previous determinations the ca1culation of the velocity of the ions has depended on the truth of an equation which has been experimentally verified as far as possible, but in the case of air the velocity may be obtained by a method not involving any theory depending on the rate of recombination of the ions.

Velocity amI Rate 0/ Recombination

0/ fhe Ions 0/ Gases

145

Two large plane plates, A and B (Fig. 3), were placed parallel to one another, 16 cm. apart, on insulating blocks C and D. The bulb was so arranged, in regard to the plates A and B, that the radiation fell on the plate A and half of the volume of air between A and B. No radiation reached the air to the left of the dotted line EF in the figure, which was 8 cm. from either plate. The plate A was connected to one terminal of a large battery of storage ceUs, the other pole being connected to earth. The plate B was connected through a contact lever, LM, mounted on an insulating block, to one pair of quadrants of the electrometer, the other being connected to earth. The pendulum interrupter was so arranged as to make the current in the primary of the induction coil, to break the electrometer-circuit by knocking away the lever LM, and then to break the battery-circuit shortly F

B

A

D

farth

D

&

7ö &fhry Qf CBI/s

c

L

Lever

D

Fig.3 afterwards. For this two other shunt levers were required, which are not shown in the figure. A condenser, N, was introduced into the electrometercircuit to increase its capacity. A steady difference of potential of 220 volts was applied between the two plates. When the bulb was excited the ions on one side have to travel over a distance of 8 cm. before they reach the plate B. The object of the experiment was to determine the interval between the starting of the rays and the arrival of the ions at the plate B. It was found that there was only a small defiection of the electrometer until after adefinite interval had elapsed, when the deft.ection increased rapidly. This was taken as the instant when the ions had reached the electrode. The deft.ection of the electrometer was proportional to the quantity of electricity that had passed from A to B during the time between making the current and breaking the electrometer-circuit. The electrometer itself was not connected with its circuit until after the contacts bad been broken.

The Collected Papers 0/ Lord Ru/her/ort!

146

It was found that after the rays had been aeting for 0·36 of a seeond the electrometer defieetion eommenced to inerease rapidly. A potential difference of 220 volts was aeting between the plates, so that the eleetromotive intensity was 13·75 volts per ern. In the time 0·36 of a seeond, therefore, under a potential gradient of 13· 75, the ions have travelled over a distanee of 8 em. This gives a velocity of the ion of 1. 6 em. a seeond, and eorresponds to the velocity of a positive or a negative ion, for the time taken was found to be independent of the sign of the ion. The sum of the veloeities of the positive and negative ions is therefore 3·2 em. per second -a result agreeing with that determined for air by a eompletely independent method. The elose agreement between these two results affords strong evidenee of the truth of the theory by the aid of whieh the velocities of the ions of different gases have been obtained. The velocities with whieh the ions move through the gas are immensely greater than the velocity of the ions in the electrolysis of liquids. The velocity of the hydrogen ion in nearly pure water is 1·08 cm. per hour under a potential gradient of 1 volt per cm.; so that the velocity of the hydrogen ion in eonduetion under the Röntgen rays (10'4 cm. per sec.) is over 36,000 times as great. From eonsiderations based on the kinetie theory of gases the veloeity of a small eharged body moving through the gas under the infiuenee of an eleetrie field may be determined.

Let

e be the charge on the positive ion; -e on the negative; mt m2 the masses of the positive and negative ions; X eleetromotive intensity; /(1 /(2 the quotient of pressure by density for the positive and negative ions.

Then, assuming that the partial pressure of the dissociated gas is small, the veloeity Ut of the positive ion is given by Ul

=

Xe D.,

mt/(l

where D 1 is the eoeffieient of interdiffusion of the positive ions and the undissociated gas. (J. J. Thomson, Brit. Assoe. Report, 1894, and Art. 'Diffusion', Encyclopaedia Britannica.) The velocity U2 of the negative ion is given by U2

=

~D2'

m2/(2

where D2 is the eoeffieient of interdiffusion of the negative ions and the gas.

/"e/ocity ami Rate of Recomhination 0/ fhe IOlls 0/ Gases

147

The sum of the velocities of thc ions is thus given by =

U

Ut

+

U2

xe xe = - D t -!--D2• milet

m2 1e2

When thc ions are of cqual mass

2xe

u=-D. mle

We have no mcans of determining D, the coefficient of interdiffusion of the ions into the gas, nor the mass and charge of the carrier. If, however, we assume that the ion carries the same charge that it does in the electrolysis of liquids, we can theoretically deduce the velocity with which a moleeule carrying the atomic charge would move through the gas. D then becomes the coefficient of interdiffusion of agas into itself, and is given by thc relation D

P= 1'5435-, p

where p- is the coefficient of viscosity of the gas and p its density. For hydrogen D Therefore

=

1· 7 and the value of

!.. = m

104 approximately.

e

- = 10-6'

mle

and

x=

therefore

u = 340 cm.

108 for 1 volt per cm.

The velocity of a molecule of hydrogen through hydrogen and carrying an atomic charge is thus 340 cm. per sec., while the experimentally determined value is only 10·4 cm. per sec. The disagreement of theory and experiment seems to point to the conclusion that either the charge is less than the charge carried by an ion in ordinary electrolytes. or that the carrier is larger than the molecule. We have not sufficient experimental evidence to decide between the two suppositions, but some experiments on the velocity of the ions in gases like hydrochloric acid and sulphuretted hydrogen seem to point to the conclusion that the carrier is larger than the molecule. We may suppose that when an ion is liberated it becomes a centre of an aggregation of molecules. This cluster of molecules is in stable equilibrium under the attractive force due to the charge on the ion and the size of the cluster is determined by the intensity of the bombardment of the molecules of the gas on its surface. Such an hypothesis would explain

The Collected Papers 0/ Lord Ruther/ord

148

the observed fact that the positive and negative ions of gases like hydrochloric acid and sulphuretted hydrogen have equal velocities; for the size of the aggregation is dependent only on the charge on the ion for the same gas, and is, therefore, the same for the positive and negative ions. In the electrolysis of hydrochloric acid we know that the velocity of the hydrogen ion is much greater than that of the chlorine ion, so that there is an essential difference between the carriers in the two cases. Proceeding on this assumption of the formation of clusters round a central nucleus, we may readily determine the diameter of the clusters to give the observed velocity of the carrier. Ir D is the coefficient of interdiffusion of a molecule of diameter 0'1. the coefficient of interdiffusion of the carrier of diameter 0'2 through the gas is given by

0'1 2

~D. 2

Therefore _

U-

Xe mK

(0'1)2 D. 0'2

The value ~ is the same for the molecule as for the cluster. For hydromK

gen, then, the ratio of the diameter of the cluster to the diameter of the molecule is given by

J

340 =5·7.

10·4

For oxygen D = 0·21 and, assuming its charge is double that of hydrogen, we get a theoretical value of the velocity of 85 cm. per sec. This corresponds to a cluster of the diameter of 5· 5 molecules. In the case of chlorine, assuming the charge on the ion is the same as that on hydrogen, we get a cluster of diameter 2·7 molecules. We see, therefore, that to explain the observed results the carrier need not be greater than five times the radius of the molecule. Further experiments, which are not yet completed, have been made to find the velocity of the ions of agas conducting under the infiuence of the radiation given out by uranium and its salts. It can be shown that the velocity of the ions in the conducting gas is the same as when the gas is acted on by Röntgen radiation, so that the carrier in the two cases is identical. Further results, however, must be reserved for a future paper. In conc1usion, I desire to express my thanks to Professor J. J. Tbomson for many valuable suggestions during the course of tbis investigation. Cavendish Laboratory July 19, 1897

The Discharge of Electrification by Ultra-violet Light by E. R UTHERFORD, M.A., B.se., 1851 Exhibition Scholar, New Zealand University, Coutts Trotter Student, Trinity College, Cambridge

From the Proceedings oj' (he Cambridge Philosophical Society, for 1898, ix, pt. 8, pp. 401-416

THE general action of ultra-violet light on the discharge of electrification has been investigated by many different experimenters. Hertz* in 1887 first drew attention to the action of ultra-violet light on a spark gap. Wiedemann and Ebertt showed that the kathode was the seat of this action and investigated the general effect on high potential discharges. Hallwachst and Righi§ observed the fact that zinc and other metals became positively electrified under the action of ultra-violet light. These results were extended by Elster and Geitelll who have published aseries of papers on the effect of ultra-violet light in causing discharge under various conditions and have also -U investigated the action of a magnetic field on the discharge at low pressures. Stoletow** investigated in detail the relation between the current and electromotive force for the discharge at low voltages and at different pressures. Most of these papers have dealt with the general character of the discharge, but the subject of the nature of the conduction and of the carrier that discharges the e1ectrification has not been specially attacked. In a very interesting paper Lenard and Wolfftt investigated the effect of a surface, on which ultra-violet light fell, on the condensation of a steam jet in the neighbourhood, and their results led them to the conclusion that many bodies were disintegrated under the action of ultra-violet light and that the particles torn off became nuclei for the condensation of the steam jet. In the light of more recent experiments these results are, however, capable of other interpretations. R. v. HelmholtzH has shown that a steam jet is acted on when chemical

*

t Wied. Annal, xxxiii, p. 241, 1888. § Phi!. Mag., xxv, p. 314, 1888. 11 Wied. Annal., xxxviii, pp. 40, 497; xxxix, p. 332; xli, pp. 162, 166; lii, p. 433; Iv, p. 684. ~ Wied. Anna!., xli, p. 166,1890. ** Journal de Russ. Phys., xxi, 1889; Journal de Physique, ix, p. 468, 1890. tt Wied. Annal., xxxvii, p. 443, 1888. :j:t Wied. Annal., xxxii, p. 1, 1887; xl, p. 161, 1890. Wied. Annal., xxxi, p. 983, 1887.

t Phi!. Mag., xxvi, p. 78, 1888.

150

The Collected Papers 0/ Lord RutherJord

action is going on in its neighbourhood. Richarz· has shown that Röntgen rays produce condensation in a steam jet, and Wilsont has recently observed that ions produced under the action of uranium and Röntgen radiation become, under certain conditions, nuc1ei for the condensation of water around them. He has also demonstrated the important fact, which appears to have been overlooked, that ultra-violet light produces c10uds in ordinary moist air quite independent of any solid body in its neighbourhood. The presence of this effect must have complicated the effects of Lenard and Wolff, and the more general results on the properties of ions in producing condensati on seem to show that possibly their results may be ascribed to the presence of free gaseous ions rather than disintegrated partic1es of metal. It is the object of this communication to give results of investigations on the nature of the carriers of the negative electrification produced under the action of ultra-violet light, and to show that probably the greater part of the electrification is carried by gaseous ions produced at the surface of the negatively electrified plate. In order to obtain a discharge with ultra-violet light, the light must fall on a negatively electrified surface. There is no discharge produced by allowing the light to fall between two plates without impinging on either. In this respect the action of ultra-violet light is very different to Röntgen and uranium radiation, which produce a volume ionization of the gas through whieh they pass. The question whether there is any volume ionization of a gas through which ultra-violet light passes was investigated by Henryt who tried the vapours of iodine and methyl iodide both of which are very powerful conductors under Röntgen rays, but with a negative result. The result of Wilson that a cloud is formed in moist air with strong ultra-violet light renders it possible that there is a slight volume ionization of the gas through which the light passes, but the effect appears to be too smaU to be determined by electrical means, and in all later experiments it is assumed that the surface of the negatively charged plate is the seat of the action of the ultra-violet light discharge. If we allow ultra-violet light to fall on a negatively electrified surface, e.g. a polished zinc plate, since the body is slowly discharged, it seems probable that ifthe discharge is due to the convection ofthe charged particles, these charged partic1es can be blown away by directing a sufficiently rapid blast of air across their path. This has been experimentally shown by Zeleny §, who showed that a negatively charged gas can be obtained by blowing past the negatively charged plate when the ultra-violet light was aeting. This gas has similar properties to the charged gas 11 obtained by the separation of ions in Röntgen ionization, for they readily give up their charge and refuse to pass through a plug of cotton-wooI. I had independently observed the same fact, and had also investigated t Camb. Philos. Soc., ix, pt. VÜ, 1897. * Wied. Annal., lix, p. 592, 1896. § Phi!. Mag., March 1898. t Proc. Camb. Phil. Soc., ix, pt. vi, 1897. 11

Rutherford, Phil. Mag., April 1897•

n,e Discharge

0/ Electrijication by Ultra-violet Light

151

the effects of blowing currents of air by electrodes, especially from the point of view of determining the velocity of the carrier of electricity. Before entering on the general results, it is necessary to draw attention to the phenomena observed by Blondlot and Bichat.* They found that if an insulated zinc plate was acted on by ultra-violet light, when all the conductors in the neighbourhood were connected to earth, the potential to which it could be raised was increased six or seven times by allowing ablast of air to impinge on the plate. They found that this action was independent of dust and moisture. It is easy in this way to raise the potential of a plate of amalgamated zinc to 15 volts in a few minutes, although the potential to which it could be raised under the action of the light alone was less than 2 volts. I have also found that the rate of leak of a body charged negatively is much more rapid when ablast of dust-free air is directed against it. A plate of polished zinc, charged to -8 volts, placed at a distance of 10 cm. from a neighbouring plate gave a rate of leak twelve times as fast as under the action of the ultra-violet light alone. The blast seems to assist the electromotive force in removing the negative charge. The presence of this action must be taken into account in all cases where currents of air impinge on negatively electrified surfaces. The following figure (Fig. 1) shows the general arrangement of the experiment to show the effect of blowing a current of air by a negatively electrified plate on which ultra-violet light is allowed to fall. A

L

Q,

E~rfJI

M

Q2

B

G

&rtll

W

c

T

Fig.l

Ablast of air from a bellows or a gasometer passed between two parallel plane electrodes Band C, and then through an insulated aluminium cylinder T. The air before reaching the plates was free from dust as far as possible by passing through a bulb G tightly packed with cotton-wool. An arc light A was used as a source of ultra-violet light. The light after passing through a quartz plate Q1, which covered an opening in a metal screen LM surrounding

* C.R., xvii, p. 29,1888.

152

The Collected Papers

0/ Lord Ruther/ord

the arc, passed through a second quartz plate Q2, through the fine wire-gauze Band then impinged on the metal plate C, which was generally of polished or amalgamated zinc. The broken lines show the position of insulators by which the gauze B, the plate C, and the tube T were all insulated from one another. An insulated wire W passed centrally down the tube T. In a particular experiment the plate C was 4·2 cm. long, 1· 5 cm. wide and O· 8 cm. from the gauze B. Since the area of the rectangular orifice through which the air passed was only 1· 2 sq. cm., velocities of the air of 300 or 400 cm. per sec. could be readily obtained. Experiment I

C was connected to the negative pole of a battery of 8 volts, the other pole being connected to earth. The gauze B was connected to one pair of quadrants of an electrometer, the other pair being connected to earth. All other parts of the apparatus were earthed. When the arc light was acting the plate C lost a negative charge which passed over to the gauze Band the electrometer needle showed a movement corresponding to 60 divs. per min. When a rapid current of air was directed between the plates, the leak to the electrometer was completely stopped. This showed that the carriers of the negative charge, which had left C, had been blown out by the rapid current of air. If C was charged to -24 volts the rate of leak from C corresponded to 170 divs. per mine but where the blast was in action the rate of leak was reduced to 8 divs. per min. or less than -lo of the charge which escaped from C reached the gauze B; when C was charged to higher voltages and the blast kept in action, a still greater proportion of the charged particles reached the gauze without being blown out. The number of carriers that reached the gauze could be raised or lowered by diminishing or increasing the velocity of the blast, the other conditions remaining the same. Experiment 2

We have seen that a whole or part of the charged carriers can be prevented from reaching the gauze B. It is now necessary to show what becomes of the carriers after being blown out from between the plates. The plate C was charged to -24 volts and the gauze connected to earth. The aluminium cylinder T was connected to one pole of a battery of 30 volts, and the wire

Thr nischal'ge 0/ Electl'ificalioll by Ultl'a-violel Light

153

W connected to the electrometer. There was no deflection of the electrometer if the arc alone was acting. Ir T was charged negatively, then when a current of air was sent between the plates the wire W became charged negatively. Ir T was charged positively there was no appreciable leak to W. This shows that a negatively charged gas has passed into the cylinder wh ich in the first case lost its charge to the central wire and in the other to the tube T. A very convenient means of testing the whole charge conveyed with the current of air is to place a plug of cotton-wool inside the tube, which has the property of completely discharging the electrification carried with the air.

Experiment 3. Discharging power o[ a metal tube If the aluminium tube T which was 30 cm. long and 1·1 cm. diameter was connected to the electrometer and the central wire removed, it was found that the gas gave up part of its charge to the tube in its passage along it. I n an experiment where the velocity of the air along the tube was about 150 cm. per sec., t of the gas was discharged in passing along the tube. Since the electrified particles tend to repel one another to the side such a discharge is to be expected. If the volume density of the electrification were uniform over the cross section of the tube, an experiment of this kind on the discharging power would allow us to calculate the velocity with which the carrier moves under its own repulsion, but this condition is here not fulfilled.

Experiment 4 All the electrification blown out between the electrodes C and B can be collected. In one experiment the rate of leak of C was observed with the blast in action. Part (1) of the charge which left C was given up to the gauze B, apart (2) was discharged on the tube T and the part (3) which escaped was caught in a cotton-wool collector. The sums of the rates ofleak to (1), (2) and (3) were found to be very nearly equal to the rate of leak of C.

Velocity o[ the carrier The general experiments on the effects of a current of air between two electrodes when the ultra-violet light is acting may be simply explained by supposing that the negatively charged particles which escape from the surface C, travel towards the gauze B with a velocity proportional to the electromotive force between the two plates. Let u = velocity of the charged partic1e for a potential gradient of 1 volt per cm., d = distance between the plates, I = length of the plates, l' =- difference of potential in volts between the plates.

154

Tlze Collected Papers

0/ Lord Rutlzelfol'd

The time t taken for the carrier to pass from one plate to the other is therefore given by d d2 t=-=v uv

u'd

Now if P = velocity of blast, the time I} taken for the current of air to pass between the electrodes is given by 1 t}

= p'

Assuming that the velocity of the air blast is constant over the cross seetion of the plates, if tl is less than t none of the charged carriers whieh leave C can reach the gauze B, but they will be all blown out with the current of air. We may suppose the carrier describes a diagonal path between the plates due to the resultant of the two impressed velocities at right angles to each other, and unless tbis L B A diagonal path cuts the gauze B the carrier will escape. Let AB, CD (Fig. 2) be the two plates. Suppose the carriers to be p D c produced uniformly along CD by the action of ultra-violet light. A Fig.2 carrier starting from C travels along the diagonal path CL and gives up its charge to the plate AB. Draw BP parallel to LC, meeting CD at P. We see that a carrier starting from P will just reach B. All carriers starting from the right of P will not reach AB but will be carried out by the eurrent, while those to the left of P will give up their charges to AB. The ratio p of the number of carriers blown out to the total number that leave CD is given by PD p= CD' the distance PD = P. t, where t is the time taken for the carrier to cross over between the plates; d2 :. PD=P X-, U'V

:. u =

P d2

7' vp'

The ratio p is determined experimentally and P can be measured, therefore the value of u can be at once obtained. Experiments performed in this way give a value of the velocity of the carrier at normal pressure of about

1. 5 cm. per sec.

Ihr Discharge

0/ Electrification by Ultra-violet Light

155

for a potential gradient of I volt per cm. This method is, however, probably not capable of accuracy, on account of the variation of the velocity of the air across the cross section of the plates and accidental disturbances due to the violent eddying motion of the air when velocities of the order of 300 or 400 cm. per sec. are used. The method is also practically restricted to the case of air on account of the large amount of gas required for an experiment, so that I was led to devise a more general and satisfactory method of determining the velocity of the carrier. In Fig. 3 the general arrangement is shown: a glass bell jar was fixed on a base plate CD of zinc through which a circular opening EF was cut. In M

R

N

c

L

A

F

E

8

D

EE EE

5

Fig.3 thc top of thc vessel there was fixed an ebonite stopper through which passed a rod L carrying at one end a polished metal plate AB. This plate AB was fixed to a ball-and-socket joint, and could be levelled by screws passing through a plate fixed to the rod. The plate AB could be raised or lowered from the outside of the vessel by a screw. The bell jar was fitted down to the base plate with sealing-wax, and the whole was placed on insulating blocks over a source of ultra-violet light S, which was either an arc light or a spark gap. The ultra-violet light first passed through a quartz plate Qh fixed over an opening in a large metal screen, then through the quartz plate Q2 covering

156

The Collected Papers

0/ Lord Rutherford

the opening EF, and then fell on the metal plate AB. The wire-gauze performed the double function of allowing part of the light to pass through and yet acting electrically as a plane metal surface. The plate AB was generally of polished zinc and was accurately levelled to the base plate. A delivery tube T was let into the base plate and the whole vessel was made airtight to a110w of exhaustion. In order to determine the velocity of the carrier, the rod L which was insulated by means of the ebonite stopper was connected to one pair of quadrants of the electrometer by the wire M, the other pair being to earth. The base plate CD was then connected to one terminal of a lOO-volt transformer, worked from the town mains, the other terminal of which was connected to earth. When the base plate is charged positively, the plate AB is charged negatively by induction, and the negative carriers, set free under the action of the ultra-violet light, start travelling towards the base plate under the influence of the electromotive force acting between the two plates. lf the plate AB is close to CD, a large number of the carriers are able to reach CD, and give up their charge before the electromotive force is reversed. All the carriers distributed between the plates at the instant of reversal travel back to the top plate, and since the top plate is charged positively by induction no more carriers are produced during that half alternation. Experimentally it was found that there was no leak to the top plate when the base plate was negatively charged and the ultra-violet light was in action. We see, therefore, that when the plates are close together the plate AB loses a negative charge. This rate of leak will evidently decrease as the distance between the plates is increased, until a certain distance is reached, when the plate AB shows no loss of charge, although the ultra-violet light and alternating E.M.F. are both acting. When the plate is at this distance the first carriers liberated when AB becomes negative by induction are able to travel nearly to the base plate, but before any can give up their charge, the E.M.F. is reversed and they travel back to the plate from which they came. All distances greater than this give us no rate of leak, but it is the object of the experiment to determine the shortest distance between the plates for which AB shows no loss of charge. This point is in general fairly sharply defined, as the table below shows, which gives the rate of leak at different distances from the base plate. Turns of Screw

1 turn 3 5 7 8 9

9·5

Deflection of Elect. in 30 sec.

off scale 290 divs. 220 120

35

2

o

ThC' Dischmy;(! nf Electr,(icatioll hy Ultra-\';o!et L(f?ht

157

N ine turns of the screw is approximately the shortest distance between the plates for which the loss of charge is negligible, and the numbers show how rapid is the decrease of deflection between seven and nine turns.

Let u = velocity of the carrier for a potential gradient of I volt per em. d = shortest distance between the plates for which the leak of AB is zero. T = time of complete alternation of the transformer. It is assumed that the E.M.F. at any time of the alternation is given by Eo sin 277

f' where E is the maximum value of the alternating E.M.F. o

The distance dx traversed by an ion in the time dt is given by dx

Eo. t d =d sm 277 T' U· t.

Now the distance passed over by the carriers wh ich first set out is equal T · h . to d m t e tIme 2' Therefore integrating both sides, we obtain

Eo T d=-·-·u d 77 ' or

1I

d 277 = EoT'

The distance d is determined experimentally, and E o and T are constants of the transformer circuit, so that u is known. Since the electrometer circuit is insulated before the E.M.F. is applied to the lower plate, the true potential between the top and base plate is less than the potential of the lower plate on account of the induction effect between the two plates. In consequence of the rapidity of the alternations, the electrometer needle shows no movement due to the induction effect, so that it is necessary to determine the correction to be applied when the needle of the electrometer remains at rest. This may be simply done as folIows: The two quadrants of the electrometer are insulated from each other, and then a steady E.M.F. is applied to the base plate. The electrometer needle is deflected, but is brought back to its original position by applying a suitable E.M.F. to the other pair of quadrants. Let r be the ratio of the applied potential to the potential V of the base plate. The potential of the electrometer circuit is r V and the true difference of potential between the plates is (I - r)V. The correction is thus made by putting (1 - r)Eo instead of E o in the equation that determines the velocity.

158

The Collected Papers 0/ Lord Ruther/ord

The following is an examp1e of adetermination of the velocity of an air ion travelling through air: d= O·8cm., T = Jl sec. Reading of alternating voltmeter 95, the maximum value of the E.M.F. Eo is therefore given by

Eo =95.y2, d2."

u=-=1·37. EoT

The correction for induction between the plates was 5 per cent, so that the final value is u = 1·44 cm. per sec. A large number of determinations have been made for the velocity of the air ion, and the mean of the results gives a velocity of about 1·45 cm. per sec. In these experiments two sources of ultra-violet light have been used,

viz. an arc lamp and a spark between two zinc terminals. For the spark two

large Leyden jars were charged up by a Ruhmkorff coil and discharged through a spark gap of about 0·8 cm. In order to get rid of electrostatic effects, the jars, coil, and spark gap were a1l placed inside a metal tank connected to earth. The opening above the sparking terminals was tightly c10sed by a quartz plate, on both sides of which fine wire-gauze, connected to earth, was stretched. This arrangement was found to screen thoroughly from electrostatic disturbances. For experiments on the velocity of the carrier the arc light is not nearly so satisfactory a source of ultra-violet rays as the spark discharge. The arc as a rule gives out more intense radiation but it has the drawback of rapidly raising the temperature of the plates and of the adjacent air. In consequence of this the velocity of the carrier is changed, and the determination of the least distance of no appreciable loss of charge cannot be made with certainty. In this respect the spark light is far more satisfactory and it is also a more constant source of rays, but it has the disadvantage of being an intermittent source of light. A source of error in the determination of the velocity, which is difficult to avoid, lies in the irregularity of the E.M.F. from the town supply. During the daytime, the load on the machines was light, and there were often rapid alterations in the E.M.F. and period of the alternations. Another source of error also lies in the assumption that the electromotive force curve of the alternator is a sine curve. It is intended to continue these experiments and to use instead of the transformer an alternating electromotive force produced by reversing the sign of a steady E.M.F. by means of a suitable revolving

"flic Discltarge

0/ Eleetrijication by Ultra-I'iolet Light

159

commutator. lt is hoped in that way to obtain a very accurate value of the velocity under varying eonditions of pressure and temperature.

Effects 0/ Different Metals Ir the theory that the diseharge of eleetrification is due to the disintegration of negatively charged particles is true, we should expect to obtain different velocities of the carrier according to the metaIon which the ultra-violet light falls. This point was tested by replacing the zine plates by similar sized plates of aluminium, lead, eopper, amalgamated zinc and a zinc plate covered with mercury sodium amalgam. Although there were differences in the rate of leak of these metals due to their different degrees of sensitiveness to ultra-violet light, it was found that the distance of no loss of charge was nearly the same in each case. The arc light was used for this series of experiments and, allowing for errors of experiments, it seems to be true that the velocity of the carrier is independent of the metaIon which the light falls. This seems to show that the carrier is produced from the gas near the plate and not from the metal itself.

Effect 01 Electromotive Force The velocity of the carrier, i.e. the velocity for a potential gradient of I volt per cm., is independent of the potential of the surface from which the carrier sets out. Experiments were also made to determine the value of the velocity for different values of the E.M.F. To test this the ultra-violet light from an arc was allowed to shine through a wire-gauze on to a metal plate wh ich was much larger than in the experiments in the bell jar. By using a suitable transformer, potential differences, which were measured by a Thomson's Multicellular Voltmeter, of 365 and 700 volts beside the usual 95 volts were obtained. Correction ror induction

Voltage

d

95 volts 365 volts 700 volts

0·775 cm.

1·53 cm. 2·07 cm.

14 per cent.

10 per cent.

7 per cent.

When the corrections for induetion are made, the radius of the velocities obtained for unit potential gradient are Voltage

Velocity

95 365

1·46 1·43 1·33

700

We see from the above table the velocities are very approximately proportional to the electromotive intensity between the plates. The actual velocity of the carrier when the E.M.F. was 700 volts was 450 cm.. per sec.

160

The Collected Papers 0/ Lord Ruther/ord

From considerations based on the kinetic theory of gases it can readily be shown that a carrier of molecular dimensions carrying an atomic charge attains a limiting velocity in a uniform field after a very short interval of time and it then moves uniformly. No correction therefore need be made for the interval that elapses before the carrier reaches its limiting velocity.

Effect 0/ Pressure on the Velocity 0/ the Carriers The effect of pressure on the velocity was investigated by means of the apparatus shown in Fig. 3. The following table shows the results obtained for air: Pressure

765mm.

323 162 140 95

58

34

Velocity

1·4cm. 3·36 7·3 7·8

11·9

20·3

33·6

These results show that down to apressure of 34 mm. the velocity of the carrier is inversely proportional to the pressure. As Stoletow has shown, it was found that the rate ofleak for a given E.M.F. increased with the diminution of pressure so that the results were capable of fair accuracy even at low pressures. The plates were never separated by more than 1· 5 cm. but the electromotive force employed was much smaller for the lower pressures. In the experiment for the pressure of 34 mm. only lo of the voltage of the transformer was used. If the law that the velocity is inversely proportional to the pressure holds for low pressures at apressure of 1 mm. with a potential gradient of 1,000 volts per cm. the carrier would travel with a velocity of 1·4 X 760 X 1000 = 106 cm. per sec. These experiments on the variation of the velocity with the pressure afford very strong evidence that the carrier of the charge is of molecular dimensions. According to the kinetic theory of gases the velocity of a charged particle of molecular dimensions varies inversely as the pressure. * If, however, the partic1e is large compared with a molecule, the velocity of the carrier is dependent only on the viscosity of the gas which is independent of the pressure within limits. The conclusion to be drawn from these results is that the carriers must be very small and comparable in size with a molecule. Many of the interesting results of Stoletowt on the relation between the • J. J. Thomson, Bril. Assoe. Report, 1894.

t Journ. de Phys., ix, p. 468, 1890.

The DIscharge vi Electl'iftc:atioll by Ultra-rivlet Light

161

current and the E.M.F. for different pressures on ultra-violet light conduction can be generally explained on the hypothesis that the rate of production of the carriers is a function of the pressure, and that the rate of escape of the negative ions is dependent on the velocity of the carrier. It is not my intention here to enter into a detailed discussion of his valuable results but the fact that the current through the gas at very low pressures (about 0·002 mm.) is independent of the pressure seems to show that either there is a slight disintegration of the electrodes by ultra-violet light or that the presence of mercury vapour is responsible for the action. Velocity in Different Gases

The velocity of the carrier depends on the gas surrounding the plates. Hydrogen and carbonic acid were used and were weIl dried before passing into the apparatus. Gas

Velocity

Air Hydrogen Carbonic acid

3·9

Ratio, air=l

1·4

1

2·8

0·78

0·56

These results were obtained with the use of the arc light when the heating of the gas prevented the velocities being obtained with as much accuracy as desired. The hydrogen ion in hydrogen travels nearly three times as fast as the air ion in air, and five times as fast as the carbonic acid ion in carbonic acid. It has been shown that the carriers which are produced under the action of ultra-violet light have very similar properties to the ions which are produced by Röntgen radiation. It is also an interesting result that the velocity of the air ion in Röntgen conduction is not very different from that obtained in ultra-violet light conduction. In a previous paper· it has been shown that the sums of the velocities of the positive and negative ions in air, hydrogen and carbonic acid are 3,2, 10·4 and 2·15 cm. respectively. Assuming for the purpose of comparison that the velocities of the positive and negative ions are equal, the table below shows the relative values of the velocities for the negative ions: Velocity in Velocity in Röntgen Ultra-violet light Conduction Conduction

Gas

Air Hydrogen Carbonic acid

1·6cm. 5·2cm. 1·07 cm.

1·4

3·9

0·78

Considering the widely different methods used for the determination of • Phil. Mag., November 1897. F

162

The Collected Papers of Lord Rutherford

the velocities in the two cases, the comparative agreement shows that possibly the carrier is the same in the two cases or in any case not greatly different. The results obtained in this paper seem to show that the gas near the surface of the negatively electrified plate is ionized under the action of ultraviolet light. The positive ion gives up its charge to the plate and the negative ion is repel1ed from the plate. This ion travels through air at normal pressure and temperature with a velocity of about 1 ·4 cm. per sec. for a potential gradient of 1 volt per cm. The velocity of the ion varies inversely as the pressure and is different for different gases. It is intended to continue these investigations on the velocity of the ions and on the general phenomena of ultra-violet light conduction. Cavendish Laboratory February 21. 1898

Some Reminiscences of Professor Ernest Rutherford during his time at McGill University, Montreal 1 by

H. L. BRONSON

EARLY in the spring of 1904, I had the good fortune to attend a Yale University Extension Lecture by Rutherford and it proved to be a turning point in my life. I was hoping to receive my Ph.D. in June and was working on a somewhat trivial problem that I found neither interesting nor creative, and the future did not look very promising. This was only shortly before E. R. was to go to London for his Bakerian and Royal Institution Lectures and he was fuH of his subject. His enthusiasm was contagious and his subject appealed strongly to Me. This was just what I needed. I now knew what Iwanted to do. I therefore wrote to him at once asking if I could come to Montreal and work under him. I had been a Teaching Fellow at Yale and I asked for and was given apart-time demonstratorship at McGill. So began three most rewarding years of research under Rutherford's guidance before he left for Manchester in 1907. I think I must have been his first foreign research student. Rutherford proved to be all that I had hoped for, not only stimulating but considerate and helpful. He seemed to have an uncanny ability to see any problem as a whole and the most direct way to attack it. He never tried to use the work of his students and younger associates to feather his own nest, but always made certain that they received more than their fair share of the credit for any work done. I had the pleasure of accompanying E. R. to New Haven when he gave the Silliman Lectures in 1905. Yale was at that time in the process of combining its two Departments of Physics into one, and building a large new Laboratory on its new campus; and Rutherford was urged to head up the whole undertaking, practically on his own terms. He discussed with me the pros and cons on the way back to Montreal, but his reaction was-'Why should I go there? They act as though the University was made for the students.' It was on this trip, I think, that he told me about some of his experiences at his Bakerian and Royal Institution Lectures of the previous year and how il1 at ease he had been. I heard from him the story that Eve teIls on p. 107 of his 'Rutherford', of how he handled Lord Kelvin on that occasion. But in telling it he showed how seriously troubled he had been until the 'sudden inspiration' came, because Kelvin was such a popular idol.

164

The Collected Papers 0/ Lord Rutherford

I might add a personal incident which indicates Rutherford's attitude towards his co-workers. During much of 1905 I was working on the decay constant of radium A, B, and C. I was using a new method which gave more accurate values than had previously been obtained, and E. R. was naturally anxious to have the results published as soon as possible. I objected because there was a discrepancy between the theoretical and experimental curves greater than could be accounted for by experimental errors. He said that I was striving for too great accuracy and that ewe should leave it for the other fellow to get the next decimal place'. I think this attitude was one of the secrets of his successfulleadership in his field. However, he did not press the point or take offence at my disagreeing with him; that is, he had respect for my personality if not for my judgment. It was really wonderful to work under a man who never took that superior attitude which made me hesitate to speak my mind freely. So I had my own way and the satisfaction of clearing up the discrepancy to the conviction of all concemed. 2

by

OTTO HAHN, FOR. MEM. R.S.

As I am one of the oldest surviving students of Emest Rutherford from his time at McGill I have been asked to contribute some personal reminiscences as an introduction to his work during this period. My memories date from the autumn of 1905 to the summer of 1906-a short period but long enough not only to leam something of the scientific powers of the young, but already famous, professor but also to form a picture of the personality of this remarkable man. The call to McGill of the young New Zealand physicist came in the year 1898, scarcely three years after he had left his distant home to work under J. J. Thomson in the Cavendish Laboratory in Cambridge. When Professor Callendar left McGill for the chair at University College, London, Rutherford had already achieved such a reputation in England by his work in Cambridge that he was immediately named as Callendar's successor. The decision came ab out mainly through the recommendation of J. J. Thomson and after an interview with the principal of McGill, Dr Peterson, and the Macdonald Professor of Physics, John Cox. Thus in the autumn of 1898 Rutherford left England for Canada. Although he stayed less than ten years, he carried out there a great series of experiments which made him famous throughout the whole scientific world. This is not the place to tell about the details ofthese researches, for they will be found in the publications reproduced in this volume. It is rather my intention to give here some personal recollections of my stay in Montreal. They will, I hope, help a Httle towards understanding some aspects of Rutherford's personality.

Reminiscences ofProfessor Ernest Ruthetfol'd at McGill Univel'sity

165

From autumn 1904 to summer 1905 I was in London working in Sir William Ramsay's laboratory, and I had, by a fortunate chance, discovered a new radioactive element which I named 'Radiothorium' . Actually, my task was to separate about 9 to 10 mg. of pure radium from a radium-barium mixture, containing very Httle radium, for adetermination of its atomic weight; but as the active barium salt originated from a thorium mineral containing uranium the discovery of the thorium transrnutation product of fairly long life was not difficult. After this 'discovery' Ramsay suggested that I should concentrate completely on radioactivity. Knowing my ignorance ofthis new subject-I was an organic chemist-I wrote to Professor Rutherford requesting that I might work with hirn for a time, mentioning 'in support ofmy request' the discovery of the new radioactive element. Rutherford wrote me a friendly letter in return, and so, with my radiothorium, I set off in the autumn of 1905 for Montreal. On the very day of my arrival in his laboratory Rutherford asked me to tell him about the 'new element'. He appeared to be a little sceptical. I was, however, able to convince hirn that I had found something quite new, not merely the more rapidly decaying thorium X previously discovered by Rutherford and Soddy. Later, Rutherford admitted to me that at first he had not believed in radiothorium. His scepticism had been reinforced by the opinion of his friend B. B. Boltwood of Yale, who had written to hirn: 'The substance appears to be a new component of thorium X and stupidity.' Neither Rutherford nor Boltwood had much faith in the radioactive work then being reported from Ramsay's laboratory. Later Boltwood and I became good friends. (A discussion with him about the period of radiothorium led me to the hypothesis of a further radioactive decay product of thorium, which I actually discovered after my return to Germany. I named this substance 'Mesothorium'.) As it happened, already in Montreal I was able to have my revenge against Rutherford's original misgivings about radiothorium. I discovered in actinium a new radioactive substance which had been overlooked in the experiments in Rutherford's laboratory on the decay products of actinium. I called this substance 'Radioactinium'. When later in Germany-partly alone, partly in collaboration with Lise Meitner-I discovered still a few more radioactive elements, now better called atomic species, Rutherford wrote to me: 'You seem to have a special smell for discovering new radioactive elements.' The atmosphere and spirit in Rutherford's laboratory were extremely happy. The number of his research students was not yet large, and so he was able to give individual attention to each of them, which he did almost every day. These students all contributed to the rapid development of the new subject ofradioactivity. R. K. McClung worked on the ßrays. Howard Bronson made accurate measurements on the decay of the active deposits; he onee said to me that his curves were more accurate than the graph paper on which

166

The Collected Papers of Lord Rutherford

he recorded his measurements, and consequently he gave up using graph paper. A. S. Eve was examining the intensity ofthe y rays from uranium and from thorium minerals. The electroscopes he made in the laboratory had, however, too high a naturalleak, for the laboratory was already contaminated, and he had to make his electroscopes at home. We were not so careful about radioactive contaminations in those times as we are today. When on one occasion Rutherford helped me to get over some trouble with the adjustment of an electrometer, he removed the trouble but made the apparatus radioactive. At that time Rutherford was chiefly occupied with the study of the IX particles, which indeed remained his first love throughout his life. He had bullt in the cellar the apparatus with which he detecmined the electric and magnetic deviation of the IX partic1es. Even at this early stage this work had convinced him that the IX partic1es were either atoms of helium or doubly-charged atoms of hydrogen. In ajoint publication in 1902 Rutherford and Soddy had already suggested that helium was the more likely alternative, on the ground that helium is contained in all radioactive minerals. The experimental proof of the production of helium by radium was therefore envisaged by Rutherford, and, when Soddy left Montreal for England in 1903, Rutherford made a working arrangement with Soddy whereby Soddy was to try to prove the formation of radium from uranium-then a quite unsolved problem, since the longlived intermediate product Ionium had not yet been discovered-while Rutherford would try to prove the production of helium by radium. This was made possible by the German chemist Professor Giesel, who, employed in the quinine factory of Buchler and Co. in Braunschweig, had made the commercial production of radium a kind of hobby, and who put radium preparations on the market at the astonishingly low price of 10 marks, later at 20 marks, per mg. (A few years later radium was selling at 150 marks per mg.) Thus it was that in 1903 both Rutherford, during a visit to London, and Sir William Ramsay were able to buy radium bromide from Giesel at f1 per mg. Soddy, who was at that time working with Ramsay, forgot the arrangement which Rutherford had made with him. U sing apreparation of about 20 mg. of radium bromide Ramsay and Soddy were able to prove spectroscopically the production of helium from radium. To confirm this important discovery, Rutherford then lent Soddy his own radium preparation. At the bottom of his heart Rutherford was a little hurt by this departure of Soddy's from the agreement. In various conversations with him when I was in Montreal, two years after the event, I noticed that he spoke of Soddy with a certain reserve. About Ramsay he was less reticent; after his brilliant work on the rare gases Ramsay did not have the 'lucky touch' in his later work on radioactivity. (Hence Rutherford's early doubts about the discovery ofradiothorium in Ramsay's laboratory.) Rutherford always spoke with the greatest respect of J. J. Thomson, and with high esteem of M. and Mme Curie. He had once had a scientific dispute

Rem;ll;sc('llces l~{ Pr(?{essor Ernest Ruthe/ford at At cGill Cllil'el'sity

167

with Henri Becquerel, so that perhaps he did not look upon him with an entirely unbiased mind. Rutherford was on very friendly terms with B. B. Boltwood, the distinguished radiochemist who discovered Ionium and who, unfortunately, died at a comparatively early age. During my stay in Montreal, A. S. Eve seemed c10sest to hirn of all his colleagues. Eve's charming wife was a sister of Miss Brooks, a former very promising student of Rutherford's. Another good friend was Professor J. C. McLennan of Toronto. Both gratitude and friendship bound Rutherford to the Director of the Macdonald Physics Laboratory, Professor John Cox, who was an excellent teacher but not so distinguished as a research worker, and who, recognizing the extraordinary gifts of his younger colleague, enabled Rutherford to devote hirnself almost exc1usively to research. Of the foreign workers in the laboratory I may mention Godlewski, the young Polish Dozent, already there when I arrived, a very capable man of whom Rutherford thought a great deal; and Dr Max Lewin, who arrived in Montreal from Göttingen at the same time as myself, and who, later, when still a young Professor in Göttingen, had to take over his father's cloth factory and was thus lost to science. Compared with later times, equipment and apparatus were very simple. We built our own ß- and y-ray electroscopes out of large tin cans, on which were placed smaller tobacco or cigarette tins. The insulation of the electroscope leaf was made of sulphur, for we had no amber. The evacuation of the apparatus used by Rutherford in his IX-ray experiments was done by means of a rather old-fashioned Toepler pump, so that the source of active deposit had already largely decayed by the time a reasonably good vacuum had been obtained. On the other hand, the whole subject was still so new that even with these primitive means it was easy to experience the joy of discovery. In Montreal, Rutherford was freely recognized as the leader in scientific research. It rnight happen, at the end of a joint colloquium with the chemists on some subject in organic chemistry, that Rutherford would make some brief comment or other and then, forgetting the subject under discussion, embark in his enthusiastic way on adescription of his latest experiments on his beloved IX partic1es. Everything else was then forgotten. Rutherford's enthusiasm and abounding vigour naturallyaffected us all. To work in the laboratory in the evening was the rule rather than the exception, particularly for us Germans, whose stay in Montreal was limited. Frequently we would spend the evening in his house, when naturally little but 'shop' was talked, not always to the pleasure of the hospitable Mrs Rutherford, who would have preferred to play the piano. He had a great, hearty laugh which echoed through the whole laboratory. I remember a publication of the German chemist W. Marckwald, which ended a long argument with Mrne Curie. Polonium, which was first discovered by Mme Curie, was also found later by Marckwald but called by hirn radiotellurium. Mme Curie had described her polonium as a bismuth-like element,

168

The Collected Papers 0/ Lord Ruther/ord

which is not quite correct; the name radiotellurium was chemically justified. Eventually it turned out that polonium and radiotellurium were identica1, and the latter name had to be dropped, although polonium is in fact a higher homologue of tellurium. Marckwald ended his paper with the familiar quotation from Romeo and luliet: 'What's in a name? That which we call a rose By any other name would smell as sweet.' When he read this Rutherford was delighted with, as he said, such a happy ending. He went round the laboratory reciting the quotation in his robust voice, and, later, whenever the name of Marckwald came up, out came the quotation. This gay and youthful unaffectedness was one of the qualities which made contact with Rutherford so enjoyable. In the neighbouring Chemistry Laboratory was an Assistant Professor by name of Evans. In spite of totally different scientific interests Evans was often invited to the Rutherfords'. There was another reason for this, apart from their friendship. Evans was always induced to recite a long rhyme which began thus: 'I started on a journey just about a week aga For the Httle town of Morrow in the state of Ohio. Said I: "My friend, I want to go to Morrow and return Not later than to-rnorrow for I have no time to burn." Said he to rne: "Now let rne see if I have heard you right. You want to go to Morrow and corne back to-rnorrow night? But if you started yesterday to Morrow, don't you see, You could have got to Morrow and returned to-day at three." , It should be said that Evans delivered these long and somewhat involved verses with striking effect. Rutherford did not care very much about his outward appearance. In his youth he had come to England from New Zealand with the aid of a scholarship, with little or no money of his own. One day in Montreal a representative of Nature presented himself in order to take a photograph of the prominent scientist for this journal. The photograph was taken in the cellar of the laboratory where Rutherford's oc-ray apparatus was set up. When the first negative was developed, the photographer was not satisfied with the result. His appearance was not elegant enough for the fastidious English public; not even a pair of cuffs showed below his sleeves. So for the next picture I had to lend him my pair of detachable cuffs. But in the second photograph not enough of them was visible. However, in the third picture they showed up in their glory, and so in 1906 I had the proud satisfaction of seeing my 'detachables' immortalized in Nature. Rutherford left Montreal in 1907. He continued his brilliant discoveries and did work, even more exciting, in Manchester and Cambridge. But I believe that later, at the height ofhis farne as Lord Rutherford ofNelson, he counted his untroubled years in Montreal among the happiest of his lire.

Uranium Radiation and the ElectricaI Conduction Produced by It by E. R UTHERFORD, M.A., D.se. formerly 1851 Science Scholar, Coutts Trotter Student, Trinity College, Cambridge,' McDonald Professor of Physics, McGill University, Montreal. From the Philosophical Magazine for January 1899, sero 5, xlvii, pp. 109-163 Communicated by Professor J. J. Thomson, F.R.S.

THE remarkable radiation emitted by uranium and its compounds has been studied by its discoverer, Becquerel, and the results ofhis investigations on the nature and properties of the radiation have been given in aseries of papers in the Comptes Rendus. * He showed that the radiation, continuously emitted from uranium compounds, has the power of passing through considerable thicknesses ofmetals and other opaque substances; it has the power ofacting on a photographie plate and of discharging positive and negative electrification to an equal degree. The gas through which the radiation passes is made a temporary conductor of electricity and preserves its power of discharging electrification for a short time after the source of radiation has been removed. The results of Becquerel showed that Röntgen and uranium radiations were very similar in their power of penetrating solid bodies and producing conduction in agas exposed to them; but there was an essential difference between the two types of radiation. He found that uranium radiation could be refracted and polarized, while no definite results showing polarization or refraction have been obtained for Röntgen radiation. It is the object of the present paper to investigate in more detail the nature of uranium radiation and the electrica1 conduction produced. As most of the results obtained have been interpreted on the ionization theory of gases which was introduced to explain the electrical conduction produced by Röntgen radiation, abrief account is given of the theory and the results to which it leads. In the course of the investigation, the following subjects have been considered: § 1. § 2.

Comparison of methods of investigation. Refraction and polarization of uranium radiation.

* Comptes Rendus, 1896, pp. 420, 501, 559, 689, 762, 1086; 1897, pp. 438, 800.

F*

170

The Collected Papers of Lord Rutherford

§ 3. § 4. § 5.

§ 6. § 7. § 8. § 9. § 10. § 11. § 12. § 13. § 14. § 15.

§ 16. § 17. § 18. § 19.

Theory of ionization of gases. Complexity of uranium radiation. Comparison of the radiation from uranium and its compounds. Opacity of substances for the radiation. Thorium radiation. Absorption of radiation by gases. Variation of absorption with pressure. Effect of pressure of the gas on the rate of discharge. The conductivity produced in gases by complete absorption of the radiation. Variation of the rate of discharge with distance between the plates. Rate of recombination of the ions. Velocity of the ions. Fall of potential between two plates. Relation between the current through the gas and e1ectromotive force applied. Production of charged gases by separation of the ions. Discharging power of fine gauzes. General remarks. § 1. Comparison of Methods o/Investigation

The properties of uranium radiation may be investigated by two methods, one depending on the action on a photographic plate and the other on the discharge of electrification. The photographic method is very slow and tedious, and admits of only the roughest measurements. Two or three days' exposure to the radiation is generally required to produce any marked effect on the photographic plate. In addition, when we are dealing with very slight photographic action, the fogging of the plate, during the long exposures required, by the vapours of substances· is liable to obscure the results. On the other hand, the method of testing the electrical discharge caused by the radiation is much more rapid than the photographic method, and also admits of fairly accurate quantitative determinations. The question of polarization and refraction of the radiation can, however, only be tested by the photographic method. The electrical experiment (explained in § 2) to test refraction is not very satisfactory. § 2. Polarization and Refraction

The almost identical effects produced in gases by uranium and Röntgen radiation (which will be described later) led me to consider the question whether the two types of radiation did not behave the same in other respects. In order to test this, experiments were tried to see ifuranium radiation could be polarized or refracted. Becquerelt had found evidence of polarization and ... Russell, Proc. Roy. Soc., 1897.

t Comptes Rendus,

1896, p. 559 •

L"rallilllll Radiation and file Eleclrical ConduCfiol1 Produced hy 11

171

refraction, but in repeating experiments similar to those tried by hirn, I have been unable to find any evidence of either. A large number of photographs by the radiation have been taken under various conditions, but in no case have I been able to observe any effect on the photographie plate which showed the presence of polarization or refraction. In order to avoid fogging of the plate during the long exposures required, by the vapours of substances, lead was employed as far as possible in the neighbourhood of the plate, as its effect on the film is very slight. Abrief account will now be given of the experiments on refraction and polarization. Relraction. A thick lead plate was taken and a long narrow slit cut through it; this was placed over a uniform layer ofuranium oxide; the arrangement was then equivalent to a line source of radiation and a slit. Thin prisms of glass, aluminium, and paraffin-wax were fixed at intervals on the lead plate with their edges just covering the slit. A photographic plate was supported 5 mm. from the slit. The plate was left for a week in a dark box. On developing a dark line was observed on the plate. This line was not appreciably broadened or displaced above the prisms. Different sizes of slits gave equally negative results. If there was any appreciable refraction we should expect the image of the slit to be displaced from the line of the slit. Becquerel* examined the opacity of glass for uranium radiation in the solid and also in a finely-powdered state by the method of electric leakage, and found that, if anything, the transparency of the glass for the radiation was greater in the finely divided than in the solid state. I have repeated this experiment and obtained the same result. As Becquerel stated, it is difficult to reconcile this result with the presence of refraction. Polarization. An arrangement very similar to that used by Becquerel was employed. A deep hole was cut in a thick lead plate and partly filled with uranium oxide. A small tourmaline covered the opening. Another small tourmaline was cut in two and placed on top of the first, so that in one half of the opening the tourmalines were crossed and in the other half uncrossed. The tourmalines were very good optically. The photographic plate was supported 1 to 3 mm. above the tourmalines. The plate was exposed four days, and on developing a black circle showed up on the plate, but in not one of the photographs could the slightest difference in the intensity be observed. Becquerel* stated that in his experiment the two halves were unequally darkened, and concluded from this result that the radiation was doubly refracted by tourmaline, and that the two rays were unequally absorbed. § 3. Theory 01 Ionization

To explain the conductivity of agas exposed to Röntgen radiation, the theoryt has been put forward that the rays in passing through the gas produce

* Comptes Rendus, 1896, p. 559.

t J. J. Thomson and E. Rutherford, Phi!. Mag., November 1896.

172

The Collected Papers 0/ Lord Ruther/ord

positively and negatively charged particles in the gas, and that the number produced per second depends on the intensity of the radiation and the pressure. These carriers are assumed to be so small that they will move with a uniform velocity through agas under a constant potential gradient. The term ion was given to them from analogy with electrolytic conduction, but in using the term it is not assumed that the ion is necessarily of atomic dimensions; it may be a multiple or submultiple of the atom. Suppose we have agas between two plates exposed to the radiation and that the plates are kept at a constant difference of potential. A certain number ofions will be produced per second by the radiation and the number produced will in general depend on the pressure of the gas. Under the electric field the positive ions travel towards the negative plate and the negative ions towards the other plate, and consequently a current will pass through the gas. Some of the ions will also recombine, the rate of recombination being proportional to the square of the number present. The current passing through the gas for a given intensity of radiation will depend on the difference of potential between the plates, but when the potential difference is greater than a certain value the current will reach a maximum. When this is the case all the ions are removed by the electric field before they can recombine. The positive and negative ions will be partially separated by the electric field, 4D.d an excess of ions of one sign may be blown away, so that a charged gas will be obtained. If the ions are not uniformly distributed between the plates, the potential gradient will be disturbed by the movement of the ions. If energy is absorbed in producing ions, we should expect the absorption to be proportional to the number of ions produced and thus depend on the pressure. If this theory be applied to uranium radiation we should expect to obtain the following results: (I) Charged carriers produced through the volume of the gas. (2) Ionization proportional to the intensity of the radiation and the pressure. (3) Absorption of radiation proportional to pressure. (4) Existence of saturation current. (5) Rate of recombination of the ions proportional to the square of the number present. (6) Partial separation of positive and negative ions. (1) Disturbance of potential gradient under certain conditions between two plates exposed to the radiation. The experiments now to be described sufficiently indicate that the theory does form a satisfactory explanation of the electrical conductivity produced by uranium radiation. In all experiments to follow, the results are independent of the sign of the charged plate, unless the contrary is expressly stated.

Uranium Radiation and the Electrical Conduction Produced by It § 4. Comp/ex Nature

0/ Uranium

173

Radiation

Before entering on the general phenomena of the conduction produced by uranium radiation, an account will be given of some experiments to decide whether the same radiation is emitted by uranium and its compounds and whether the radiation is homogeneous. Röntgen and others have observed that the x-rays are in general of a complex nature, inc1uding rays of wide differences in their power of penetrating solid bodies. The penetrating power is also dependent to a large extent on the stage of exhaustion of the Crookes tube. In order to test the complexity of the radiation, an electrical method was employed. The general arrangement is shown in Fig. 1.

8 E~rfJI E~rfJI

A

8

E~rfJI

Fig. 1 The metallic uranium or compound of uranium to be employed was powdered and spread uniforrnly over the centre of a horizontal zinc plate A, 20 cm. square. A zinc plate B, 20 cm. square, was fixed parallel to A and 4 cm. from it. Both plates were insulated. A was connected to one pole of a battery of 50 volts, the other pole of which was to earth; B was connected to one pair of quadrants of an electrometer, the other pair of which was connected to earth. Under the influence of the uranium radiation there was a rate of leak between the two plates A and B. The rate of movement of the electrometer needle, when the motion was steady, was taken as a measure of the current through the gas. Successive layers of thin metal foil were then placed over the uranium compound and the rate of leak determined for each additional sheet. The table (p. 174) shows the results obtained for thin Dutch meta!. In the third column the ratio of the rates of leak for each additional thickness of metal leaf is given. Where two thicknesses were added at once, the square root of the observed ratio is taken, for three thicknesses the cube root. The table shows that for the first ten thicknesses of met al the rate of leak diminished approximately in a geometrical progression as the thickness of the metal increased in arithmetical progression.

174

The Collected Papers 01 Lord Rutherlord THICKNESS OF METAL LEAF 0·00008 CM. LAYER OF URANIUM OXIDE ON PLATE

Numberof Layers

Leak: per min. in scale divisions

0

91

1

77

2

60

3

49

4

42

5

33

6

24·7

8

15·4

10

9·1

13

5·8

Ratio for each layer

0·85 0·78 0·82 0·86 0·79 0·75 0·79 0·77 0·86

It will be shown later (§ 8) that the rate of leak between two plates for a saturating voltage is proportional to the intensity of the radiation after passing through the metal. The voltage of 50 employed was not sufficient to saturate the gas, but it was found that the comparative rates of leak under similar conditions for 50 and 200 volts between the plates were nearly the same. When we are dealing with very small rates of leak, it is advisable to employ as small a voltage as possible, in order that any small changes in the voltage of the battery should not appreciably affeet the result. For this reason the voltage of 50 was used, and the eomparative rates of leak obtained are very approximately the same as for saturating electromotive forces. Since the rate of leak diminishes in a geometrieal progression with the thickness of metal, we see from the above statement that the intensity of the radiation falls off in a geometrie al progression, i.e. according to an ordinary absorption law. This shows that the part of the radiation considered is approximately homogeneous. With increase of the number of layers the absorption commences to diminish. This is shown more clearly by using uranium oxide with layers of thin aluminium leaf (see table, p. 175). It will be observed that for the first three layers of aluminium foil, the

Uranilll11 Radiatioll anti the EleCII'kal COllduction Produced by /1

175

intensity of the radiation falls off according to the ordinary absorption law, and that. after the fourth thickness, the intensity of the radiation is only slightly diminished by adding another eight ]ayers. THICKNESS OF ALUMINIUM FOIL

Number of layers of Aluminium foil

Leak per minute in scale divisions

0

182

1

77

2

33

3

14·6

4

9·4

12

0·0005

CM.

Ratio

0·42 0·43 0·44 0·65

7

The aluminium foil in this case was about 0·0005 cm. thick, so that after the passage of the radiation through 0·002 cm. of aluminium the intensity of the radiation is reduced to about one-twentieth of its value. The addition of a thickness of 0·00 1 cm. of aluminium has only a small effect in cutting down the rate of leak. The intensity is, however, again reduced to about half of its value after passing through an additional thickness of O· 05 cm., which corresponds to one hundred sheets of aluminium foil. These experiments show that the uranium radiation is complex, and that there are present at least two distinct types of radiation-one that is very readily absorbed, which will be termed for convenience the oe radiation, and the other of a more penetrative character, which will be termed the ß radiation. The character ofthe ßradiation seems to be independent ofthe nature ofthe filter through which it has passed. It was found that radiation of the same intensity and of the same penetrative power was obtained by cutting off the oe radiation by thin sheets of aluminium, tinfoil, or paper. The ß radiation passes through all the substances tried with far greater facility than the oe radiation. For example, a plate of thin cover glass placed over the uranium reduced the rate ofleak to one-thirtieth of its value; the ßradiation, however, passed through it with hardly any loss of intensity. Some experiments with different thicknesses of aluminium seem to show, as far as the results go, that the ß radiation is of an approximately homogeneous character. The following table gives some of the results obtained for the ß radiation from uranium oxide:

The Collected Papers 0/ Lord Rutherford

176

ß RADIATION Thickness of Aluminium

0·005 0·028 0·051 0·09

RateofLeak

1

0·68 0·48 0·25

The rate of leak is taken as unity after the CI: radiation has been absorbed by passing through ten layers of aluminium foil. The intensity of the radiation diminishes with the thickness of metal traversed according to the ordinary absorption law. It must be remembered that when we are dealing with the ß radiation alone, the rate of leak is, in general, only a few per cent of the leak due to the CI: radiation, so that the investigation of the homogeneity of the ßradiation cannot be carried out with the same accuracy as for the oe radiation. As far, however, as the experiments have gone, the results seem to point to the conclusion that the ß radiation is approximately homogeneous, although it is possible that other types of radiation of either small intensity or very great penetrating power may be present.

§ 5. Radiation emitted by different Compounds

0/ Uranium

All the compounds of uranium examined gave out the two types of radiation, and the penetrating power of the radiation for both the CI: and ß radiations is the same for all the compounds. The following table shows the results obtained for some of the uranium compounds. TmCKNESS OF ALUMINIuM FOIL

Number of layers of Aluminium foil

Uranium metal

0 1 2 3

1 0·51 0·35

5 12

0·15

4

0·0005

CM.

Proportionate Rate of Leak Uranium Uranium Uranium Potassium Sulphate Nitrate Oxide

1 0·43 0·28 0·17 0·15

1 0·42 0·18 0·08 0·05

1 0·42 0·27 0·17 0·12

0·125

0·04

0·11

Uraniunz Radiation and the Elee/deal Conduetion Produeed by It

177

Fig. 2 shows graphically some of the results obtained for the various uranium compounds. The ordinates represent rates of leak, and the abscissae thicknesses of aluminium through which the radiation has passed. The different compounds of uranium gave different rates of leak, but, for convenience of comparison, the rate of leak due to the uncovered salt is taken as unity. 511

so

10

IN/)IJ,CTI()~ IN/)IJ,CTI()~

28.

Curr

ent C

urve

URAN/UM POTASS/UM SULPHATE URAN/UM OX/DE· I

30 20 10 Thickness of Aluminium: each division 0 -00012 cm.

50

Fig.2 It will be seen that the rate of decrease is approximately the same for the first layer of metal, and that the rate of decrease becomes much slower after four thicknesses of foil. The rate of leak due to the ß radiation is a different proportion of the total amount in each case. The uranium metal was used in the form of powder, and a smaller area of it was used than in the other cases. For the experiments on uranium oxide a thin layer of fine powder was employed, and we see, in that case, that the ß radiation bears a much smaller proportion to the total than for the other compounds. When a thick layer of the

178

The Collected Papers o[ Lord Ruther[ord

oxide was used there was, however, an increase in the ratio, as the following table shows: Number of Iayers of Aluminium foil

o

1 2 4 8 12

18

Rate of Leak Thin Iayer of Thick Iayer of Uranium Oxide Uranium Oxide

1 0·42

0·18 0·05 0·04

1 5

0·12 0·113

0·11

The amount of the IX radiation depends chiefly on the surface of the uranium compound, while the ß radiation depends also on the thickness of the layer. The increase of the rate of leak due to the ß radiation with the thickness of the layer indicates that the ßradiation can pass through a considerable thickness of the uranium compound. Experiments showed that the leak due to the IX radiation did not increase much with the thickness ofthe layer. I did not, however, have enough uranium salt to test the variation of the rate of leak due to the ß radiation for thick layers. The rate of leak from a given weight of uranium or uranium compound depends largely on the amount of surface. The greater the surface, the greater the rate ofleak. A small crystal ofuranium nitrate was dissolved in water, and the water then evaporated so as to deposit a thin layer of the salt over the bottom of the dish. This gave quite a large leakage. The leakage in such a case is due chiefly to the IX radiation. Since the rate of leak due to any uranium compound depends largely on its amount of surface, it is difficult to compare the quantity of radiation given out by equal amounts of different salts: for the result will depend greatly on the state of division of the compound. It is possible that the apparently very powerful radiation obtained from pitchblende by Curie* may be partly due to the very fine state of division of the substance rather than to the presence of a new and powerful radiating substance. The rate ofleak due to the ßradiation is, as a rule, small compared with that produced by the IX radiation. It is difficult, however, to compare the relative intensities of the two kinds. The IX radiation is strongly absorbed by gases (§ 8), while the ß radiation is only slightly so. It will be shown later (§ 8) that the absorption of the radiation by the gas is approximately proportional to the number of ions produced. If, therefore, the ß radiation is only slightly '" Comptes Rendus. July 1898, p. 175.

Uranium Radiation and the Electrical Conductiol1 Produced by It

179

absorbed by the gas, the number of ions produced by it is small, i.e. the rate of leak is small. The comparative rates of leak due to the 1X and ßradiations is thus dependent on the relative absorption of the radiations by the gas as well as on the relative intensity. The photographic actions of the 1X and ß radiations have also been co mpared. A thin uniform layer ofuranium oxide was sprinkled over a glass plate; one half of the plate was covered hy a piece of aluminium of sufficient thickness to practically ahsorh the 1X radiation. The photo graphie plate was fixed about 4 mm. from the uranium surfaee. The plate was exposed 48 hours, and, on developing, it was found that the darkening of the two halves was not greatly different. On the one half ofthe plate the action was due to the ßradiation alone, and on the other due to the 1X and ß radiations together. Exeept when the photographie plate is elose to the uranium surfaee, the photographie action is due prineipally to the ß radiation.

§ 6. Transparency

0/ Substances

to the two Types

0/ Radiation

If the intensity of the radiation in traversing a substanee diminishes aceording to the ordinary absorption law, the ratio r of the intensity of the radiation after passing through a distance d of the suhstance to the intensity when the suhstance is removed is given by

r = e-Ad, where ,\ is the coeffieient of absorption and e = 2·7. In the following table a few values of ,\ are given for the 1X and ß radiations, assuming in each ease that the radiation is simple and thatthe intensity falls off aecording to the above law: Substance

Dutch metal Aluminium Tinfoil Copper Silver Platinum Glass

A

A

for the cx radiation for the ~ radiation

2700

1600 2650

15

108 49 97

240 5·6

The ahove results show wh at a great difference there is in the power of penetration of the two types of radiation. The transparency of aluminium for the ß radiation is over one hundred times as great as for the 1X radiation. The opacity of the metals aluminium, copper, silver, platinum for the ß radiation follows the same order as their atomic weights. Aluminium is the most transparent of the metals used, hut glass is more transparent than aluminium for the ß radiation. Platinum has an opacity sixteen times as great as

180

The Collected Papers of Lord Rutherford

aluminium. For the a. radiation, aluminium is more transparent than Dutch metal or tinfoil. For a thickness of aluminium 0·09 cm. the intensity of the ß radiation was reduced to 0·25 of its value; for a thickness of copper 0·03 cm. the intensity was reduced to 0·23 of its value. These results are not in agreement with some given by Becquerel, * who found copper was more transparent than aluminium for uranium radiation. The ß radiation has a penetrating power of about the same order as the radiation given out by an average x-ray bulb. Its power of penetration is, however, much less than for the rays from a 'hard' bulb. The a. radiation, on the other hand, is far more easily absorbed than rays from an ordinary bulb, but is very sirnilar in its penetrating power to the secondary radiationt sent out when x-rays fall upon a metal surface. It is possible that the a. radiation Is a secondary radiation set up at the surface of the uranium by the passage of the ßradiation through the uranium, in exactly the same way as a diffuse radiation is produced at the surface of a metal by the passage of Röntgen rays through it. There is not, however. sufficient evidence at present to decide the question.

§ 7. Thorium Radiation While the experiments on the complex nature of uranium radiation were in progress, the discoveryt that thorium and its salts also emitted a radiation, which had general properties similar to uranium radiation, was announced. A few experiments were made to compare the types of radiation emitted by uranium and thorium. The nitrate and the sulphate of thorium were used and gave similar results, although the nitrate appeared to be the more active of the two. The leakage effects due to these salts were of quite the same order as those obtained for the uranium compounds; but no satisfactory quantitative comparison can be made betwen the uranium and thorium salts as the amount of leak depends on the amount of surface and thickness of the layer. It was found that thorium nitrate when first exposed to the air on a platinum plate was not a steady source ofradiation, and for a time the rate ofleak varied very capriciously, being sometimes five times as great as at others. The salt was very deliquescent, but after exposure of some hours to the atmosphere the rate of leak became more constant and allowed of rough comparative measurements. Thorium sulphate was more constant than the nitrate. The absorption of the thorium radiation was tested in the same way as for uranium radiation. The following table gives some of the results. The • Comptes Rendus. 1896, p. 763. t Perrin, Comptes Rendus, cxxiv, p. 455; Sagnac, Comptes Rendus, 1898. t G. C. Schmidt, Wied. Annal., May 1898.

Ul'aniul11 Radiation and the Elec/rical Conduction Produced by It

181

aluminium foi! was of the same thickness (0,0005 cm.) as that used in the uranium experiments: Number oflayers Leak per minute in of Aluminium foil scale divisions

o 4

8 12 17

200

94

37 19 7·5

The curve showing the relation between the rate of leak and the thickness of the metal traversed is shown in Fig. 2 (p. 177), together with the results for uranium. It will be seen that thorium radiation is different in penetrative power from the (X radiation of uranium. The radiation will pass through between three and four thicknesses of aluminium foil before the intensity is reduced to one-half, while with uranium radiation the intensity is reduced to less than a half afcer passing through one thickness of foil. With a thick layer of thorium nitrate it was found that the radiation was not bomogeneous, but rays of a more penetrative kind were present. On account of the inconstancy of thorium nitrate as a source of radiation, no accurate experiments have been made on this point. The radiations from thorium and uranium are thus botb complex, and as regards the (X type of radiation are different in penetrating power from each other. In all the experiments on uranium and thorium, care was taken that no stray radiation was present which would obscure the results. Such precautions are very necessary when the rate of leak, due to the radiation transmitted through a considerable thickness of metal, is only a small percentage of the total. The method generally employed was to cover the layer of active salt with the meta! screen, and then place in position over it a large sheet of lead with a rectangular hole cut in it of smaller area than that of the layer of salto The lead was pressed tightly down, and the only radiation between the parallel plates had to pass through the metal screen, as the lead was too thick to allow any to go through. § 8. Absorption

0/

Uranium Radiation by Gases

The (X radiation from uranium and its compounds is rapidly absorbed in its passage through gases. The absorption for hydrogen, air, and carbonic acid was determined, and was found to be least in hydrogen and greatest in carbonic acid. To show the presence of absorption, the following arrangement (Fig. 3) was used: A layer of uranium-potassium sulphate or uranium oxide was spread uniformly over a metal plate P, forming a lamella of 11 cm. diameter. A glass

182

The Collected Papers

0/ Lord Rutherford

vessel G, 12 cm. in diameter, was placed over the layer. Two parallel metal plates A and B, 1·5 cm. apart, were insulated from each other by ebonite rods. A circular opening 7 cm. in diameter was cut in the plate A, and the opening covered by a sheet of aluminium foil 0·0005 cm. thick. The plate B was connected through a rod R to a screw adjustment S, so that the condenser AB could be moved as a whole parallel to the base plate. The system AB was adjusted parallel to the uranium surface and did not rotate with the screw. The rod R passed through a short glass tube fixed in the ebonite plate C. A short piece of indiarubber tubing T was passed over the glass tube and a projecting flange in which the rod R was screwed. This served the same

eil

c EAIiTH

eil

R

eil

elf"""

eil

P-N!,~~'!.~

p CARTII

Fig. 3 purpose as the usual stuffing box, and allowed the distance of AB from the uranium to be adjusted under low pressures. The plate A was connected to one pole of a battery of 60 volts, the other pole of which was to earth. The plate B was connected through the screw to one pair of quadrants of an electrometer, the other pair ofwhich was to earth. In order to avoid the collection of an electrostatic charge on the glass surface due to the conduction between the uranium and the glass near it, it was found very necessary to coat the inside of the glass cylinder with tinfoil. The tinfoil and base plate P were connected to earth. Since the surface of the uranium layer may be supposed to be giving out radiation uniformly from all parts, the intensity ofthe radiation at points near the centre of the uranium surface should be approximately uniform. If there were no absorption of the radiation in the gas, we should expect the intensity

Lranillll1 Radiation and the Electrical Conduction Produced by 11

HG

of the radiation to vary but slightly with distances from the surface small compared with the diameter of the radiating surface. The radiation passing through the aluminium produces conductivity between A and B (Fig. 3), and the rate of leak depends on the intensity of the radiation which has passed through a certain thickness of gas and the aluminium foil. As the system AB is moved from the base plate, if there is a rapid absorption of the radiation in the gas, we should expect the rate of leak to fall offrapidly, and this is found to be the case. The following table gives the results obtained for air, hydrogen, carbonic acid, and coal-gas. For the first reading the distance d of the aluminium foil from the base plate was about 3·5 mm.

Distance of Al. foil from Uranium

d d + 1·25 mm. d+2'5 mm. d + 3·75mm. mm. d+5 d + 7·5 mm. d + 10 mm. d + 12·5 mm. d + 15 mm.

Hydrogen

1

Rate of leak between plates Carbonic Air Acid Coal-gas

1 0·74 0·57 0·41 0·32

1 0·67

0·84 0·67

0·45 0·31 0·21 0·16

1 0·81 0·63 0·39

0·53

0·22

The rate of leak for the distance d is taken as unity in each gas for the purpose of comparison. The actual rates of leak between A and B for the distance dis given in the following table : Gas

Hydrogen Coal-gas Air Carbonic acid

Rate of leak in scale divisions per minute

25 35 28 18

The results of the previous table are shown graphically in Fig. 4, where the ordinates represent currents and the abscissae distances from the base plate. It will be seen that the current decreases most rapidly in carbonic acid and least in hydrogen. As the distance from the base-plate increases in arithmetical progression, the rate of leak diminishes approximately in geometrical progression. The rapid decrease of the current is due to the absorption of the radiation in its passage through the gas. The decrease of the current in air

The Collected Papers 0/ Lord Ruther/ord

184

at 190 mm. pressure is also shown in the figure. Since the absorption is smaller for air at this pressure than at normal pressure, the rate of leak diminishes much more slowly with the distance. In the above experiments both the IX and ßradiations produce conductivity in the gas. A thin layer of uranium oxide was, however, used, and in that case the rate of leak due to the ß radiation may be neglected in comparison with that produced by the IX radiation. The results that have been obtained on the variation ofthe rate ofleak with distance may be simply interpreted on the theory of the ionization of the gas

h

::t ~

:s

~

URAN/UM OX/DE· OX/DE·

10

/5

/5

Fig.4 througb which the radiation passes. It is assumed that tbe rate of ionization is proportional to the intensity of the radiation (as is the case in Röntgen-ray conduction), and that the intensity of the radiation near the uranium surface is constant over a plane parallel to that surface. This is very approximately the case if the distance from the uranium surface is small compared with the diameter of the radiating surface. For simplicity we will consider the case of an infinite plane of uranium giving out homogeneous radiation. If I be the intensity of the radiation close to the uranium surface, the intensity at a distance x is equal to Ie-i.x where A is the coefficient of absorption of the gas. The intensity is diminished in passing through the layer of aluminium foll A (Fig. 3) in a constant ratio of all distances from the uranium. The intensity at a distance x after passing throughl the aluminium is thus Kle-i.x

Cranium Radiation and fhe Electrical Conduction Produced by 11

185

where K is a constant. The rate of production of the ions between two parallel planes between A and B (Fig. 3) at distances x + dx and x from the uranium is therefore proportional to Kle-i..xdx. If r be the distance of A from the uraniurn, and I the distance between A and B, the total number of ions produced per second between A and B is proportional to

J

Hr

~Ke-),xdx,

or to

KI

Ae-i..r {I - e-i../}.

When a 'saturating' electromotive force (see § 16) acts between A and B, the current is proportional to the total nurnber of ions produced. Now, as the system AB is moved from the radiating surface,

~ (1

- e-i..I) is a constant for

any particular gas. We thus see that the rate of leak is proportional to e-N , or the rate of leak decreases in geometrical progression as the distance r increases in arithmetical progression. This result allows us to at once deduce the value of the coefficient of absorption for different gases from the data we have previously given. The results are given in the following table: Gas

Value of A

Hydrogen Air Carbonic acid Coal-gas

0·43 1·6 2·3 0·93

or, to express the same results in a different way, the intensity of the radiation from an infinite plane of uranium is reduced by absorption to half its value after having passed through 3 mrn. of carbonic acid 4· 3 rnrn. of air 7·5 mrn. of co al-gas 16·3 mrn. of hydrogen We see that the absorption is least in hydrogen and greatest in carbonic acid, and follows the same order as the density of the gases. The values given above are for the oe radiation. The ßradiation is not nearly so rapidly absorbed as the oe, but, on account of the small electricalleakage produced in its passage through the gas, it was not found feasible to measure absorption in air or other gases. The absoprtion of the oe radiation by gases is very much greater than the absorption of rays from an ordinary Crookes' tube. In a previous paper* it has been shown that the value of ,\ for the radiation from the particu1ar bulb • Phil. Mag., April 1897.

The Collected Papers 0/ Lord Ruther/ord

186

used was o· 01. The absorption coefficient for the oe: radiation is 1· 6, or 160 times as great. The absorption of the ß radiation in gases is probably of the same order as the absorption for ordinary x-rays. § 9. Variation

0/ Absorption

with Pressure

The absorption of the oe: radiation increases with increase of pressure and very approximately varies directly as the pressure. The same apparatus was used as in Fig. 3, and the vessel was kept connected to an air pump. The variation of the rate of leak between A and B for different distances from the base plate was determined for pressures of 760, 370, and 190 mm., and the results are given below: Rate of leak between plates Air 760 mm. Air 370mm. Air 190mm.

Distance of A from Uranium

d (=3·5 mm.) d 2·5rnm. d+ 5 mm. d 7·5mm. d+ 10 rnrn. d+ 12·5rnrn. d+ 15 rnrn.

1 67 0·45 0·31 0·21 0·16

+ +

1

1

0·71 0'78

0·51 0·36

0·59

For the purpose of comparison the rate of leak at the distance dis taken as unity in each case. It can readily be deduced frorn the results that the intensity of the radiation is reduced to half its value after passing through 4·3 rnrn. ofair at 760 rnrn. 10 rnrn. of air at 370 mrn. 19·5 rnrn. ofair at 190 mm. The absorption is thus approximately proportional to the pressure for the range that has been tried. It was not found feasible to rneasure the absorption 2t lower pressures on account of the large distances through which the radiation rnust pass to be appreciably absorbed. A second rnethod of measuring the absorption of the radiation in gases, which depends on the variation of the rate of leak between two plates as the distance between thern is varied, is given in § 12.

§ 10. Effect

0/ Pressure

on the Rate

0/ Discharge

Becquerel* has given a few results for the efIects of pressure, and showed that the rate of leak due to uranium dirninished with the pressure. Beattie and S. de Srnolant also investigated the subject, and came to the conclusion that in ... Comptes Rendus, p. 438 (1897).

t Phi!. Mag., xliii, p. 418 (1897).

Uran;ul1I Radiation and the Electrical Conduction Produced by It

187

some cases the rate of leak varied as the pressure, and in others as the square root of the pressure, according to the voltage employed. Their tabulated results, however, do not show any elose agreement with either law, and in fact, as I hope to show later, the relation between the rate of leak and the pressure is a very variable one, depending to a large extent on the distances between the uranium and the surrounding conductors, as weIl as on the gas employed. The subject is greatly complicated by the rapid absorption of the

"

~

.;;;.

~

\)

ISO

PRES fURR WMA S. 300 -4S0

60D

750

Fig.5 radiation by gases, but all the results obtained may be interpreted on the assumption that the rate of production of ions at any point varies directly as the intensity of the radiation and the pressure of the gas. To determine the effects of pressure, an apparatus' similar to Fig. 3 was used, with the difference that the plate A was removed. The uranium compound was spread uniformly, over the central part of the lower plate. The movable plate, which was connected with the electrometer, was 10 em. in diameter and moved parallel to the uranium surface. The base plate was connected to one pole of a battery of 100 volts, the other pole of which was connected to earth. The rate of movement of the electrometer needle was taken as a measure of the current between the plates. In some cases

188

The Collected Papers of Lord Rutherford

the uranium compound was covered with a thin layer of aluminium foil, but although this diminished the rate of leak the general relations obtained were unaltered. The following tables give the results obtained for air, hydrogen, and carbonic acid at different pressures with a potential difference of 100 volts between the plates-an amount suflicient to approximately 'saturate' the gases air and hydrogen. Much larger voltages are required to produce approximate saturation for carbonic acid. Air: Uranium oxide on base plate. Plates about 3·5 mm. apart. AIR

Pressure

Current

IDm.

760 600 480 365 210 150 100 50 35

1 0·86 0·74 0·56 0·32 0·23 0·17 0·088 0·062

For hydrogen and carbonie acid. Plates about 3·5 mm. apart. HYDROGEN

Pressure

Current

IDm.

760 540 335 220 135

CARBONIC ACID

Pressure

Current

IDm.

1 0·73 0·46 0·29 0·18

760 410 220 125 55

1 0·92 0·69 0·38 0·175

The current at atmospheric pressure is in each case taken as unity for comparison, although the actual rates ofleak were different for the three gases. Fig. 5 (p. 187) shows these results graphically, where the ordinates represent current and the abscissae pressure. The dotted line shows the position of the curve if the rate of leak varied directly as the pressure. It will be observed that for all three gases the rate of leak first of all increases directly as the pressure, and then increases more slowly as the pressure increases. The difference is least marked in hydrogen and most marked in carbonic acid. In hydrogen the rate of leak is nearly proportional to the pressure. The relation between the rate of leak and the pressure depends also on the

Uranium Radiation and the Electrical Conductioll Produced hy It

189

distance between the plates. The following few numbers are typical of the results obtained. There was a potential difference of 200 volts between the plates and the rate of leak is given in scale divisions per mm.

Pressure

Rate of Leak Distance between Distance between plates 2'5 mm. plates 15 mm.

mm.

187

11

47

41

127

21

376

752

83

For small distances between the plates the rate ofleak is more nearlyproportional to the pressure than for large distances. The differences between the results for various gases and for different distances receive a simple explanation if we consider that the intensity of the radiation falls off rapidly between the plates on account of the absorption in the gas. The tables given for the relation between current and pressure, where the distance between the plates is small, show that when the absorption is small, the rate ofleak varies directly as the pressure. For small absorption the intensity of the radiation is approximately uniform between the plates, and therefore the ionization ofthe gas is uniform throughout the volume ofthe gas between the plates. Since under a saturating electromotive force the rate of leak is proportional to the total ionization, the above experiments show that the rate of production of the ions at any point is proportional to the pressure. It has been previously shown that the absorption of the radiation is approximately proportional to the pressure. Let q = rate of production of the ions near the uranium surface for unit pressure. = coefficient of absorption of the gas for unit pressure.

"0

The total number of ions produced between the plates, distant d apart, per unit area of the plate is, therefore, easily seen to be equal to pq

or to

~

J;-p>'x dx,

(1 - e-p~d

),

since we have shown that the ionization and absorption are proportional to the pressure. If there is a saturating electromotive force acting on the gas, the ratio of the rate of leak at the pressure PI to that at the pressure P2 is equal to

190

The Collected Papers 0/ Lord Rutherford

the ratio r of the total number of ions produced at the pressure PI to the total number at pressure P2 and is given by 1 - e-P1A•d 1 - e-PIAod·

r= -:-----

Now PIAO is the coefficient of absorption of the gas for the pressure PI. Ir the absorption is small between the plates, PIAod and P2Aod are both small and the value of r reduces to r=PI

P2'

or the rate of leak when the pressure is small is proportional to the pressure. Ir the absorption is large between the plates at both the pressures PI and P2, the value of r is nearly unity, i.e. the rate of leak is approximately independent of the pressure. Experimental results on this point are shown graphically in Fig. 7 (p. 194). For intermediate values of the absorption, the value of r changes more slowly than the pressure. With the same distance between the plates, the difference between the curves (Fig. 5) for air and hydrogen is due to the greater absorption of the radiation by the air. The less the absorption ofthe gas, the more nearly is the rate of leak proportional to the pressure. For carbonic acid the rate of leak decreases far more slowly with the pressure than for hydrogen; this is due partly to the much greater value of the absorption in carbonic acid and partly to the fact that 100 volts between the plates was not sufficient to saturate the

gas.

If we take the rate of leak between two parallel plates some distance from the source of radiation, we obtain the somewhat surprising result that the rate of leak increases at first with diminution of pressure, although a saturating electromotive force is applied. The arrangement used was very similar to that in Fig. 3. The rate of leak was taken between the plates A and B, which were 2 cm. apart, and the plate A was about 1·5 cm. from the uranium surface. The following table gives the results obtained: Pressure mm.

Current

760 645 525 380 295 180 100 49

1 1·46 2 2·2 2·05 1·6 1·04 0·58

Liranium Radiation anti the Electrical Conductioll Produced by It

191

The current at atmospheric pressure is taken as unity. The results are represented graphically in Fig. 6. The rate of leak reaches a maximum at apressure of less than half an atmosphere, and then decreases, and at apressure of 100 mm. the rate ofleak is still greater than at atmospheric pressure.

t....

~

t....

~

I'REj SURE INMM. • liOO

400

200

--

80G

Fig. 6 This result is readily explained by the great absorption of the radiation at atmospheric pressure and the diminution of absorption with pressure. Let d1 = distance of plate A from the uranium. d2 = distance of plate B from the uranium. With the notation previously used, the total ionization between A and B (on the assumption that the radiating surface is infinite in extent) is readily seen to be equal to

~ {e-p"J..d.

_ e-p"Aod.}.

This is a function of the pressure, and is a maximum when dle-P"J.od. - d2e-p"Ae d• = 0, Le. when

dl

log e d2

= -

p'Ao(d2

-

d l ).

The value of pA O for air at 760 mm. is 1·6. If d2 = 3 cm., d l = 1, the leak is a maximum when the pressure is about one-third of an atmosphere. On account of the large distance of the plates

192

The Collected Papers of Lord Ruthelford

from the uranium surface in the experimental arrangements, no comparison between experiment and theory could be made. In all the investigations on the relation between the pressure and the rate of leak, large electromotive forces have been used to ensure that the current through the gas is proportional to the total ionization of the gas. With low voltages the relation between current and pressure would be very different, and would vary greatly with the voltage and distance between the electrodes as well as with the gas. It has not been considered necessary to introduce the results obtained for small voltages in this paper, as they are very variable under varying conditions. Although they may all be simply explained on the results obtained for the saturating electromotive forces they do not admit of simple calculation, and only serve to obscure the simple laws which govem the relations between ionization, absorption, and pressure. The general nature of the results for low voltages can be deduced from a consideration of the results given for the connection (see § 16) between the current through the gas and the electromotive force acting on it at various pressures. The above results for the relation between current and pressure may be compared with those obtained for Röntgen radiation. Perrin* found that the rate of leak varied directly as the pressure for saturating electromotive forces when the radiation did not impinge on the surface of the metal plates. This is in agreement with the results obtained for uranium radiation, for Perrin's result practically asserts that the ionization is proportional to the pressure. The results, however, of other experimenters on the subject are very variable and contradictory, due chiefly to the fact that in some cases the results were obtained for non-saturating electromotive forces, while, in addition, the surface ionization at the electrodes greatly complicated the relation, especially at low pressures. § 11. Amount of Ionization in Different Gases

It has been shown that the a; radiation from uranium is rapidly absorbed by air and other gases. In consequence of this the total amount of ionization produced, when the radiation is completely absorbed, can be determined. The following arrangement was used: A brass ball 2·2 cm. in diameter was covered with a thin layer of uranium oxide. A thin brass rod was screwed into it and the sphere was fixed centrally inside a bell jar of 13 cm. diameter, the brass rod passing through an ebonite stopper. The bell jar was fixed to a base plate, and was made airtight. The inside and outside of the bell jar were covered with tinfoil. In practice an E.M.F. of 800 volts was applied to the outside of the bell jar. The sphere, through the metal rod, was connected to one pair of quadrants of an electrometer. It was assumed that, with such a large potential difference between the bell jar and the sphere, the gas was approximately saturated and the rate of movement of the electrometer needle was proportional to the total number of ions produced in the gas. The • Comptes Rendus. cxxiii, p. 878.

PROFESSOR RUTHERFORD AT McGILL UNIVERSITY Pastel by R. G. Matthews, 1907

This page intentionally left blank

CralliulIl Radiation alUl fhe Electrical Conduction Produced by Ir

193

foUowing wen: somc of the results obtained, the rate of leak due to air being taken as 100. Gas

Adr

Hydrogen Oxygen Carbonic Acid Coal-gas Hydrochloric Acid Gas Ammonia Gas

Total Ionization

100 95 106 96 111 102 101

The results for hydrochloric acid and ammonia are only approximate, for it was found that both gases slightly altered the radiation emitted by the uranium oxide. For example, before the introduction of the gas the rate of leak due to air was found to be 100 divisions in 69 sec.; after the introduction of hydrochloric acid 100 divisions in 72 sec.; and with air again after the gas was removed 100 divisions in 74 sec. The rate of leak is greatest in coal-gas and least in hydrogen, but all the gases tried show roughly the same amount of ionization as air. In the case considered both kinds of radiation emitted by uranium are producing ionization in the gas. By covering over the uranium oxide with a few layers of thin tinfoil it was found that, for the arrangement used, the rate of leak due to the penetrating ray was smaU in eomparison with the rate of leak due to the cx radiation. The effeet of diminution ofthe pressure on the rate ofleak for air, hydrogen, and earbonic acid is shown in Fig. 7, where the abscissae represent pressure and the ordinates rate of leak. In the case of air and carbonic acid it was found that the rate of leak slightly increased at first with diminution of pressure. This was ascribed to the fact that even with 800 volts aeting between the uranium and the surrounding conductor the saturation for atmospheric pressure was not complete. It will be observed that the rate ofleak in air remains practically constant down to apressure of 400 mm., and for carbonic acid down to a pressure of 200 mm. In hydrogen, however, the change of rate of leak with pressure is more rapid, and shows that aU the radiation emitted by the uranium was not completely absorbed at atmospheric press ure, so that the total ionization is probably larger than the value given in the table. Assuming that there is the same energy of radiation emitted whatever the gas surrounding the uranium and that the radiation is almost completely absorbed in the gas, we see that there is approximately the same amount of ionization in aU the gases for the same absorption of energy. This is a very interesting result, as it affords us some information on the subject of the relative amounts of energy required to produce ionization in different gases. In whatever process ionization may consist there is energy absorbed, and the energy required to produce aseparation of the same quantity of electricity G

194

The Collected Papers ofLord RutlzerJord

(which is carried by the ions of the gas) is approximately the same in all the gases tried. From the results_ we have just given, it will be seen how indefinite it is to speak of the conductivity of agas produced by uranium radiation. The ratio of the conductivities for different gases will depend very largely on the distance apart of the electrodes between which the rate of leak is observed. When the distance between the electrodes (e.g. two parallel plates) is small, the rate of leak is greater in carbonic acid than in air, and greater in air than in hydrogen. As the distance between the plates is increased, these values tend to approximate equality. If, however, the rate of leak is taken between two plates some

~ ~

o

PRES$JRE IN MMSI

200

480

&00

800

Fig.7 distance from the radiating surface (e.g. the plates A and Bin Fig. 3), the ratio of the rates of leak for different gases will depeud on the distance of the plate A from the surface of the uranium. If the plail. A is several centimetres distant from the uranium, the rate of leak will be greater with hydrogen than with air, and greater in air than in carbonic acid-the exact reverse of the other case. These considerations will also apply to what is meant by the conductivity of agas for uranium radiation. In a previous paper* I found the coefficient of absorption of agas for Röntgen rays to be roughly proportional to the conductivity of the gas. The conductivity in this case was measured by the rate of leak between two plates elose together and not far from the Crookes tube. The absorption in the air

* Phi!.

Mag., April 1897.

Uranium Radiation and the Electrical Conduction Produced by It

195

between the bulb and the testing apparatus was small. If it were possible to completely absorb the Röntgen radiation in agas and measure the resulting conductivity, the total current should be independent of the gas in which the radiation was absorbed. This result follows at onee if the absorption is proportional to the ionization produced for all gases. The results for uranium and Röntgen radiation are thus very similar in this respect. § 12. Variation

0/

the CUl'rent betweell two Plates with the Distance between them

The experimental arrangement adopted was similar to that in Fig. 3 with the plate A removed. Two horizontal polished zinc plates 10 cm. in diameter were placed inside a bell jar. The lower plate was fixed and covered with a uniform layer of uranium oxide, and the upper plate was movable, by means of a screw, parallel to the lower plate. The bell jar was airtight, and was connected with an air pump. The lower plate was connected to one pole of a battery of 200 volts, the other pole of which was earthed, and the insulated top plate was connected with the electrometer. The exterior surface of the glass was covered with tinfoil connected to earth. The following table gives the results of the variation of the rate of leak with distance for air at pressures of 752, 376, and 187 mm. The results have been corrected for change of the eapacity of the electrometer circuit with movement of the plates. Distance between plates mm.

2·5

Rate of leak in scale divisions per minute 752 mm. 376 mm. 187 mm.

5

41 70

12·5 15

109 123 128

7·5 10

92

21

40 53 65 76 83

20

36

47

The results are shown graphically in Fig. 8, where the abscissae represent distances between the plates and the ordinates rates of leak. The values given above correspond to saturation rates ofleak; for 200 volts between the plates is sufficient to very approximately satllrate the gas even for the greatest distance apart of 1. 5 cm. It will be observed that the rate of leak increases nearly proportional1y to the distance between the plates for short distances, but for air at atmospheric pressure increases very slowly with the distance when the distances are large. If there were no appreciable absorption of the radiation by the gas, the

196

The Collected Papers of Lord Rutherford

ionization would be approximately uniform between the plates, provided the diameter ofthe uranium surface was large compared with the greatest distance between the plates. The saturation rate of leak would, in that case, vary as the distance. If there is a large absorption of the radiation by the gas, the ionization will be greatest near the uranium and will fall offrapidly with the distance. The saturation rate ofleak will thus increase at first with the distance, and then tend to a constant value when the radiation is completely absorbed between the plates. 200 VpLTS '(IETW(EN P,lJATIS

~ ~

IJ}fTAN4E Bl7WEE~ PLATts IN "'MS. Q

5

10

JS

Fig. 8

The results given in the previous table allow us to determine the absorption coefficient of air at various pressures. My attention was first drawn to the rapid absorption of the radiation by experiments of this kind. The number of ions produced between two parallel plates distant d apart is equal to

f:.-.>.xdx,

pq

i.e. to

t

(1 - e-PAod),

assuming the ionization and the absorption are proportional to the pressure. The notation is the same as that used in § 10.

Uranium Radiation and the Electrical Conduction Produced by It

197

For the pressure p the saturation rate of leak between the plates is thus proportional to 1 - e-p"A.,d. If p and d are varied so that p X dis a constant, the rate of leak should be a constant. This is approximately true as the numbers previously given (see Fig. 8) show. It must, however, be borne in mind that the conditions, on which the calculations are based, are only approximately fulfilled in practice, for we have assumed the uranium surface to be infinite in extent and that the saturation is complete. The variation of the rate of leak with distance agrees fairly c10sely with the theory. When pAod is small the rate of leak is nearly proportional to the distance between the plates and the pressure of the gas. When pAod is large the rate of leak varies very slowly with the distance. The value of PA o can be deduced from the experimental results, so that we have here an independent method of determining the absorption of the radiation at different pressures. The lower the pressure the more uniform is the ionization between the plates, so that the saturation rate of leak at low pressures is nearly proportional to the distance between the plates. This is seen to be the case in Fig. 8, where the curve for apressure of 187 mm. is approximately a straight line. Similar results have been obtained for hydrogen and carbonic acid. § 13. Rate

0/ Recombination 01 the Ions

Air that has been blown by the surface of a uranium compound has the power of discharging both positive and negative electrification. The following arrangement was used to find the duration of the after-conductivity induced by uranium radiation: A sheet of thick paper was covered over with a thin layer of gum arabic, and then uranium oxide or uranium potassium sulphate in the form of fine powder was sprinkled over it. After this had dried the sheet of paper was formed into a cylinder with the uranium layer inside. This was then placed in a metal tube T (Fig. 9) of 4 cm. diameter. Ablast of air from a

c

,. URAN/UM

A

U

N RA

/U

M

L

B

U

N RA

/U

M

URAN/UM

Fig.9 gasometer, after passing through a plug C of cotton-wool to remove dust, passed through the cylinder T and then down a long metal tube connected to earth. Insulated electrodes A and B were fixed in the metal tube. The electrometer, could be connected to either of the electrodes A or B. In practice the quadrants

198

The Collected Papers of Lord Rutherford

of the electrometer were first connected together. The electrode A or Band the electrometer were then charged up to a potential of 30 volts, and the quadrants then separated. When the uranium was removed there was no rate of leak at either A or B when a rapid current of air was sent through the tube. On replacing the uranium cylinder and sending a current of air along the tube, the electrometer showed a gradual loss of charge which continued until the electrode was discharged. When the electrode A was charged to 30 volts there was no rate ofleak of B. The rate of leak of B or A is thus proportional to total number of ions in the gas. The ions recombine in the interval taken for the air to pass between A and B. The rate of leak of B for a saturating voltage, when A is to earth, is thus less than that of A. For a particular experiment the rate of leak of the electrode A was 146 divisions per minute. When A was connected to earth, the saturation rate of leak of B was 100 divisions per minute. The distance between A and B was 44 cm., and the mean velocity of the current of air along the tube 70 cm. per second. In the time, therefore, of 0·63 sec. the conductivity of the gas has fallen to 0·68 of its value. Ifwe assume, as in the case ofRöntgenized air, * that the loss of conductivity is due to the recombination of the ions, the variation of the number with the time is given by dn

dt

=

-rx.n2 ,

where n is the number of ions per cubic centimetre and rx. a constant. If N is the number of ions at the electrode A, the number of ions n at B after an interval t is given by 1

1

Ii- N=rx.t. Now the saturating rates of leak at A and B are proportional to N and n, and it can readily be deduced that the time taken for the number of ions to recombine to half their number is equal to 1· 3 sec. This is a much slower rate of recombination than with Röntgenized air near an ordinary Crookes tube. The amount of ionization by the uranium radiation is, in general, much smaller than that due to Röntgen rays, so that the time taken for the ions to fall to half their number is longer . The phenomenon of recombination of the ions is very similar in both uranium and Röntgen conduction. In order to test whether the rate of recombination of the ions is proportional to the square of the number present in the gas, the following experiment was performed: A tube A (Fig. 10) was taken, 3 m.long and 5·5 cm. in diameter. A cylinder D, 25 cm. long, had its interior surface coated with uranium oxide. This

* Phil.

Mag., November 1897.

L °ral/iulI/ Radiation alUl fhe Electrical COllducl;OIl Produced by lt

199

cylinder just fitted the large tube, and its position in the tube could be varied by means of strings attached to it, which passed through corks at the ends of the long tube. The air was forced through the tube from a gasometer, and on entering the tube A passed through a plug of cotton-wool, E, in order to remove dust from the air and to make the current of air more uniform over the cross-section of the tube. The air passed by the uranium surface and then through agauze L into the testing cylinder B of 2· 8 cm. diameter. An insulated rod C, 1·6 cm. in diameter, passed centrally through the cylinder Band was connected with the electrometer. The cylinders A and B were connected to one pole of a battery of 32 volts. the other pole of which was to earth.

B _

B _

E

A

L

URAN/UM CYI.INDER

D

EARTH

EARTH

Fig. 10 The potential difference of 32 volts between Band C was sufficient to almost completely remove an the ions from the gas in their passage along the cylinder. The rate of leak of the electrometer was thus proportional to the number of ions in the gas. The following rates of leak were obtained for different distances of the uranium cylinder from the gauze L. Distance of Uranium cylinder from L

T

d cl ~ 25 cm. d.:..... 50cm. d d 100cm. d -':'0200 cm.

t t + 1 sec. t -i- 2 sec. t -+- 4 sec. t -t 8 sec.

Rate of leak in scale divisions Calculated per minute rates of leak

*159 111 *87 62 39

*159 112 *87 60 37

The first column of the table gives the distances of the end of the uranium cylinder from the gauze L. d (about 20 cm.) was the distance for which the first measurement was made. In the second column the time intervals taken for the air to pass over thc various distances are given. The value of t corresponds to the distance d. The mean velocity ofthe current of air along the tube was about 25 cm. per sec.

200

The Collected Papers 0/ Lord Rutherford

In the third column are given the observed rates of leak, and in the fourth column the calculated values. The values were calculated on the assumption that the rate of recombination of the ions was proportional to the square of the number present, i.e. that

-dn = _ an2 dt

'

where n is the number of ions present and cx is a constant. The two numbers with the asterisk were used to determine the constants of the equation. The agreement of the other numbers is c10ser than would be expected, for in practice the velocity of the blast is not constant over the cross-section, and there is also a slight loss of conductivity of the gas due to the diffusion of the ions to the side of the long tube. It will be observed that the rate of recombination is very slow when a small number of ions are present in the gas, and that the air preserves one quarter of its conducting power after an interval of 8 sec. § 14. Velocity

0/ the Ions

The method* adopted to determine the velocity of the ions in Röntgen conduction cannot be employed for uranium conduction. It is not practicable to measure the rate of recombination of the ions between the plates on account of the very sma11 after-conductivity in such a case; and, moreover, the inequality of the ionization between the two plates greatly disturbs the electric field between the plates. A comparison of the velocities, under similar conditions, of the ions in Röntgen and uranium eonduetion ean, however, be readily made. The results show that the ions in the two types of eonduetion are the same. In order to compare the veloeities an apparatus similar to Fig. 10 was used. The ions were blown by a charged wire A, and the eonduetivity of the gas tested immediately afterwards at an electrode B, whieh was fixed elose to A. The eleetrode A was eylindrieal and fixed centrally in the metal tube L, whieh was connected to earth. For convenienee of caleulation it is assumed that the eleetric field between the eylinders is the same as if the eylinders were infinitely long. Let a, b be the radü of the electrode A and the tube L (internal) ; Let V be the potential of A (supposed positive). The electromotive intensity X (without regard to sign) at a distance r from the centre of the tube is given by

v

* fhit.

X=b r logea Mag., November 1897.

UraniulI/ Radiation and (he Electrical Conduction Produced by It

201

Let u1 U2 be the velocities of the positive and negative ions for a potential gradient of 1 volt per cm. If the velocity is proportional to the electric force at any point, the distance dr traversed by the negative ion in the time dt is given by dr =

XU2

dt,

b loge- r dr

a dt=--

or

VU2

Let 1'2 be the distance from the centre from which the negative ion can just reach the electrode in the time ( taken for the air to pass along the electrode. Then (rl- a2 ) b t = _w> loge-' U2 a If P2 be the ratio of the number of the negative ions that reach the electrode A to the total number passing by, r2 2 -

then

P2 = b2 _ a2

pib 2 Therefore

a2

U2

=

b

a 2) loge ~

-

2V. t

(1)

Similarly the ratio PI of the number of positive ions that give up their charge to the extern al cylinder to the total number is given by PI(b 2 u\ =

-

b a 2) loge Ci

2V. t

(2)

In tbe above equations it is assumed that tbe current of air is uniform over tbe cross-section of tbe tube, and tbat the ions are uniformly distributed over the cross-section; also, tbat tbe movement of tbe ions does not appreciably disturb tbe electric field. Since the value of t can be calculated from the velocity of the current of air and the length of the electrode, the values of the velocities of the ions under unit potential gradient can at once be determined. The equation (l) shows that P2 is proportional to V, Le. that the rate of leak ofthe electrode A varies directly as the potential of A, provided the value of V is not large enough to remove all the ions from the gas as it passes by the electrode. This was experimentally found to be the case. In the comparison of the velocities the potential V was adjusted to such a value that P2 was about one-half. This was determined by testing the rate of leak at B with a saturating electrornotive force. The arnount of recombination G*

202

The Collected Papers of Lord Rutherford

of the ions between the electrodes A and B was vcry smalI, and could be neglected. The uranium cylinder was then removed, all the other parts ofthe apparatus remaining unchanged. An aluminium cylinder was substituted for the uranium cylinder, and x-rays were allowed to faH on the aluminium. The bulb and induction-coil were placed in a metal box in order to carefully screen off all electrostatic disturbanees. The rays were only allowed to fall on the central portion ofthe cylinder. The intensity ofthe rays was adjusted so that, with the same current of air, the rate of leak was comparable with that produced by the uranium. It was then found that the value of P2 was nearly the same as for the uranium conduction. For example, the rate ofleak of B was reduced from 38 to 14 scale-divisions per minute by charging A to a certain small potential, when the air was blown by the surface of the uranium. When Röntgenized air was substituted, the rate of leak was reduced from 50 to 18 divisions per minute under the same conditions. The values of P2 were O· 63 and o· 64 respectively. This agreement is eloser than would be expected, as the bulb was not a very steady source of radiation. This result shows that the ions in Röntgen and uranium conduction move with the same velocity and are probably identical. The velocity of an ion in passing through agas is proportional to !!...., where e is the charge carried by the m ion, and m its mass. Unless e and m vary in the same ratio it follows that the charge carried by the ion in uranium and Röntgen conduction is the same, and also that their masses are equal. It was found that the velocity ofthe negative ion was somewhat greater than that of the positive ion. This has been shown to be the case of ions produced by Röntgen rays. * The difference of velo city between the positive and negative carrier is readily shown. The rate of leak of B is observed when charged positive]y and negatively. When B was charged positively the rate of leak measured the number of negative ions that escaped the electrode A, and when charged negatively the number of positive ions. The rate of leak was always found to be slightly greater when B was charged negatively. This is true whether Ais charged positively or negatively, and shows that there is an excess of positive ions in the gas after passing by the electrode A. The difference ofvelocities ofthe ions can also readily be shown by applying an alternating electromotive force to the electrode A sufficient to remove a large proportion of the ions as the air passes by. The issuing gas is always found to be positively charged, showing that there is an excess of positive over negative ions. A large number of determinations of the velocities of the uranium ions have been made, with steady and alternating electromotive forces, when the air passed between concentric cylinders or plane rectangular plates. In consequence of the inequality of the velocity of the current of air over the

* Zeleny,

Phi!. Mag., July 1898.

[!J'anium Radiation and the Electrical Conduction Produced by It

203

cross-section of the tube, and other disturbing factors which could not be allowed for, the determination could not be made with the accuracy that was desired. For an accurate determination, a method independent of currents of air is very desirable. ~

15. Potential Gradient between two Plates

The normal potential gradient between two plates is altered by the movement of the ions in the electric field. Two methods were used to determine the potential gradient. In the first method a thin wire or strip was placed between two parallel plates one of which was covered with uranium. The wire was connected with the electrometer, and after being left some time took up the potential in the air elose to the wire. In the second method the ordinary mercury- or water-dropper was employed to measure the potential at a point. For the first method two large zinc plates were taken and placed horizontal and parallel to one another. A layer of uranium oxide was spread over the lower plate. The bottom plate was connected to one pole of a battery of 8 volts, and the top plate was connected to earth. An insulated thin zinc strip was placed between the plates and parallel to them. The strip was connected with the electrometer, and gradually took up the potential of the point. By moving the strip the potential at different points between the plates could be determined. The following table is an example of the results obtained. PLATES

4·8 cm.

Distance from top plate

o

0·6

1·2 2·1 3 ·1 4·8

APART;

8

VOLTS BETWEEN PLATES

Potential in volts with Uranium

o

2·5

3·8

5·9 7 8

Potential in volts without Uranium

o 1

2 3·5 5·2 8

The third column is calculated on the assumption that without the uranium the potential falls off uniformly between the plates. The method given above is not very satisfactory when the strip is elose to the plates, as it takes up the potential ofthe point very slowly. The water- or mercury-dropper was more rapid in its action, and gave results very similar to those obtained by the first method. Two parallel brass plates were placed vertically and insulated. One plate was connected to the positive and the other to the negative pole of a battery. The middle point of the battery was placed to earth. The water-dropper was connected with thc

204

The Collected Papers 0/ Lord Rutherford

electrometer. The potential at a point was first determined without any uranium near. One plate was then removed, and an exact1y similar plate, covered witb tbe uranium compound, substituted. Tbe potential of tbe point was tben observed again. In tbis way tbe potential at any point witb and witbout the uranium could be determined. The curve shown in Fig. 1l is an example of the potential gradient observed between two parallel plates 6·6 cm. apart. The dotted line represents the potential gradient wben the uranium is removed. The ordinates represent volts and the abscissae distances from the plate covered with the uranium compound.

-2

-2 ~

.~

s: ~

o -~ "'t

()ISTAIVCE 8E7WEEN"'\2LATGJ IN cM.t. 6 -2 -241 -22

~

-~

~

-2

-2

~

Fig. 11

It will be observed that tbe potential gradient is diminisbed near the uranium and increased near tbe other plate. The point of zero potential is displaced away from the uranium. From curves showing the potential gradient between two plates, the distribution of free electrification between the plates can be deduced. By taking the

first differential of the curve we obtain ~'V , tbe electric force at any point, and by taking the second differential of the

~e we obtain ~;, wbicb is equal

to - 47Tp, wbere p is tbe volume density of electrification at any point. In order to produce the disturbance of the electric field sbown in Fig. 11, there must be an excess of ions of one kind distributed between the plates. Such a result follows at once from what has been said in regard to the inequality of tbe ionization between the plates due to the absorption of tbe radiation.

Uranium Radiation and the Elecu'ical Conduction Produced by I

205

It was found that the potential gradient approached more and more its undisturbed value with increase of the electromotive force between the plates. The dis placement of the point of zero potential from the uranium surface increased with diminution of electromotive force. For example, for two plates 51 mm. apart, charged to equal and opposite potentials, the points of zero potential were 28, 30, 33 mm. from the uranium when tbe differences of potential between the plates were 16, 8, and 4 volts respectively. When the uranium was charged positively, the point of zero potential was more displaced tban when it was charged negatively. This is due to the slower velocity of the positive ion. The slope of potential very elose to the surface of the uranium has not been investigated. The deviation from the normal potential slope between the plates depends very largely on the intensity of the ionization produced in the gas. With very weak ionization the normal potential gradient is only slightly affected. Child* and Zelenyt have shown tha~ the potential gradient between two parallel plates exposed to Röngten rays is not uniform. In their cases the ionization was uniform between the plates, and the disturbance in the field manifested itself in a sudden drop at both electrodes. In the case considered for uranium radiation, the ionization is too small for this effect to be appreciable. The disturbance of the field is due chiefly to the inequality of tbe ionization, and does not only take place at the electrodes. § 16. Relation between Current and Electromotive Force

The variation with electromotive force of the current through agas exposed to uranium radiation has been investigated by Becquerel,! and later by de Smolan and Beattie. § The general relation between the current through the gas and the E.M.F. acting on it is very similar to that obtained for gases exposed to Röntgen radiation. The current at first increases nearly proportionally with the E.M.F. (provided the E.M.F.'s of contact between the metals are taken into account), then more slowly, till finally a stage is reached, which may be termed the 'saturation stage', where there is only a very slight increase of current with a very large increase of electromotive force. As far as experiments have gone, uranium oxide, when immersed in gases which do not attack it, gives out a constant radiation at adefinite temperature, and the variation of the intensity of radiation with the temperature over the ordinary atmospheric range is inappreciable. For this reason it is possible to do more accurate work with uranium radiation than with Röntgen radiation, for it is almost impossible to get a really steady source of x-rays for any length of time. It was the object of these experiments to determine the relation between current and electromotive force with accuracy, and to see whether the gas really becomes saturated, Le. whether the current appreciably increases with * Wied. Annal., April 1898, p. 152. t Phi!. Mag., July 1898. ~

Comptes Rendus, pp. 438, 800 (1897).

§ Phi!. Mag., xliii, p. 418 (1897).

206

The Collected Papers of Lord Rutllerford

electromotive forces when the electromotive forces are great, but still not suffieient to break down the gas and to produee eonduetion in the gas without the uranium radiation. A null method was devised to measure the eurrent, in order to be independent of the eleetrometer as a measuring instrument and to merely use it as an indieator of differenee of potential. Fig. 12 shows the general arrangement of the experiment. A and B were two insulated parallel zine plates: on the lower plate A was spread a uniform layer of uranium oxide. The bottom plate was eonneeted to one pole of a battery of a large number of storage eells, the other pole of which was to earth. The

E

T

TT

TT

B EARTH

IIRIINIU/tl _ _A

EARTII

E EARTII

c

D

F

Fig. 12 insulated plate B was connected to one pair of quadrants of an eleetrometer, the other pair of which was to earth. Under the infiuenee of the uranium the air between the plates A and B is made a partial conductor, and the potential of B tends to become equal to that of A. In order to keep the potential of B at zero, B is connected through a very high resistance T of xylol, one end of whieh is kept at a steady potential. If the amount of electricity supplied to B through the xylol by the battery is equal and opposite in sign to the quantity passing between A and B, the potential of B will remain steadily at zero. In order to adjust the potential to be applied to one end of the xylol-tube T, a battery was conneeted through resistance boxes R 1 R 2 • the wire between being eonneeted to earth. The ratio ofthe E.M.F. e aeting on Tto the E.M.F. E of the battery is given by e R1 E= R 1 +R2' In praetiee, R l + R 2 was always kept constant and equal to 10,000 ohms, and, in adjusting the resistance, plugs taken from one box were transferred to the other. The value of eis thus proportional to Rh and the amount of eurrent supplied to B (assuming xylol obeys Ohm's Law) is proportional to R l • Ifthe

l

raniUll1

Radiatio11 alUl the Electrical Conduction Produced by It

207

resistances are varied till the electrometer remains at the 'earth zero', the current between the plates is proportional to R I • If the value of the E.M.F. applied is too great the needle moves in one direction, if too small in the opposite direction. For fairly rapid leaks the current could be determined to an accuracy of 1 per cent; but for slow electromotive leaks this accuracy is not possible on account of slow changes of the electrometer zero when the quadrants are disconnected. The following tables show the results of an experiment with uranium oxide. The surface of the uranium was 14 cm. square. In order to get rid of stray radiation at the sides lead strips, which nearly reached to the top plate, were placed round the uranium. 16 volts were applied to the resistance box, and a resistance of 10,000 ohms kept steadily in the circuit. Plates 2·5 cm. apart Volts Current R 1

0·5 1 2 4 8

16 37·5 112 375 800

425 825 1,570 2,750 3,750 4.230 4,700 5,250 5.625 5.825

Plates 0'5 cm. apart Volts Current R]

0·125 0·25 0·5 1 2 4 8

16

100 335

1.400

2,800 4,300 5.250 5.650 6,200 6,670 6,950 7,400 7,850

Under the column of volts the difference of potential between A and Bis given. The current is given in terms of the resistance R 1 required to keep the electrometer at the earth zero. It will be observed that for the first few readings Ohm's Law is approximately obeyed, and then the current increases more gradually till for large E.M.F.s the rate ofincrease is very slow. For the p1ates O· 5 cm. apart the rate of leak for 335 volts is only 50 per cent greater than the rate of leak for 1 volt. The same general results are obtained if the surface of the uranium is bare or covered with thin metal. The disadvantages of covering the surface with tin or aluminium foil are (1) that the intensity of the radiation is considerably decreased; (2) that the ions diffuse from under the tinfoil through any small holes or any slight openings in the side. The drawback of using the uncovered uranium in the form of fine powder, is that under large e1ectric forces the fine uranium particles are set in motion between the plates and cause an additional leakage. In practice, the rate of leak was measured with potential differences too small to produce any appreciable action of this kind. In order to investigate thc current-e1ectromotive-force relations for different gases the same method was used, but the leakage in this case took place between two concentric cylinders. The apparatus is shown in the lower part of

208

The Collected Papers ofLord Rutherford

Fig. 12: C and D were two concentric cylinders ofbrass 4·5 and 3·75 cm. in .diameter, insulated from each other.The ends ofthe cylinder D were closed by ebonite collars, and the central cylinder was supported in position by brass rods passing through the ebonite. The surface C was uniformly covered with uranium oxide. The cylinder D was connected to one pole of a battery, the other pole of which was to earth. The cylinder C was connected to the electrometer. The following tables show the results obtained for hydrogen, carbonic acid, and air. Distance between cylinders O' 375 cm. HYDROGEN

CARBONIC ACID

AIR

Volts

Current

Volts

Current

Volts

Current

0 -0,062 0·125 0·25 0'5 1 2 4 8 16 108 216

122 125 123 142 150 160 163 165 168 172 178 185

0 -0,125 0·25 0·5 1 2 4 8 16 36 108 216

95 205 255 305 355 405 460 520 590 705 787 820

+1 2 4 8 36 108 216

418 451 495 533 601 615 630

The above results are expressed graphically in Fig. 13, where the ordinates represent current on an arbitrary sca1e and the abscissae volts. In the tables given for hydrogen and carbonic acid it will be observed that the current has a definite value when there is no external electromotive force acting. The reason for this is probably due to the contact difference of potential between the uranium surface and the interior brass surface of the outside cylinder. When the external cylinder was connected to earth the inside cylinder became charged* to -0,12 volt after it was left a short time. In consequence of this action, for small electromotive forces the rates ofleak are different for positive and negative. Results of this kind are shown more clearly in Fig. 14, which gives the current-electromotive-force curves for hydrogen and carbonic acid for small voltages. When there is no extern al electromotive force acting, the current has a fixed value; if the uranium is charged positively, the current increases slowly with the voltage; when the uranium is charged negatively, the current is at first reversed, becomes zero, and rapidly increases with the • This phenomenon has been studied by Lord Kelvin, Beattie, and S. de Smolan, and it has been shown that metals are charged up to small potentials under the inftuence of uranium radiation. The steady difference of potential between two metal plates between which the radiation falls is the same as the contact difference of potential. An exactly similar phenomenon has been stlldied by Perrin (Comptes Rendus, cxxiii, p. 496) for x-ra)'s.

Uraniul1I Radiation ami fhe Electrical Conduction Produced by /1

CARB!NIC ACID AS

AIR

....

~

HY/JI ()6~N

~

\.)

·375 cms

CYlINlJeRS

APART

VOLTS 40

I

60

ItO

160

Fig. 13

zo ~N

I ()6 HY/J

zo

IC

R

N BO

CA

~ ~ ~ ~

IC

AC

N

GE

O DR

HY

VO I. TS -$

zo Fig. 14

-$

200

209

210

The Collected Papers of Lord RutlzeJ:{ord

voltage until for about one volt between the plates the positive and negative currents are nearly equal. The curve for carbonic acid with a positive charge on the uranium is also shown. It will be seen that the initial slope of the curve is greater for carbonic acid than for hydrogen. It is remarkable that the current with zero E.M.F. for hydrogen is about two-thirds of its value when 216 volts are acting between the plates. The ions in hydrogen diffuse more rapidly than in air, and, in consequence, a large proportion of the negative ions reach the uranium and give up their charge to it before recombination can take place. If the radiation fell between two plates of exactly the same metal, the inequality between the positive and negative current values for low voltages would almost disappear, but even in that case there would still be an apparent current through the gas, due to the fact that the negative carriers diffuse with greater rapidity than the positive. Effects of this kind have been studied for Röntgen radiation by Zeleny.* For large E.M.F.s no appreciable difference in the value of the current could be detected whether the uranium was positively or negatively charged, i.e. positive and negative electrifications are discharged with equal facility. For the different gases the current tends more rapidly to a saturation value in hydrogen than in air, and more rapidly in air than in carbonic acid. In all these cases there is still a slight increase of current with increase ofE.M.F. long after the 'knee' of the saturation curve has been passed, and in no case has complete saturation been observed at atmospheric pressure, even for a potential gradient of 1,300 volts per cm. The explanation of the general form of the curves showing the relation between current and electromotive force for ionized gases has been given in a previous paper. t In the case of uranium conduction the phenomenon is still further complicated by the want of uniformity of ionization between the plates and the resulting disturbance of the electrostatic field due to the excess of ions of one kind between the plates. The ionization of the gas is greatest near the uranium surface, and falls off rapidly with the distance. The rate of recombination of the ions thus varies from point to point between the plates, being grea test near the surface of the uramum. The equations which express completely the relation between the current and electromotive force for the rate of leak between two parallel plates, one of which is covered with uranium, are very complex and cannot be expressed in simple form. The disturbance of the electrostatic field between the plates, due to the movement of the ions, has to be considered as wen as the variable rates of recombination at the different points, and the difference of velocity between the positive and negative ions. The great difficulty in producing complete saturation, i.e. to reach a stage ... Phi!. Mag., July 1898. t J. J. Thomson and E. Rutherford, Phi!. Mag., November 1896.

('rünilllll RadiL1fiol/ ami '"e Electrical COl/duction Produced by 11

211

when all the ions produccd reach thc clectrodes, may be due to one or more of three causes: (l) Rapid rate of recombination of the ions very ne ar the surface of the uramum. (2) Presence of very slow moving ions together with the more rapidly moving carriers. (3) An effect of the electric field on the production of the ions. The effect of (3) is probably very smalI, for there is no experimental evidence of any such action unless the electromotive forces are very high. That the slow increase of the current in strong fields is due to (1) rat her than (2) receives some support from an experiment that has been recently tried. Instead of measuring the current with the uranium covering one electrode, the air which had passed over uranium was forced between two concentric cylinders between which the electromotive force was acting. The rate of leak was found to only increase 2 or 3 per cent when the E.M.F. was increased from 16 to 320 volts. This increase is much smaller than in the results previously given. Since the effect of (2) would be present in both cases, this experiment seems to show that the difficulty in removing an the ions from the gas is not due to the presence of some very slow-moving carriers.

Effect

0/ Pressure

Some current electromotive-force curves for small voltages have been obtained at different pressures. Examples of the results are shown in Fig. 15, which gives the relation between the current and the electromotive force at pressures of 760, 380, 190, and 95 mm. of mercury. These results were obtained with a different apparatus and by a different method to that given in Fig. 12. Two parallel insulated metal plates, about 3 cm. apart. one of which was covered with uranium oxide, were placed inside an airtight vessel. One plate was connected to earth and the other to the electrometer. The plate connccted to the electrometer was then charged up to a potential of 10 volts. On account of the presence of the uranium oxide the charge slowly leaked away, and the rate of movement of the electrometerneedle measured the current corresponding to different values of the electromotive force. The method did not admit of the accuracy of that previously employed (see Fig. 12). The rate of leak for small fractions of a volt could not be determined, so that in the curves (Fig. 15) it is assumed that the current was zero when the electromotive force was zero. This is probably not quite accurate owing to the slight contact difference of potential between the plates, so that there was a small initial current for zero external electromotive force. The general results show that the gas tends to become more readily saturated with diminution of pressure. The variation of the current with the E.M.F.

212

The Collected Papers

0/ Lord Ruther/ord

depends on two factors-the velocities of the ions, and their rate of recombination. Some experiments on the velocity of the carriers· in ultra-violet light conduction showed that the velocity of the ions in a given electric field is inversely proportional to the pressure. This is probably also true for the ions in Röntgen conduction; so that under the pressure of 95 mm. the ions would move eight times as fast as at atmospheric pressure. The variation of the rate of recombination with pressure has not yet been determined.

AI R /90 MM • • 71YDIlIJSEII--."'OMIII. to-

~

AI R /90 MM •

~ ~

1I1L7S.

2

4

4

4

Fig. 15 The curve for hydrogen at atmospheric pressure is also given in Fig. 15, and shows that hydrogen is about as easily saturated as air at 190 mm. pressure.

§ 17. Separation

0/ the

Positive and Negative Ions

It is a simple matter to partially separate the positive and negative ions in

uranium conductions and produce an electrified gas. The subject of the production of electrification by passing a current of air over the surface of uranium enclosed in a metal vessel has been examined by Beattie, t who found the electrification obtained was ofthe same sign as the charge on the uranium. His results admit of a simple explanation on the theory of ionization. The gas near the surface of the uranium is far more strongly ionized than that some distance away on account of the rapid absorption of the radiation by the air.

* Proc. Comb. Phi!. Soc., February 21,1888.

t Phi!. Mag., July 1897, xliv. p. 102.

Uranium Radiation and the Electrical Conduction Produced by It

213

For convenience of explanation, let us suppose a piece of uranium, charged positively, placed inside a metal vessel connected to earth, and a current of air passed through the vessel. Under the influence of the electromotive force the negative ions travel in towards the uranium, and the positive ions towards the outer vessel. Since the ionization is greater near the surface of the uranium, there will be an excess of positive ions in the air some distance away from the uranium. Part of this is blown out by the current of air, and gives up its charge to a filter of cotton-wool. The total number of negative ions blown out in the same time is much less, as the electromotive intensity, and therefore the velocity of the carrier, is greater near the uranium than near the outside cylinder. Consequently there is an excess of positive ions blown out, and a positively electrified gas is obtained. As the potential difference between the electrodes in increased, the amount of electrification obtained depends on two opposing actions. The velocity of the carriers is increased, and consequently the ratio of the number of carriers removed is diminished. But if the gas is not saturated, with increase of electromotive force the number of ions travelling between the electrodes is increased, and for small voltages this increase more than counterbalances the diminution due to increase of velocity. The amount of electrification obtained will therefore increase at first with increase of voltage, reach a maximum, and then diminish; for when the gas is saturated no more ions can be supplied with increase of electromotive force. This is exactly the result which Beattie obtained, and which I also obtained in the case ofthe separation of the ions of Röntgenized air. The fact that more positive than negative electrification is obtained is due to the greater velocity with which the negative ion travels. (See § 14.) The properties of this electrified gas are similar to that which has been found from Röntgen conduction. The opposite sign of the electrification obtained by Beattie for uranium induction, and by myself for Röntgen conduction, * is to be expected on account of the different methods employed. For obtaining electrification from Röntgenized air a rapid current of air was directed dose to the charged wire. In that case the sign of the electrification obtained is opposite to that of the wire, as it is the carriers of opposite sign to the wire which are blown out before they reach the wire. In the case of uranium the current of air filled the cross-section of the space between the electrodes; and it has been shown that under such conditions electrification of the same sign as the uranium is to be expected.

§ 18. Discharging Power

0/ Fine Gauzes

Air blown over the surface of uranium loses all trace of conductivity after being forced through cotton-wool or through any finely divided substance. In this respect it is quite similar to Röntgenized air. The discharging power of cotton-wool and fine gauzes is at first sight surprising, for there is considerable evidence that the ions themselves are of molecular dimensions, and might

* Phi!. Mag., April 1897.

214

The Collected Papers 0/ Lord Ruther/ord

therefore be expected to pass through small orifices; but a little consideration shows that the ions, like the mo1ecu1es, are continually in rapid motion, and, in addition, have free charges, so that whenever they approach within a certain distance of asolid body they tend to be attracted towards it, and give up their charge or adhere to the surface. On account ofthe rapidity of diffusion* ofthe ions, the discharging power of a meta1 gauze, with openings very 1arge compared with the diameter of a carrier, may be considerab1e. The tab1e below gives some results obtained for the discharging power of fine copper gauze. The copper gauze had two strands per millimetre, and the area occupied by the metal was rough1y equa1 to the area of the openings. The gauzes filled the cross-section ofthe tube at A (Fig. 9), and were tight1y pressed together. The conductivity of the air was tested after its passage through the gauzes, the velocity of the air along the tube being kept approximately constant. The rates ofleak per minute due to the air after its passage through different numbers of gauzes is given below. Number of Gauzes

o 1 2

3 4 5

Rate of leak in divisions per minute

44

32·5 26·5

19'5 10·5 6

After passing through five gauzes the conductivity of the air has fallen to Iess than one-seventh of its original value. Experiments were tried with gauzes of different degrees of coarseness with the same general result. The discharging power varies with the coarseness ofthe gauze, and appears to depend more on the ration of the area of metal to the area of the openings than on the actua1 size of the opening. If a copper gauze has such apower of removing the carriers from the gas, we can readily see why a small plug of cotton-wool should compietely abstract the ions from the gas passing through it. The rapid 10ss of conductivity is thus due to the smallness of the carrier and the consequent rapidity of diffusion. §

19. General Remarks

The cause and origin of the radiation continuous1y emitted by uranium and its salts still remain a mystery. All the results that have been obtained point to the conclusion that uranium gives out types of radiation which, as regards their effect on gases, are similar to Röntgen rays and the secondary radiation emitted by metals when Röntgen rays fall upon them. If there is no polarization or refraction the similarity is compiete. J. J. Thomsont has suggested that • Townsend, Phi!. Mag., June 1898.

t Proc. Camb. Phi!. Soc., vol. ix, pt. viii, p. 397 (1898).

L ral/iulII Radiatiol/ wu/ the E/c('t/'i('({/ tOllductiol/ P/'oc!lIced hy 11

215

thc regrouping of the constitucnts of thc atom may give rise to electrical effects such as are produced in the ionization of agas. Röntgen's* and Wiedemann's';' results see m 10 show that in thc process of ionization a radiation is emitted which has similar properties to easily absorbed Röntgen radiation. The energy spent in producing uranium radiation is probably extremely small, so that the radiation could continue for long intervals oftime without much diminution of internal energy of the uranium. The effect of the temperature of the uranium on the amount of radiation given out has been tried. An arrangement similar to that described in § 11 was employed. The radiation was completely absorbed in the gas. The vessel was heated up to about 200 0 C.; but not much difference in the rate of discharge was observed. The results of such experiments are very difficult to interpret, as the variation of ionization with temperature is not known. I have been unable to observe the presence of any secondary radiation produced when uranium radiation falls on a metal. Such a radiation is probably produced, but its effects are too small for measurement. In conclusion. I desire to express my best thanks to Professor J. J. Thomson for his kindly interest and encouragement during the course of this investigation. Cavendish Laboratory September I, 1898

* Wied.

Annal., lxiv (1898).

t Zeit. f.

Electrochemie, ii., p. 159 (1895).

Thorium and Uranium Radiation (preliminary note)

by E. R UTHERFORD, M.A., D.se. Macdonald Professor of Physics, McGill University, Montreal and R. B. OWENS, E.E. Tyndall Fellow Columbia University, Macdonald Professor 0/ Electrical Engineering, McGill University, Montreal

From Transactions ofthe Royal Society ofCanada, 1899, section III, vol. 2, pp. 9-12 (Presented by Professor Cox, read May 26th, 1899)

IN 1896 Becquerel discovered that the compounds of the meta! uranium continuously emitted aradiation similar in character to Röntgen rays. These rays have the property of acting on a photographie plate in the dark, and of making the gas through which they pass a partial conductor of electricity. They also have the power of passing through considerable thicknesses of metal and in general behave very similarly to X-rays emitted from a so-called "soft" tube. In 1897 Schmidt (Wied. Annal, May, 1898) found that the compounds of thorium had the same property as uranium. In a paper by one of us (Rutherford, Phi!. Mag., lan., 1899) the radiation emitted by uranium has been considered in detail. The present paper is an extension of the investigation to the radiation emitted by thorium and its compounds, and a comparison of the two types of radiation. The methods of investigation are similar to those used in the previous work. The intensities of the radiations were compared electrically by measuring the rate of discharge of electrification produced by the rays. A layer of the radio-active substance was uniformly spread on a small platinum plate which rested on the top of a larger brass one. An insulated parallel brass plate about 4 cms. distant was connected to one pair of quadrants of a delicate electrometer, the other pair of which was earthed. The lower plate was connected to one pole of a battery of 95 volts, the other pole of which was also to earth. When the quadrants of the electrometer were separated the top plate gradually acquired the potential of the lower plate, and the rate of movement of the electrometer needle was taken as a measure of the current through the gas; the gradual charging of the top plate being due to the movement of the charged partic1es or ions produced by the radiation throughout the volume of the gas. It has been shown (loc. cit.) that the

Thorium and Uranium Radiation

217

intensity of the radiation is proportional to the current through the gas, i.e., to the rate of movement of the electrometer needle when a large E.M.F. acts between the plates. It was found that layers of this aluminium foil over the active substance cut off the intensity of the radiation gradually, but not so rapidly as in the case of uranium. The decrease of the intensity for the first few layers of aluminium foil follows a regular absorption law. After a certain thickness of foil has been added thc intensity of the transmitted radiation diminishes very slowly and a much larger proportional thickness of metal is required to reduce the intensity still further. These results show that the radiation emitted from thorium is complex, and that different types of radiation are emitted, some of which are more readily absorbed than others. Using aluminium foil, the amount of the more penetrating type of radiation depends largely on the thickness of the layer of the active substance. With thin layers of the radio-active material, the rate of leak due to the more penetrating rays is hardly measurable compared with that due to the more absorbable rays. The action of paper in cutting down the intensity of radiation is worthy of remark. The first thickness of ordinary foolscap paper cuts down the intensity of the radiation to about two-thirds of its value. The addition of successive layers changes the intensity but litde; an extra layer, several millimetres thick, is quite transparent to the rays. Uranium radiation is in this respect different from thorium radiation, for a few thicknesses of paper are almost opaque to the uranium rays. It was found that a11 the compounds of thorium examined, viz., thorium oxide, thorium sulphate, and thorium nitrate gave out the same kinds of radiation as measured by the transparency of aluminium foil for the rays. The intensities of the radiation differed largely; for equal weights of active substance being greatest for thorium oxide. The nature of the radiations is thus independent of the particular state of chemical combination of the compound, but depends only on the presence of thorium in the material. Thorium radiation is absorbed in its passage through the air, but not so rapidly as the uranium rays. Before an infinite plane of thorium oxide the intensity of the radiation at a distance of 10 mms. from the surface would only be one-half of that at the surface. For uranium there would be the same diminution in about 4 mms. By means of a specially constructed apparatus, the absorption of thorium radiation at different pressures was examined. From apressure of one-half to four atmospheres the absorption was found to be direct1y proportional to the pressure of the gas. The method of measurement was similar to that described in a previous paper. It was very early observed that the radiation from thorium oxide was not constant, but varied in a most capricious manner. This was the more peculiar as the sulphate and nitrate were fairly constant. All the compounds of

218

The Collected Papers 01 Lord Ruthel:ford

uranium also give out a radiation which remains remarkably constant and probably varies very little with time. Becquerel has found that a specimen of uranium which had been kept in a dark room for two years gave out the same intensity of radiation as at first. The inconstancy of the radiation from thorium oxide was examined in detail, as it was thought it might possibly give some eIue as to the cause and origin of the radiation emitted by these substances. It was found that if the substance was inclosed in a lead box with a door, the rate of leak was much slower with the door open than eIosed. The addition of a slight draught of air caused by opening or shutting the door of the room diminished the rate of leak still more. Under similar conditions the rate of leak due to the sulphate and nitrate of thorium and the uranium compounds is not appreciably affected. The sensitiveness of thorium oxide to slight currents of air is very remarkable, and made it difficult to work with. With the air quite still, the substance in a few minutes regained its normal activity. The recovery was quite gradual. On covering the radioactive substance with aluminium foil the action was reduced, but was still quite marked. In order to investigate the matter in more detail, the substance was placed in an air-tight vessel. When a current of air was passed through by means of a water pump on a pair of bellows, the rate of leak rapidly diminished to ab out one-third of its value, and did not change much with considerable increase in the velocity of the blast. On cutting off the current of air, the substance gradually recovered its normal state. The effect was independent of the amount of moisture present, as air bubbled through water gave the same effect as air which had passed through a column of calcium chloride and phosphorus pent-oxide. It was independent of the presence of carbonic acid. A current of coal gas gave the same effect as a current of air. The reduction of the rate of leak by a given current of air depends on the thickness of the layer of the active substance. With thin layers the diminution was small, but with Jayers several millimetres thick the action was greatly increased. Layers of thin aluminium foil over the thorium oxide only partially diminished the effect. The effect persisted unchanged if the same air was passed backwards and forwards through the apparatus by means of a suitable aspirator. This seems to show that it is not the presence of any substance in the air existing in small quantity, that produces the effect. A large number of experiments of various kinds have been tried, but so far, no eIue has been obtained as to why this action should be so manifest in thorium oxide. It appears as if in the pores of the thick layer of thorium oxide some change takes place with time, which increases the intensity of the radiation, and if the result of the action is continually removed, the intensity of the radiation is diminished. This would explain why the action is shown chiefly in thick layers, and depends on the current of air. All experiments so far tried show that the action is :pot due to the r~moval of the ions between the plates by the air

ThoriulII allli Uraniul/I Radiation

219

currents. but is rat her the result of a change in the condition of the radioactive substance at or near its surface. The phenomena of electrical conduction produced by thorium radiation are, as far as the investigations have gone, strictly similar to those produced by Röntgen rays and uranium radiation, and obey the same laws. For high press ures of air, thorium oxide behaves differently from uranium, but the difference is probably dosely connected with the action of currents of air on the intensity of the radiation, which we have already considered. A special apparatus was constructed to examine the properties of uranium and thorium radiation at high pressures. The investigation is not yet completed but has led to some interesting results. It was found that the rate of recombination of the ions is very rapid, especially at high pressures. If we take the case of thorium radiation, the intensity is reduced to one-half, after passing through 10 mms. of air at atmospheric pressure, and assuming that the absorption is proportional to the pressure the intensity would be reduced to one half in a distance of 0·5 mms. for apressure of 20 atmospheres. The ionization is thus very intense near the surface of active substances, and the ions recombine with great rapidity after they are formed, so that very large electromotive forces are required to carry them across to the electrodes before any appreciable amount of recombination has occurred. The rapidity of recombination depends largely on the amount of dust or nudei present in the air through which the ions are passing. A striking experiment to illustrate this action can be readily shown by filling the vessel in which the radio-active substance is placed with tobacco smoke. The rate of leak immediately falls to less than one-tenth of its value. The intensity of the radition emitted by the substance is unaltered, but a great number of the ions in their passage between the plates give up their charges to the nudei in the smoke, and the current is correspondingly diminished.

A Radioactive Substance emitted from Thorium Compounds by E. RUTHERFORD, M.A., B.SC., Macdonald Professor of Physics, McGill University, Montreal

From the Phi[osophica[ Magazine for January 1900, sero 5, xlix, pp. 1-14 Communicated by Professor J. J. Thomson, F.R.S.

IT has been shown by Schmidt* that thorium compounds give out a type of

radiation similar in its photographie and electrical actions to uranium and Röntgen radiation. In addition to this ordinary radiation, I have found that thorium compounds continuously emit radioactive partic1es of some kind, which retain their radioactive powers for several minutes. This 'emanation', as it will be termed for shortness, has the power of ionizing the gas in its neighbourhood and of passing through thin layers of metals, and, with great ease, through considerable thicknesses of paper. In order to make c1ear the evidence of the existence of a radioactive emanation, an account will first be given of the anomalous behaviour of thorium compounds compared with those of uranium. Thorium oxide has been employed in most of the experiments, as it exhibits the 'emanation' property to a greater degree than the other compounds; but what is true for the oxide is also true, but to a less extent, of the other thorium compounds examined, viz. the nitrate, sulphate, acetate, and oxalate. In a previous papert the author has shown that the radiation from thorium is of a more penetrating character than the radiation from uranium. Attention was also directed to the inconstancy of thorium as a source of radiation. Owenst has investigated in more detail the radiation from thorium compounds. He has shown that the radiations from the different compounds are of the same kind, and, with the exception of thorium oxide in thick layers, approximately homogeneous in character. The intensity of thorium radiation, when examined by means of the electrical discharge produced, is found to be very variable; and this inconstancy is due to slow currents of air produced in an open room. When the apparatus is placed in a c10sed vessel, to do away with air currents, the intensity is found to be practically constant. The sensitiveness of thorium oxide to

* Wied. Annal., May 1898.

t Phi!. Mag., January 1899, p. 109.

t Phil.

Mag., Oetober 1899, p. 360.

A Radioactive Substance emitted from Thorium Compounds

221

slight currents of air is very remarkable. The movement of the air caused by the opening or closing of a door at the end of the room opposite to where the apparatus is placed, is often sufficient to considerably diminish the rate of discharge. In this respect thorium compounds differ from those of uranium, which are not appreciably affected by slight currents of air. Another anomaly that thorium compounds exhibit is the ease with which the radiation apparently passes through paper. The following table is an example of the way the rate of leak between two parallel plates, one of which is covered with a thick layer of thorium oxide, varies with the number of layers of ordinary foolscap paper placed over the radioactive substance. I Thickness of each Layer of Paper = 0·008 cm. 50 volts between plates TABLE

Number of Layers of Paper

o

Rate of Discharge

1 0·74 0·74

1 2 5 10 20

0·72

0·67 0·55

In the above table the rate of leak with the thorium oxide uncovered is taken as unity. It will be observed that the first layer reduced the rate ofleak to o· 74, and the five succeeding layers produce very Httle effect. The action, however, is quite different if we use a thin* layer of thorium oxide. With one layer of paper, the rate of discharge is then reduced to less than /6 of its value. At first sight it appears as if the thorium oxide gave out two types of radiation, one of which is readily absorbed by paper, and the other to only a slight extent. If we examine the radiation given out by TABLE

11

Thickness of Paper = 0·0027 cm. Number of Layers of Thin Paper

o 1 2 3

Rate of Discharge

1 0·37 0·16 0·08

* To produce a thin layer on a plate, the oxide, in the form of a fine powder, was sprinkled by means of a fine gauze, so as to cover the plate to a very small depth. By a thick layer is meant a layer of oxide over a millimetre in thickness.

222

The Collected Papers

0/ Lord Ruther/ord

a thin layer of thorium oxide, by placing successive layers of thin paper upon it, we find the radiation is approximately homogeneous, as the Table Il (p. 221) shows. The rate of leak of the bare salt is taken as unity. If the radiation is of one kind, we should expect the rate of discharge (which is proportional to the intensity of the radiation) to diminish in geometrical progression with the addition of equal thicknesses of paper. The above figures show that this is approximately the case. With a thick layer of thorium oxide, by adding successive layers of thin paper, we find the rate of discharge gradually diminish, till after a few layers it reaches a constant value. The amount that is cut off by the first layer of foolscap paper (see Table I) is of the same kind of radiation as that which is emitted by a thin layer of oxide. On directing a slight current of air between the test plates, the rate of discharge due to a thick layer of thorium oxide is greatly diminished. The amount of diminution is to a great extent independent of the e1ectromotive force acting between the plates. Under similar conditions with uranium, the rate of leak is not appreciably affected. With a thin layer of oxide, the diminution of the rate of leak is small; but with a thick layer of oxide, the rate of leak may be reduced to less than one-third of its previous value. If two thicknesses of foolscap paper are placed over the thorium oxide, the resulting rate of leak between the plates may be diminished to less than ;}o of its value by a slight continuous blast of air from a gasometer or bellows. The phenomena exhibited by thorium compounds receive a complete explanation if we suppose that, in addition to the ordinary radiation, a large number of radioactive particles are given out from the mass of the active substance. This 'emanation' can pass through considerable thicknesses of paper. The radioactive partic1es emitted by the thorium compounds gradually diffuse through the gas in its neighbourhood and become centres of ionization throughout the gas. The fact that the effect of air currents is only observed to a slight extent with thin layers of thorium oxide is due to the preponderance, in that case, of the rate of leak due to the ordinary radiation over that due to the emanation. With a thick layer of thorium oxide, the rate of leak due to the ordinary radiation is practically that due to a thin surface layer, as the radiation can only penetrate a short distance through the salto On the other hand, the 'emanation' is able to diffuse from a distance of several millimetres below the surface of the compound, and the rate of leak due to it becomes much greater than that due to the radiation alone. The explanation of the action of slight currents of air is c1ear on the 'emanation' theory. Since the radioactive particles are not affected by an electrical field, extremely minute motions of air, if continuous, remove many of the radioactive centres from between the plates. It will be shown shortly that the emanation continues to ionize the gas in its neighbourhood for several minutes, so that theremoval ofthe particles from between the plates diminishes the rate of discharge between the plates.

..J Rwlioaci ire .')'UhSIIII/(,(! emiTfecl ji'ol11 Thoriul11 COl11pounds

/Jura/ion

(~l

the Rad;oactil'ity

223

0/ t!re Emanatioll

The emanation gradually loses its radioactive power. The following method was adopted to determine the rate of decay of the intensity of the radiation of the radioactive particles emitted by thorium oxide. A thick layer of thorium oxide was enc10sed in a narrow rectangular paper vessel A (Fig. 1), made up of two thicknesses of foolscap paper. The paper cut off the regular radiation almost entirely, but allowed the emanation to pass through. The thorium thus enclosed was placed inside a long metal tube B. C A

B

D

EAR1H

Fig. 1 One end of the tube was connected to a large insulated cylindrical vessel C, which had a number of small holes in the end for the passage of air. Inside C was fixed an insulated electrode, D, connected with one pair of quadrants of a Thomson e1ectrometer. The cylinder, C, was connected to one terminal of a battery of 100 volts, the other terminal of which was connected to earth. A slow current of air from an aspirator or gasometer, which had been freed from dust by its passage through a plug of cotton-wool, was passed through the apparatus. The current of air, in its passage by the thorium oxide, carried away the radioactive particles with it, and these were gradually conveyed into the large cylinder C. The electrometer needle showed no sign of movement until the radioactive partic1es were carried into C. In consequence of the ionization of the gas in the cylinder by the radioactive particles, a current passed between the electrodes C and D. The value ofthe current was the same whether C was connected with the positive or negative pole of the battery. When the current of air had been flowing for some minutes, the current between C and D reached a constant value. The flow of air was then stopped, and the rate of leak between C and D observed at regular intervals. lt was found that the current between C and D persisted for over ten minutes.

The Collected Papers 0/ Lord Ruther/ord

224

The fol1owing is aseries of observations. TABLE

III

Potential Difference 100 volts Time in seconds

Current

28 62 118 155 210 272 360

0·69

o

1

0·51 0·23 0·14

0·067 0·041 0·018

CURRENT

4

TIME TIME TIME TIME 2

3

.

Fig.2

4 5

Ei

.,

•

Fig. 2, curve A, shows the relation existing between the current through the gas and the time. The current, just before the fiow of air is stopped, is taken as unity. It will be observed that the current through the gas diminishes in a geometrical progression with the time. It can easily be shown, by the theory of ionization, that the current through the gas is proportional to the intensity of the radiation emitted by the radioactive particles. We therefore see that the intensity of the radiation given out by the radioactive particles falls off in a geometrical progression with the time. The result shows that the intensity of the radiation has fallen to one-half its value after an interval of about one

.1 Radioactire Substallce clI/itlct! jj'o/ll ThoriulI/ Compounds

225

minute. The rate of leak duc to the emanation was too small for measurement after an interval of ten minutes. If the ionized gas had been produced from a uranium compound, the duration ofthe conductivity, for voltages such as were used, would only have been a fraction of a second. The rate of decay of intensity is independent of the electromotive force acting on the gas. This shows that the radioactive particles are not destroyed by the electric field. The current through the gas at any particular instant, after stoppage of the flow of air, was found to be the same whether the electromotive force had been acting the whole time or just applied for the time ofthe test. The current through the gas in the cylinder depends on the electromotive force in the same way as the current through a gas made conducting by Röntgen rays. The current at first increases nearly in proportion to the electromotive force, but soon reaches an approximate 'saturation' value. The duration of the radioactivity was also tested by another method. The paper vessel containing the thorium oxide was placed inside a long brass cylinder over 200 cm. in length. A slow current of air (with a velocity of about 2 cm. per sec. along the tube) was passed over the thorium oxide along the tube, and then between two insulated concentric cylinders. The rate of leak between the two concentric cylinders (potential difference 270 volts) was observed when the air had been passing sufficiently long to produce a steady state. The rates of Ieak were observed for varying positions of the thorium oxide along the tube. Knowing the velocity of the current of air along the tube, the time taken to carry the radioactive particles to the testing apparatus could be determined. In this way it was found that the rate of decay was about the same as determined by the first method, i.e. the intensity fell to half its value in about one minute. In this apparatus experiments were also tried to see whether the radioactive partic1es moved in an electric field. The experiments on the effect of a current of air on the rate of discharge naturally suggest that possibly one of the ions was so large that it moved extremely slowly even in strong electric fields. The results obtained showed that the particles did not move with a greater velocity than lOO~OOÖ cm. per sec. for a potential gradient of one volt per centimetre; and it is probable that the particles do not move at all in an electric field. By blowing the emanation into an inductor, no evidence of any charge in the emanation could be detected. We may therefore conclude that the emanation is uncharged, and is not appreciably affected by an electric field.

Properties

0/ the

Emanation

The emanation passes through a plug of cotton-wool without any loss of its radioactive powers. It is also unaffected by bubbling through hot or cold water, weak or strong sulphuric acid. In this respect it acts like an ordinary gas. H

226

The Collected Papers

0/ Lord Ruther/ord

An ion, on the other hand, is not able to pass through a plug of cotton-wool, or to bubble through water, without losing its charge. The emanation is similar to uranium in its photographic and electrical actions. 1t can ionize the gas in its neighbourhood, and can affect a photographie plate in the dark after several days' exposure. Russell· has shown that the active agent in producing photographie action in the case of metals, paper, etc., is due to hydrogen peroxide. Hydrogen peroxide apparently has the power of passing in some way through considerable thicknesses of special substances, and in this respect the emanation resembles it. Hydrogen peroxide, however, does not ionize the gas in its neighbourhood. The action ofhydrogen peroxide on the photographie plate is purely a chemica1 one; but it is the radiation from the emanation, and not the emanation itself, that produces ionizing and photographie actions. The radioactive emanation passes through all metals if sufficiently thin. In order to make certain that the emanation passed through the material to be examined and did not diffuse round the edges, the radioactive substance was placed in a square groove of a thick lead plate. Two layers of paper were pasted tightly over the opening to cut off the regular radiation. The material to be tested was then firmly waxed down on the lead plate. The following numbers illustrate the effect of different metals. The rate of discharge, due to the emanation between two parallel plates 4 cm. apart, was observed. Aluminium Foil, thickness Number of Layers

o 1 3 6

= 0·0008 cm.

Rate of Discharge

1

0·66 0·42 0·16

Cardboard, thickness 0·08 cm. Layers

Rate of Discharge

1 2

0·40 0·21

o

1

The emanation passed readilythrough several thicknesses of gold- and silverleaf. A plate of mica, thickness 0·006 cm., was completely impervious to the emanation. When a thick layer of thorium oxide, covered over with several thicknesses of paper, is placed inside a closed vessel, the rate of discharge due to the emanation is small at first, but gradually increases, until after a few minutes a steady state is reached.

* Proc.

Roy. Soc., 1897.

A Radioarli\'e Suhstal1re emitted from Thorium

Compollnd~

227

These results are to be expected, for the emanation can only slowly diffuse through the paper and the surrounding air. A steady state is reached when the rate of loss of intensity due to the gradual decay of the radioactivity of the emanation is reeompensed by the number of new radioactive centres supplied from the thorium compound. Let n = number of ions produced per seeond by the radioactive particles between the plates. Let q = number of ions supplied per second by the emanation diffusing from the thorium. The rate of variation of the number of ions at any time t is given by dn

dl = q - An,

where Ais a constant. The results given in Table III show that the rate of diminution of the number of ions is proportional to the number present. Solving the equation, it is seen that loge(q - An) where A is a constant. When

= -At + A, n = 0;

1=0,

therefore

A = logeq.

Thus

n--q (1 -e-'At) .

-"

With a large potential difference between the test plates the current i through the gas at any time is given by i = ne,

where e is the charge on an ion.

~; =

When a steady state is reached,

0; and the maximum number N of

ions produced per second by the radioaetive particles between the plates is given by

N='i,\' and the maximum eurrent 1 is given by

1= Ne. Therefore

i

1=

1 -- e-Ät •

228

The Collected Papers of Lord RutherJol'd

The current thus increases according to the same law as a current of electricity rises in a circuit of constant inductance. This result is confirmed by an experiment on the rise of the current between two concentric cylinders. The thorium oxide enclosed in paper was placed inside the cylinder. A current of air was sent between the cylinders in order to remove the emanation as rapidly as it was formed. The current of air was then stopped and the current between the two cylinders observed, by means of an electrometer, for successive intervals after the current of air ceased. Table IV gives the results obtained. IV Length of cylinder = 30 cm. Internal diameter outer cylinder = 5·5 cm. External diameter inner cylinder = 0·8 cm. 100 volts between cylinders. TADLE

Time in seconds

o

Current in Scale Divisions per second

2·4

7·5

3·3

2,3*

6·5

4·0

10·0

12·5

5·3 6·7 9·6 12·5 18·4 24·4

13·8 17·1

19·4 22·7 25·3 25·6 25·6

30·4

48·4

The results are expressed in Fig. 2, curve B, where the ordinate represents current and the abscissa time. It will be observed that the curve of rise of the current is similar in form to the rise of an electric current in a circuit of constant inductance. The current reaches half its value about one minute after the current of air has stopped-a result which agrees with the equation given, for e-At = ~ when t = 60 sec. (see Table IV). At the instant of stopping the current of air the current has adefinite value, since most of the ions given off by the emanation, before it is blown out of the cylinders, reach the electrodes. When the source of the emanation is removed, q = 0, and the decay of the number ofions produced by the emanation is given by the equation

dn

dt =

->..n.

* Editor's Footllofe:- This and the following figures in the same column should read: 23, 40, 53, ete. This, and similar minor errors, left as in original.

A Radioactire Substance em it ted from Thorium COII/pounds

229

If n = N when t = 0, it is easily seen that 11

N = e- i ."

or

;

1= e-i.t; i.e. the current through the gas diminishes in a geometrical progression. After 20 minutes the current through the gas is only about one millionth part of its initial value. It has been shown that e-u = ~ when t = 60 sec. Therefore and

A=

N

=

Cf.-

A

1

86'

=

86q;

or the total number ofions produced per second when a steady state is reached is 86 times the number of ions supplied per second by the emanation. The amount of emanation from thorium oxide increases with the thickness of the layer. When 1 gramme of thorium oxide was spread over a surface of 25 cm.2, the amount of discharge due to the ordinary radiation had practically reached a maximum. The rate ofleak due to the emanation for the same thickness was small. With 9 grammes of oxide spread over the same area, the rate of leak due to the emanation had reached about half its maximum value, wh ich for that case corresponded to four times the rate of leak caused by the ordinary radiation. The emanation thus still preserves its radioactive properties after diffusing through several millimetres of thorium compound. The emanation is given out whatever the gas by which the thorium is surrounded. The action is very similar whether air, oxygen, hydrogen, or carbonic acid is used. The rate of discharge due to the emanation diminishes with lowering of the pressure of the air surrounding it. Only a few observations have been made, but the results seem to point to a uniform rate of emission of the emanation at all pressures ; but since the intensity of the ionization of the gas varies directly as the pressure, the rate of leak decreases with lowering of the pressure. The amount of the emanation, so far as the experiments have gone, is also independent of the quantity of water-vapour present. The power of emitting radioactive particles is not possessed to any appreciable extent by other radioactive substances besides thorium. All the compounds of thorium examined possess it to a marked degree, and it is especially large in the oxide. Two different specimens of the oxide have been used, one obtained from Schuchart of Germany, and the other from Eimer and Amend of New York. The oxide is prepared by the latter by igniting thorium nitrate obtained from monazite sand.

230

The Collected Papers 0/ Lord Ruther/ord

The amount of discharge caused by the emanation is increased several times by the conversion of the nitrate into the oxide; but at the same time, the rate of discharge due to the ordinary radiation emitted by the thorium is increased in about an equal ratio. The conversion of the nitrate into the oxide took place below a red heat. On heating in a muffle for some time at white heat, the amount of emanation continually diminished, till after four hours' exposure to the heat, the rate of discharge due to the emanation was only ,.to of the value immediately after its conversion into oxide. Both thorium oxalate and sulphate act in a similar manner to the nitrate; but the emanation is still given off to a considerable extent after continued heating. In considering the question of the origin and nature of the emanation, two possible explanations naturally suggest themselves, viz.: (1) That the emanation may be due to fine dust particles ofthe radioactive substance emitted by the thorium compounds. (2) That the emanation may be a vapour given off from thorium compounds.

The fact that the emanation can pass through metals and targe thicknesses of paper and through plugs of cotton-wooI, is strong evidence against the dust hypothesis. Special experiments, however, were tried to settle the question. The experiments of Aitken and Wilson* have shown that ordinary air can be completely freed from dust particles by repeated small expansions of the air over a water surface. The dust particles act as nuclei for the formation of small drops, and are removed from the gas by the action of gravity. The experiment was repeated with thorium oxide present in the vessel. The oxide was enc10sed in a paper cylinder, which allowed the emanation to pass through it. After repeated expansions no c10ud was formed, showing that for the expansions used the particles of the emanation were too small to become centres of condensation of the water-vapour. We may therefore conclude, from this experiment, that the emanation does not consist of dust partic1es of thorium oxide. It would be of interest to examine the behaviour of the emanation for greater and more sudden expansions, after the manner employed by C. T. R. Wilsont in his experiments on the action of ions as centres of condensation. The emanation may possibly be a vapour of thorium. There is reason to believe that a11 metals and substances give off vapour to some degree. If the radioactive power of thorium is possessed by the moleeules of the substance, it would be expected that the vapour of the substance would be itself radioactive for a short time, but the radioactive power would diminish in consequence ofthe rapid radiation of energy. Some information on this point could probably be obtained by observation of the rate of diffusion of the emanation into gases. It is hoped that experimental data of this kind will lead to an approximate determination of the molecular weight of the emanation. • Trans. Ro}. Soc .• 1897. t Phi!. Trans. Ro}. Soc.• clxxxix, 1897.

.4 Radioadil'(, Suhstan('r rmittrdFml1 Thorium ('on1pmmd"i

231

Experiments have been tried to see if the amount of the emanation from thorium oxide is sufficient to appreciably alter the pressure of the gas in an exhausted tube. The oxide was placed in a bulb connected with a Plücker spectroscopic tube. The whole was exhausted, and the pressure noted by a McLeod gauge. The bulb of thorium oxide was disconnected from the main tube by means of a stopcock. The Plücker tube was refilled and exhausted again to the same press ure. On connecting the two tubes together again, no appreciable difference in the pressure or in the appearance of the discharge from an induction coil was observed. The spectrum of the gas was unchanged. Experiments, which are still in progress, show that the emanation possesses a very remarkable property. I have found that the positive ion produced in a gas by the emanation possesses the power of producing radioactivity in all substances on wh ich it falls. This power of giving forth aradiation lasts for several days. The radiation is of a more penetrating character than that given out by thorium or uranium. The emanation from thorium compounds thus has properties which the thorium itself does not possess. A more complete account of the results obtained is reserved for a later communication. McGill University, Montreal September 13, 1899

Radioactivity Produced in Substances by the Action of Thorium Compounds by E. R UTHERFORD, M.A., B.SC.

Macdonald Professor of Physics, McGill University, Montreal From the Philosophical Magazine for February 1900, sero 5, xlix, pp. 161-192 Communicated by Professor J. J. Thomson, F.R.S.

THORIUM compounds under certain conditions possess the property of producing temporary radioactivity in all solid substances in their neighbourhood. The substance made radioactive behaves, with regard to its photographic and electrical actions, as if it were covered with a layer of radioactive substance like uranium or thorium. Unlike the radiations from thorium and uranium, which are given out uniformly for long periods of time, the intensity of the excited radiation is not constant, but gradually diminishes. The intensity falls to halfits value about eleven hours after the removal ofthe substance from the neighbourhood of the thorium. The radiation given out is more penetrating in character than the similar radiations emitted by uranium and thorium and the radioactive derivatives from pitchblende, radium, * and polonium. t Attention was first drawn to this phenomenon of what may be termed 'excited radioactivity' by the apparent failure of good insulators, like ebonite and paraffin, to continue to insulate in the presence of thorium compounds. The apparatus first used is shown in Fig. 1. Two insulated plates, Band C, were placed parallel to one another. In a shallow square depression LM in the plate C, a layer of thorium oxide was placed and covered with severallayers of foolscap paper. The whole was enc10sed in a lead vessel A, with a door in the side to allow the plate C to be readily moved. The crossed lines show the position of insulators. The plate C was connected to the + pole of a battery of 50 volts, the other terminal of which was to earth. The plate B was connected to one pair of quadrants of a delicate Thomson electrometer with a replenisher and gauge, the other pair of quadrants of which was connected to earth. With the arrangement in the figure, when B is insulated, there can be no conduction current from C along or through the insulators, since the earthconnected vessel intervenes. If the thorium-covered plate C was removed, and abrass one of the same dimensions substituted, there was no appreciable movement of the electrometer needle. If, however, the plate C, covered with * Curie, Comptes Rendus, 1898, p. 175. t Curie, ibid., December 26, 1898.

Radioaeti\'it)' Produced b)' Action of Thorium Compounds

233

thorium oxide, were left in the vessel for several hours with the plate B charged -, on removal of C and the substitution of a non-active metal plate, the movement of the electrometer needle showed that B was receiving a charge. On reversing the battery, the current was reversed in direction but equal in amount. The current between the plates gradually decreased with the time, and became inappreciable after a few days. By replacing the thorium oxide, the experiment could be repeated. It was at first thought that possible dust particles from the thorium oxide might have escaped from under the paper and in some way adhered to the upper plate. An examination of the plate B, however, revealed no trace of

+

M

A

C

L

EARTH M

EARTH ;ARTN

Fig. 1 thorium oxide on its surface. The plate made the air a conductor in its neighbourhood, as if it were covered with a thick layer of radioactive substance. If the surface of the plate was carefully scrubbed with sand- or emery-paper, the radioactive power was to a great extent destroyed. It was found possible to make the plate B active, even if the thorium oxide were covered with thirty layers of foolscap paper tightly waxed down so as to prevent the escape of dust particles. If the plate C was charged - and B + the plate B no longer became radioactive, but the top layer of paper over the thorium was found to be active on its upper side to about the same extent as the plate Bin the previous case; i.e. the negatively charged surface was made active in both cases. All the compounds of thorium examined have the power of causing radioactivity in substances. The oxide, however, gives far the largest effects, and has consequently been used in most of the experiments. The thorium compounds used were supplied by Messrs Eimer and Amend, New Y ork. The oxide was obtained by igniting the nitrate which had been manufactured from monazite sand. If the oxide is heated for some hours to a white heat in a platinum crucible, it loses its power of exciting radioactivity in substances to a very large extent. H*

234

The Collected Papers 0/ Lord Ruther/ord Comparison

0/ Intensities 0/ Radiation

The intensity of the radiation, excited in substances in the manner described, was in a1l cases compared by the electrical method. In general, for the purposes of measurement, the radioactivity was excited in flat plates or circular cylinders. For flat plates the testing apparatus was similar to Fig. 1. The brass plates corresponding to B and C were 5 cm. apart, with a potential difference of 50 volts between them. The current between the plates, measured by the rate of movement of the electrometer needle, was taken as proportional to the intensity of the radiation at the surface. With radioactive cylinders, the active cylinder was placed in a larger cylinder and concentric with it. The current for 50 volts between the cylinders was taken as a measure of the intensity of the radiation at the surface. For experiments, extending in some cases over several days or weeks, it was necessary that for each observation the electrometer should be of the same degree of sensitiveness. This was roughly ensured by the Thomson replenisher and gauge, attached to the electrometer. For small variations from the standard sensitiveness, the values of the current were corrected by observing the number of divisions on the electrometer scale corresponding to the E.M.F. of a Clark cello As in the course of this paper it will be necessary to compare the intensity of the radiation from radioactive plates and cylinders, abrief theoretical discussion will be given of the relation that exists between the intensity of the radiation, the area of the active surface, and the maximum current through the gas. Two cases will be considered: (1) When the radiation is given out uniformly from a plane surface and the current through the gas is measured between two parallel planes. (2) When the radiation is given out from a cylinder and the current measured between concentric cylinders.

Case I. We will first consider the case of a uniformly radioactive plate C of area S, which is placed between two large parallel plates A and B (Fig. 2a). We will suppose the plate C to be of large dimensions compared with the distance d of the plate C from A, and to give out radiation equally from all points of its surface. The gas is ionized by the passage of the radiation through it, and the ions produced travel to the plates A and C under the influence of the electric field. In consequence of the energy required to ionize the gas, the intensity of the radiation diminishes in its passage through it. Suppose the radiation is homogeneous in character, and that Ais the coefficient of absorption of the radiation by the gas. Let lobe the intensity of the radiation at the surface of the plate. Since the plate is large compared with the distance d, the value of the intensity may be considered approximately equal

Radio(l('fjrit)' Produced hy Action ofTllOrill1ll CompOll17ds

235

at equal distances from the surface C. In consequence of the absorption of the radiation by the gas, the intensity 1 at a distance x from the active plate is given by 1 = loe- Ax• Let dn be the number of ions produced per second between two planes dx from it. parallel to C and distant x and x

+

8

E

an

rI Cl

3

J)

b

l'

8

an Fig.2b

Fig.2a

Since the rate of production of ions is proportional to the intensity of the radiation, the total number of ions n produced per second between A and C, distant d apart, is given by

n= =

f:

KSI",r"'dx, where K is a constant,

Kflo(I -

rAtI).

If Eis the charge on an ion, the current i through the gas, when an E.M.F. is applied sufficient to remove all the ions before recombination takes place, is given by i= nEo iA Therefore Slo = KE(l _ r Ad) ;

or the product of the intensity of the radiation and the area of the active surface is proportional to the current through the gas. It is of interest to develop the above equation from considerations of the energy required to produce an ion. Let W be the average amount of energy used up in producing an ion in the gas. We will assume that the absorption of the energy of the radiation in its passage through the gas is due solely to the production of ions. On account of the absorption, the intensity of the radiation varies from 10 at the surface of the active plate to lorAd at the surface of the top plate. 1f n be the total number of ions produced, we thus obtain

n. W =- Slo( I -- e-Ad),

236

The Collected Papers

0/ Lord Ruther/ord

where the energy absorbed over an area S is given by the right-hand side of the equation; or SI; _ Wi o - e(l _ e-Ad) , where current i = ne, as before. Some experiments given in previous papers* point to the conc1usion that the energy required to produce an ion may possibly be the same for all gases at all pressures, and it has been shown by Professor J. J. Thomson and Mr Townsend that the charge of the ionst in different gases is the same. If such is the case, W/ e is a constant for all gases and the current through the gas will depend only on A, d, and S/o• Case II. We will now consider the case of a radioactive cylinder, where the current is measured between two concentric cylinders. Let Fig. 2b represent a cross-section of the cylinders. Let a = radius of radioactive cylinder D, b = radius of concentric cylinder E. Suppose length of cylinder D to be large compared with the distance between the cylinders. If A is the coefficient of absorption ofthe radiation, the intensity 1 at a distance r (outside D) from the centre is easily seen to be _1 = a_ e-')..(r-a)

r

10

'

where 10 = intensity ofradiation at the surface, since without any absorption the value of 1 would fall off inversely as the distance. The total energy of the radiation near the surface of the external cylinder is given per unit length by a

10 b e-')..(a-b) • 27Th, the energy per unit length elose to the surface of the active cylinder by

10. 27Ta.

The total energy absorbed in the gas is thus equal to 10.27Ta{1 - e-')..(b-a)}.

If n = the number of ions produced per second due to the length I of the active rod, W.n = 10.27Tal{1 - e-')..(b-a)}

= 10.S{1 -

e-')..(b-a)},

where S is surface-area of active cylinder; or

W.i S/o = E{I ~. e~')..(b-a)}' where i = nE, =

Ai

-, where A

W

= -e = constant.

• Rutherford, Phil. Mag., January 1899. t J. J. Thomson, Phi!. Mag., December 1898; J. S. Townsend, Trans. Roy. Soc., 1899.

Radioactivity Produced by Action 01 Thorium C01upounds

237

In both of the cases considered, half the radiation has been absorbed in the substance which is made radioactive, and the other half passes through the gas, since the radiation is given out from the surface in all directions. In the case of complete absorption of the radiation in the passage through the gas, the maximum current i is given by

SIo = Ai. An investigation is now in progress to determine the value of A, that is,

W!f:. If Ais determined, the intensity ofthe radiation can at once be expressed

in absolute measure.

Conditions for the Production o[ Radioactivity in Substances

In order to confine the induced radioactivity produced by thorium compounds to any particular conductor, it is necessary that it should be charged - and all other bodies in the field +. In order to produce radioactivity in all bodies in the neighbourhood, no electric fie1d is required. If thorium oxide is placed in a c10sed vessel connected to earth, the sides of the vessel and any solid bodies near, whether conductors or insulators, become radioactive. If, in addition, the surface of the thorium oxide is covered with paper or thin aluminium foil, the side of the paper away from the oxide becomes radioactive. When no electromotive forces are acting, the amount of radioactivity in a given time per unit area is greater the nearer the body to the thorium oxide. With electromotive forces acting, the substance to which the radioactivity is due appears to travel along the lines of force from the + to the - charged body. It is thus possible to concentrate the radioactivity on small plates or fine wires by placing them in a c10sed metal vessel connected to earth and charging them -. If the bodies are all uncharged, the partic1es producing radioactivity, by the process of diffusion through the gas, are carried to the sides of the bodies and adhere to them. A fine wire fixed in the centre of avessei on the bottom of which the active salt is placed becomes only slightly radioactive, since onlya few of the active partic1es reach its surface. The c10ser a body is to the thorium, other conditions remaining unaltered, the more active it becomes. Fig. 3 shows the general arrangement for concentrating the activity on a sm all area of a conductor. A metal vessel V was connected to the + pole of a battery of small lead accumulators of 300 volts, the other pole of which was to earth. A thick layer of thorium oxide was placed in the bottom of the vessel and covered with several thicknesses ofpaper. A brass tube D was fixed in the side of the vessel and metallically connected with it. A fine platinum wire AB was fixed on the end of a stouter brass rod BC. The brass rod was fixed centrally in the cylinder D and insulated from it. The end of the brass rod B was placed weIl inside the cylinder D. The conductor AC was connected to earth. The fine wire is thus the only body exposed in the fie1d with acharge, and,

238

The Collected Papers of Lord Rutherford

under the influence of electric forces, the active particles are carried to the wire AB and adhere to its surface. The same general resuIts are obtained whether the surface of the thorium oxide is bare or covered with paper or thin layers of metal foil. Two or three layers of paper almost completely cut off the ordinary radiation* from thorium; so the effect cannot be due to the direct radiation from its surface. In this way I ha ve been able to cause a piece of platinum wire of length 1 cm. and diameter 0·018 cm., i.e. with a surface area of 0·056 cm., to give more than 20 times the rate of discharge given by a thick layer of uranium oxide of 25 sq. cm. area. A rate of movement of an electrometer needle of 200 divisions in 5 sec. is quite easily obtained from the action of such a small active V

c

D B

A

THOR OXIDE

Fig. 3 surface. (One volt gave a defiection of 40 divisions on the electrometer scale, and the capacity of the whole circuit was about 50 electrostatic units.) I have spoken of using a platinum wire, but any other metal wire will serve equally weIl. Using large electromotive forces and a large surface of thorium oxide, it would be quite possible to increase the radioactivity of unit area of the conductor to more than 20 times the value cited in the above case. So far as the results obtained indicate, there is no limit to the amount of increase, since we can suppose the area of the - charged conductor diminished and the amount of thorium increased. In practice, however, a limit would soon be reached, as it would be difficult to cause aU thc radioactive partic1es to move to the small conductor without very largc elcctric forccs.

Connection between the 'Emanation' from Thorium and 'Excited' Radioactivity In a previous papert I have shown that compounds of thorium emit some kind of radioactive material or 'emanation', which is able to pass through t Phi!. Mag., January 1900. * E. Rutherford, Phi!. Mag., January 1900.

RadioaNh'ity Pl'Oducecl hy Action

0/ Thorium ('ompowulli

239

eonsiderable thieknesses of paper and thin layers of metal, and preserves its radiating power for several minutes. These partic1es diffuse through the gas and become centres of ionization throughout the volume of the gas. The eurrent passing between two charged plates, on one of which is spread thorium oxide, is greatly diminished by directing a slow eontinuous blast of air between the plates. As the partic1es have no charge, they may be readily removed from between the plates by a eurrent of air even in a strong electric field. There is a very c10se connection between this 'emanation' and excited radioaetivity-in fact, the emanation is in some way the direct cause of the latter. The following facts will serve to show the c10se eonnection that exists: (1) All thorium compounds examined are able to make substances radioactive, but to different degrees. The greater the amount of emanation, the greater the amount of induced radioactivity. As an example, thorium oxide is the most active of all thorium compounds in produeing radioactivity and giving out the emanation. A thin Iayer of thorium oxide gives out very little emanation, and is only slightly effective in producing radioaetivity. (2) Substances are made radioactive when the active compound is covered with several Iayers of paper or thin metal foil. The emanation also readily passes through paper and thin metal foil. Two or three layers of ordinary foolseap paper completely cut off the ordinary radiation given out by thorium compounds, but do not much diminish the amount of induced radioactivity. (3) A slow current of air, which quickly removes the emanation as it appears, also diminishes the power of producing radioactivity. The amount of induced radiation is greater in c10sed than in open vessels, on account of the disturbance of air currents in the latter case. (4) Thorium oxide which had been heated to a sufficiently high temperature gave out very little emanation and produced little radioaetivity. Speaking generally , it may be said that the presence of the emanation is necessary for the production of radioactivity in substances, and that the amount of radioactivity depends upon the amount of the 'emanation'. A radioactive substance like uranium, which gives out no emanation, produces no trace of excited radioactivity. An experiment now to be described throws a further light on the question. The general arrangement of the experiment is shown in Fig. 4. A slow current of air from a gas-bag, after bubbling through sulphurie acid, passed down through a reetangular wooden vessel, 60 em. in length. In order to remove spray and dust and to equalize the current of air over the crosssection, the air was passed through cotton-wool at W. A metal plate covered the bottom of the vessel and was charged +. Four insulated meta! plates, A, B, C, D, placed at equal distances, were attached to a top metal plate connected to earth. Thorium oxide covered with paper was placed under the electrode A. The current of air was passed through the vessel at the steady rate of about

240

The Collected Papers 01 Lord Rutlle/ford

0·2 cm. per sec. for aperiod of 7 hours, with 300 volts between the lower and upper plates. The following results were obtained for the current due to the emanation which reached A, H, C, D and the corresponding radioactivity produced: Relative current due to emanation

Plate Plate Plate Plate

Relative 'excited' radioactivity

1

A

1

0·55

H

0·43 0·16

0·18

C D

0·061

0·072

The current due to the emanation which reaches A, and the radioactivity produced in A, is in each case taken as unity for the purpose of comparison.

EARTH

w B

A

c

D

THOItIUM OXIDE

Fig.4 It will be observed that radioactivity is produced on the plates some distance away from the thorium oxide, and is roughly proportional to the emanation current at the plate. We may conc1ude from this experiment that the radioactivity is, in some way, due to the 'emanation', or to something that accompanies it, but is not caused by the direct action of a radiation from thorium oxide.

Absorption

0/ the

Radiation by Substances

All radioactive substances, as weIl as bodies made radioactive in the manner described, ionize the gas in their neighbourhood and act upon a photographic plate in the dark. A simple method of testing whether two types of radiation are the same, is to determine the absorption of the radiation by layers of thin metal foil. If the absorption is different for the two types of radiation, we may consider them distinct kinds of radiation. The current between two parallel plates, one surface of which was radioactive, was determined when successive layers of a substance of equal thickness were placed over the radioactive plate. The following table is an example of

Radioaclivity Produced by Action ofThorium Compounds

241

the way the current (which is proportional to the intensity of the radiation) diminishes with successive layers of aluminium foil over a plate of zinc, which had been made radioactive: Zinc plate = 12 x 18 cm. Thickness of foil = 0·0004 cm. 95 volts between plates. No. of layen alum. foH

Current for thin layer of thorium oxide

Current for radiation from zinc

o 1 2 3 4

0·71

1 0·57

0·43 0·32

0·13

6

0·155

1

0·36 0·23

0·55

5

0·084 0·056

The third column of the table gives the variation of the current with thickness of foil for a thin layer of thorium oxide, and serves as a basis of comparison with the excited radiation. The current for the bare radioactive surface is in each case taken as unity for the purpose of comparison. Fig. 5, curves A, B, show these results graphically when the ordinates denote current and the abscissae thickness of aluminium. It will be observed that the radiations from zinc and thorium oxide are quite different in character, the radiation from the former being far more 100

90

80 70

60 60

30

CUIlRENT

60

20 10 ~AYERS AI.UMINIU"" POIL

4-

2

3

Fig. 5

4-

5

6

242

The Collected Papers of Lord Rutherford

penetrating as regards aluminium. Both types of radiation are approximately homogeneous. The current, which is proportional to the intensity of the radiation, diminishes approximately in a geometrical progression as the thickness of the metal increases in arithmetical progression. The same general difference is shown for the two types of radiation by testing their comparative absorption by thin layers of paper, gold-Ieaf, silver foil, and Dutch metal. The following table is an example of the absorption of the radiation from a zinc plate and a thin layer of thorium oxide for thin tissue paper: Thickness of layer of paper = 0·0030 cm. Potential difference between plates = 50 volts. No. of layers of paper

Radiation from zinc

1 2

1 0·57 0·35

o

3 4

0·20 0·12

Current

Radiation from thorium oxide

1 0·37 0·16 0·080

0·055

This method can also be used to compare the radiations from the various metals when made radioactive. In this way it was found that all the substances tried, viz. Cu, Pb, Pt, Al, Zn, brass, cardboard, paper, which had been made radioactive, gave out radiations of the same penetrating power. It was also found that the same type of radiation was given out from polished and dull surfaces, and that it was unaffected by the concentration of the radioactivity. Since the same radiation is given out by all the metals and non-metallic substances likecardboard or paper, under varying conditions, we may conc1ude that either the substance itself which has been made radioactive plays no direct part in determining the kind of radiation, or that all exert exacdy the same action. The 'excited' radiation is also of a more penetrating character than that given out by uranium, thorium, and the pitchblende derivatives radium and polonium. Absorption of the Radiation in Air

The absorption of the induced radiation in air was also determined. The method employed was similar to one previously used and described by the author* for determining the absorption of uranium radiation by different gases. A similar apparatus has been employed by Owenst for thorium radiation. Two insulated parallel plates, kept a fixed distance apart, could be moved

* Phi!. Mag., January 1899, p. 124.

t lbid., October 1899, p. 378.

Radioactivity Produced by Action

0/ Thorium

243

Compounds

by means of a screw to different distances from the parallel radioactive surface. The radiation from the active surface passed through a circular opening in the lower plate, covered with thin aluminium foH, and was stopped by the upper plate. The current between the two fixed plates for a large voltage was determined for different distances from the radioactive plate. If the radius of the active surface is large compared with the distance of the lower of the pair of plates from it, the current between the plates for a distance x of the lower plate from the active surface varies as e-AX, where Ais the coefficient of absorption of the radiation in the gas. The following table gives the results obtained for the radiation from a lead surface which had been made strongly radioactive: LEAD RADIATION

Distance from surface

d(=3 mm.) d + 6·25mm. d + 12·5mm. d+ 18·7mm. d+25 mm. d+31·2mm. d+ 37·5mm.

Current

1 0·79 0'59 0·46 0·35 0·27 0·21

The current is taken as unity when the measurements began at a distance d = 3 mm. from the active lead plate. For the purposes of comparison, the numbers obtained in a similar manner for thin layers of thorium oxide and uranium oxide on a bare plate are given below. THORIUM RADIATION

Distance

d( = 2·25 mm.) d 5mrn. d +- 10 rnrn.

+

d -+- 15 rnrn. d -I- 20mm.

Current

1 0·73 0·50 0·35 0·25

URANIUM RADIATION

Distance

d(= 2·25 rnrn.) d + 2·5rnrn. d I 5 rnrn. d -I- 7·5 rnrn. d + 10 rnrn. d + 15rnm. d + 20mm.

Current

1

0·685 0·445 0·296 0·188 0·088 0·059

The curves in Fig. 6 show the results graphically. It will be seen that the intensity of the radiation falls off approximately in a geometrical progression as the distance increases in arithmetical progression. Curves of absorption of thorium radiation in air at different pressures have been obtained by Owens. *

* Owens,

Phi!. Mag., October 1899.

244

The Collected Papers of Lord Rutherford

The distances through which the three types of radiation from uranium, thorium, and active lead pass through air at ordinary pressures and temperatures before the intensity is reduced to one-halfits value, are about 4, 10, and 16·5 mm. respectively. 100

90 80

70 60

50

~

40

!\,)

~

30

zo 10

I)I$TANtE IN eH. r, MMPLAii'

'5

J·O

'·5

2·0

2·5

3·0

Fig.6 Assuming that the intensity falls off as e-AX, the values of'\ for the types of radiation are given below. Value of A

'Excited' radiation Uranium radiation Thorium radiation

0·42 1·6

0·69

The order of absorption in air of the above three types of radiation is the same as for aluminium and paper. The 'excited' radiation is of a more penetrating kind than the easily absorbed type (the oe radiation)* given out by uranium, but much less than the ß type. The radiations from radium and polonium are also more readily absorbed in air than the excited radiation iso Duration of the Radioactivity

If a plate or wire which has been made radioactive is removed from the action of the thorium, the intensity of the radiation diminishes according to a very simplelaw. '" Rutherford, Phil. Mag., January 1899, p. 116.

Radioactivity Produced by Action

0/ Thorium

Compounds

245

A large number of experiments have been made on the duration of the induced radioactivity in various substances under varying conditions. A typical table of the results obtained is given below for a rod of brass which has been made active. In order to test the rate of deeay of the intensity, the aetive rod was plaeed inside a eylinder and eoneentrie with it. The eurrent between the two eylinders for a potential difference of 50 volts was measured in the usual manner, and at invervals of several hours. Length of rod = 31 . 5 em. Diameter = 0·40 em. Testing eylinder, inside diameter

= 7·3

Time in hours

em. Current

o

1 0·640 0·474 0·196 0·138 0·103 0·0370 0·0186 0·0086

7·9 11·8 23·4 29·2 32·6 49·2 62·1 71·4

The value ofthe maximum eurrent, whieh is taken as unity, was 1·6 X 10-11 ampere. Fig. 7 shows graphieally the results obtained. The results show that the eurrent through the gas (whieh is proportional to the intensity ofthe radiation) 100 90

80 70

50 40 30

CUIlRENT

60 11

20

'0

TIME 111 HIJIJR$ 30 20

10

40

Fig.7

50

60

70

246

Tlte Collected Papers 0/ Lord Rutherford

diminishes in geometrical progression with the time. The time taken for the intensity of the radiation to fall to half its value is about eleven hours. If 10 be the intensity at the beginning, the intensity 1 after a time t is given by

1 = loe-Lt, where L is a constant. The above law appears to hold accurately for all substances made radioactive. No difference in the rate of decay has been observed, whether the radiation is on a plate oflarge area or concentrated on a fine wire. The rate of decay is also independent of the substance made radioactive. A piece of paper, mica, or metal, all give the same rate ofloss ofintensity. As far as experiments have gone, the rate of decay is unaffected by the pressure of the gas surrounding it, or whether the air is dry or full of moisture. The same rate of decay has always been obtained under all the conditions tried, provided the surface is not acted on mechanically or by chemica1s. The mean value of L deduced from the above results is L

= 0·0000189.

In a previous paper· I have shown that the radioactive 'emanation' from thorium compounds quickly loses its radioactive power. The intensity in that case falls to half its value in about one minute, while the intensity of the 'excited' radiation falls to half its value in about eleven hours, or one decays 660 times faster than the other. The law offalling off of intensity is the same in the two cases. On page 236 it has been shown that the current i (for a 'saturating' E.M.F.) between two cylinders is given by

i=

SIo

A

L_

{I - e-AC.u-u)},

with the same notation as before. The intensity 1 of the radiation after a time t is given by I=/~,

and the total quantity of electricity passing between the cylinders during the time taken for the intensity to fall to zero is given by

Q=

SI. {I I cooidt = -.!! A

e-)'(b-O)}

= SIo {I _ e-AC.b-a)}· LA ' if io = initial current, it is clear that

Q =~. L

*

Phil. Mag., January 1900.

Icoe-Ltdt 0

Radioacth'ity Produced by Action

0/ Thorium

Compounds

247

In the case given in the last table, the initial eurrent was 1·6 X 10- 11 ampere, and the value of L = 0·0000189; therefore the total quantity of eleetricity passing between the cylinders is equal to 8·5 X 10-7 coulomb. The total quant i ty of electricity separated, if the radiation has been completely absorbed in the gas, is obviously 1-

1 e-),(b-a)

of this quantity. In the above ease, a = 0·20 em., b = 3·65 em., " = 0·42. Therefore quantity passing between cylinders = 11·1 X 10-7 coulomb. lncrease

0/ lnduced

Radioactivity with Time

If a plate or wire is exposed to the action of thorium oxide in a c10sed vessel, the radioaetivity at first increases nearly proportionally with the time, and then more slowly, finally tending to a maximum value after several days' exposure. The table given below is an example of the results obtained for a square zinc plate, area 86 sq. em., exposed in a metal vessel, with a potential differenee of 300 volts between thorium and surface to be made active. The plate was removed from the action of the thorium at intervals for sufficient time to determine the eurrent produeed by it between two eharged parallel plates, as in Fig. 1. Time of exposure in hours

1·58 3·25 5·83 9·83 14·00 23·41 29·83 47·00 72·50 96·00

Current

0·063 0·105 0·289 0·398 0·586 0·773 0·834 0·898 0·951 1·00

The eurrent after four days' exposure is taken as unity, as the rate of leak had nearly reached its maximum value. The maximum value of the eurrent produeed by the active plate between two test plates 4 em. apart was 1 ·7 X 10-11 ampere. Fig. 8 shows the results graphieally. From the table it will be seen that the intensity has reaehed half its final value in about twelve ho urs. We will now consider the conditions which influenee the increase of the intensity of radiation from a given surface exposed to the action of a thorium

248

The Collected Papers

0/ Lord Ruther/ord

compound. We will suppose that the surface to be made radioactive is negatively charged. Two opposing actions are evidently at work. Fresh radioactive particles are being continually carried to the plate, while the intensity of the radiation given out by the active surface continually diminishes, owing to the radiation of energy. A steady state will be reached when the rate of increase of intensity due to the supply of fresh radioactive particles is equal to the rate of decrease of the intensity due to the radiation of energy from the active surface. 100

90

80

~

70 ~

60

!t.;i

50

40 30 I

20

'0 '0

TIME IN HOURS i 20 30 +0 50 60 70 80 90 100

Fig.8 Let I be the intensity of the radiation at the surface of the plate at any time. The rate of diminution of the intensity is equal to LI, since the intensity I at any time is given by

1= loe-LI and

dl dt

=

-LI.

Let q be the rate of increase of the intensity due to the steady supply of radioactive material. Then dI dt =q -LI or

log.,(q - LI) But 1= 0 when t = O.

=

-Lt

+ A.

Rodioocth'ity Pmduced by Artiol1 Therefore A=

0/ Thorium Compoll1uls

249

1

-I lo~q.

Therefore log., or

q-LT = -LI q

1=

t

(1 - e-L ,).

When LI is very large, the maximum value of the intensity 10 is given by q

10=1

and

1

To = 1 - cL';

or the equation representing the rise of intensity of the radiation is the same as the rise of an electric current in a circuit of constant self-induction. The curve whieh is shown in Fig. 8 is in rough agreement with this equation. For example, the intensity of the radiation has risen to half its value in about twelve hours. Now e- Lt = I when I = 11 hours, i.e. according to theory, the eurrent should have reaehed half its value in about eleven hours. There is a divergence between the theoretical and observed results in the first part of the curve. The rate of inerease of intensity is slower at first than the theory would suggest. It is probable, however, that the rate of supply of radioaetive material does not reach a steady value for a considerable time after the exposure ofthe plate, and such a cause would account for the results observed. ather results obtained, under different conditions, all show too small a value of the intensity for the first few hours of exposure. We have so rar assumed that the radioaetive particles were eonveyed to the surfaee under the infiuence of an eleetrie field. The equations whieh have been given will, however, apply equaIly weIl to the ease of diffusion. If no eleetromotive forces are aeting, the radioaetive partic1es diffuse through the gas and adhere to the surfaee on whieh they impinge. A steady state will be reached when the rate of supply of fresh radioaetive particles due to diffusion is balanced by the deeay of the radiation from the surfaee. The maximum intensity of the radiation on any surfaee in the neighbourhood of a thorium eompound is thus proportional to the number of radioaetive partic1es that reaeh it by the processes of diffusion.

E./JeCI

0/ E.M.F. on Ihe Amounl 0/ Radioactivity

The amount of induced radioaetivity in a given time increases with the voltage for small voltages, but so on reaches a point beyond whieh large increases in the E.M.F. have a very small effeet on it. In order to investigate the relation in detail, the following arrangement was employed:

250

The Collected Papers o[ Lord Ruther[ord

Two insulated concentric brass cylinders A and B (Fig. 9) were used, of diameters 5·5 and 0·7 cm. respectively. The ends were c10sed with paraffin stoppers C and D. The cylinder A was connected with the + pole of a battery, the other pole of which was to earth. The cylinder B was connected to the electrometer in the usual manner. A layer of thorium oxide in a paper envelope was placed along the bottom of the cylinder A. The wh oIe was exposed to the action of thorium oxide for three days. The intensity of the radiation given out by B had, after that interval, nearly reached its maximum value.

o

A+

o

BTHDRIUM OXIDE

\larth Fig.9

The following measuremems were made for each experiment at different voltages: (1) The current between A and B was measured with the thorium oxide in the cylinder. (2) After the thorium oxide had been removed and the air blown out, the current between A and B was again determined. (3) The cylinder B was then removed and a non-active one of the same dimensions substituted and the current again observed. The electrometer was brought to the same sensitiveness every day by means of a Thomson replenisher. For rapid rates of discharge, a condenser of 0·001 microfarad capacity was placed in the electrometer circuit. The current (3) inc1udes the smalileak (if any) over the insulators plus the current due to the radioactivity produced in the end paraffin stoppers. The current (2) was due to the radioactive cylinder B, together with the current (3). The current (1) was due to the 'emanation' from thorium oxide plus (2) and (3).

Radio{/clivily Produccd hy Action ofThoriulI1 Compounds

251

From these three observations it was therefore possible to determine: (a) The rate of diseharge due to the thorium alone; (b) The rate of diseharge due to radioactive eylinder B alone;

Ce) The rate of discharge due to radioactivity on the sides and ends of vessel. In the following table the results are given for different voltages between cylinders. Results are in divisions per second of electrometer-scale. Voltage

Emanation

Radioactivity on cylinder B

Radioactivity on sides and ends

Total radioactivi ty

5·3 12·2 25·3 30·0 32·2

1·32 4·02 4·53 6·69 7·40 9·91

5·58 3·06 1·97 1·51 0·59 0·75

6·90 7·08 6·50 8·20 7·99 10·66

0 3 5 10 20 310

Fig. 10 shows the results graphically. Curve A shows the variation of the eurrent due to the 'emanation' with voltage; curve B, the variation of the amount of induced radioaetivity on inside cylinder; and eurve C, the variation of the amount of radioactivity on the sides and ends. The ordinates of curves Band C are inereased three times in order to show them on about the same scale as A. It will be observed that the shapes ofthe curves A and Bare similar. 100

tMANATIOW

A

l'l.OTItJ RAD.DN-'~OQt3·3

B

90 ...

80

~-

~

70 ~ (J

SO

50 40

30 ZO

10

EKtlt1"ED )rAD.DN SIDES1t3·f E.M.F./N "O~TS '

JO

20

30

Fig. 10

40

SO

C

252

The Collected Papers 0/ Lord Rutherford

The 'knee' of both curves occurs for about the same voltage. The curve C shows that as the voltage diminishes, the amount of radioactivity on the sides and ends increases, reaching a maximum when the voltage is zero. The currents due to the radioactivity of B, given in the third column of the table, are for 50 volts between the cylinders. The value given for 310 volts is probably too large, as it was measured for 310 volts between plates, instead of 50 as in the other experiments. In the fifth column is given the total current due to the radioactivity on ends and sides plus the action of cylinder B. It will be observed that the resulting values are not very different, except the value for 310 volts, which, for reasons above explained, is probably too large. It looks as if a certain number of radioactive partic1es were given out from the thorium and that these were carried to various parts of the vessel, the effeet due to the whole number being about the same as if they were all eoncentrated on the negative electrode.

Case 0/ Diffusion 0/ Radioactive Particles The case where no voltage is acting is one of special interest, for there the diffusion of the radioactive particles is alone operative. A loose layer of paper was placed over the paper envelope containing the thorium oxide. The paper envelope bent into the are of a eirc1e covered about one quarter of the eircumference of the cylinder. The following numbers give the rate of leak, in divisions per second, due to the radioactivity on different portions of the vessel: Radioactivity on inside eylinder Radioactivity on paper Radioactivity on outside eylinder and stoppers Total radioaetivity

1 ·32 divisions per second 2·26 divisions per second 3 . 32 divisions per seeond 6·90

The current due to the total radioactivity is thus about the same as the current when 20 volts aets between eylinders. The experiments on the effect of voltage extended for more than a month, and some of the results showed that the thorium oxide was not a constant source of radiation during the whole of that time. The variations were not, however, sufficiently large to obscure the general nature of the results.

Effect

0/ Pressure on Radioactivity

The diminution of the pressure of the gas from 760 to 20 mm. had very little action on the amount of 'excited' radioactivity on the - charged electrode. The following apparatus was employed: A brass cylinder B (Fig. 11 a), with an ebonite stopper C, through which passed a brass rod A, was connected with a mercury pump. The thorium oxide inside a paper envelope was placed inside the cylinder. B was connected

RlIdioacliI'ity Prot!uced hy Aclion of T!wriu!1l Compozmds

253

to + pole of a battery of 50 volts, and C to - pole. The apparatus was exhausted to the required pressure as rapidly as possible, and rod A exposed for several hours. The rod was then removed, and the current due to its radioactivity tested inside another cylinder D (Fig. Ilb). Thc battery and electro-

Topump B Topump

B

B

Fig. l1a Exciting Apparatus

o

Topump Fig. I1b Testing Apparatus meter connections are seen in the figure. On account of the press of other work, it was not found possible to take observations at regular intervals, but the table given below suffices to show the general nature of the results.

mm.

Timeof exposure hours

760 175 16 4·5 1·7 0·45 0·04

5·25 20 4·1 5·4 4·8 14·3 25

Press ure

Divisions persecond

2·37 9·83 2'15 1·96 0·72 0·38 0·34

Divisions per second. Interpolated values

13·1 13·9 15'1 10'3 4'5 0·65 0·44

The third column gives the current in divisions per second due to the radioactive rod in the testing vessel. In the fourth column are given the divisions per second corresponding to an exposure of the rod for the same time (three days) in each case, at the particular pressure. The results are interpolated from the first two columns with the help of the curve given in Fig. 8. The

254

The Collected Papers 0/ Lord Ruther/ord

results must only be considered approximate, and merely serve to give a comparative estimate of the radioactivity at each pressure. The general results are clear. The radioactivity is about the same at 16 mm. as at 760 mm. Between the pressures of 16 and 4·5 mm. the amount begins to diminish, until at 0·45 mm. it is only -lö of the value at atmospheric pressure. Still further diminution of pressure does not have much effect. A special experiment on the distribution of the radioactivity at low pressures throws some light on the phenomena. If we expose a rod charged at a low pressure to the action of thorium, it will be found that the rod is only slightly radioactive, while the top of the paper over the thorium oxide and the sides of the vessel are strongly radioactive. At atmospheric pressure, other conditions remaining the same, it will be found that most of the radioactivity is confined to the rod, and only a slight amount is produced on the paper and sides of the vessel. It appears as if the radioactive particles are unable to be all carried to the negative electrode at low pressures. This may be due to the increased rate of diffusion of the active particles at low pressures, or more probably to the small number of ions produced by the 'emanation' at low pressures. It is found that the current through the gas due to the 'emanation' falls off nearly proportionally to the pressure, so that the number of ions present between cylinders at low pressures is a very small fraction of those at atmospheric pressure. The following table gives results of the variation of the current, due to the emanation, with pressure of air in the apparatus of Fig. l1a. Pressure of gas mDl.

Current duc to emanation

760 587

1

0·819

402 214

0·582 0·297 0·203 0·133

145 93 25

0-046

Fig. 12 shows the results graphically. The curve is nearly a straight line. If the conveyance of the radioactive particles to the electrode is due to the movements of the ions between cylinders, at low pressures the number of ions may be too small to be effective in that respect.

Effect

0/ Gases

The apparatus shown in Fig. 11 was used. The amount of radioactivity produced on the central rod was not very different whether the gas was hydrogen, air, or carbonic acid. No definite difference was observed whether the gas was

Radioactil'ity Produced by Actioll

0/ Thorium

Compounds

255

free from water-vapour or not. The amount of current due to the emanation from thorium oxide was found, however, to vary greatly with the gas. Taking the current due to air as unity with 50 volts acting, the currents due to the 'emanation' were Air 1 0·35 H 1·1 CO2 These numbers are not necessarily proportional to the ionization constants of the gas, as the current produced depends on the relative absorption of the rays between the cylinders. 100 90

80 70

60

SO

~

40 I~

30

~

ZO 10

o

PRESS/) 'RE DF AIR IN NMS. 100 fOO 300 400 500 600 700 100 900 1000

Fig. 12 These results, together with those obtained for lowering of the pressure in air, show that there is no evident quantitative connection between the current due to the emanation and the amount of induced radioactivity. Chemical and Mechanical Actions on the Radioactive Surface We have previously considered the conditions which govern the production and decay of the induced radioactivity. We will now describe some experiments that have been made to try and throw some light on the question as to what the induced radioactivity is really due to. If the radioactivity is caused by some radioactive dust deposited on the substance, we should expect to find evidence of it by examining the surface with a microscope, or by noting whether there is any increase in weight. A fine

256

The Collected Papers

0/ Lord Ruther/ol'd

piece of platinum wire, which had been carefully weighed, was made strongly radioactive by five days' exposure to thorium oxide covered over with paper. Within the limits of accuracy of the balance no certain variation of the weight could be detected. The increase of weight, if any, was certainly less than -iö of a milligramme. On examination by a microscope no collection of dust particles on the surface could be observed. We may conclude from tbis experiment, that if the radioactivity is due to the deposition of radioactive particles on the surface, these partic1es must be extraordinarily radioactive compared with their weight. A rough estimate shows that the radioactivity of the surface layer must be at least a million times greater than that of uranium or thorium. The amount of radiation from an active surface is always lessened by mechanical actions, such as rubbing the surface with a cloth or fine sand-paper. In order to completely remove the radioactivity, it is necessary to remove the surface layer by long scouring with sand- or emery-paper. Ablast of air directed against a radioactive plate has no appreciable effect on the amount of radiation given out. A radioactive platinum wire or plate can be heated white hot without much altering the amount of radiation given out from it. A strongly active fine wire is more affected than a plate; but that is probably chiefly due to action of the flame-gases upon it. Chemical Actions

The radioactivity of a platinum plate is not much affected by dipping it in water, caustic soda, or nitric acid, whether hot or cold. Sulphuric or hydrochloric acid has the power of rapidly destroying the intensity of the radiation in a few minutes. A copper-sulphate solution, if only slightly acid, does not act on the wire rapidly. The following example of a test shows the effect of several of the solutions on a radioactive platinum plate. After each immersion the plate was washed in water and dried over a Bunsen flame. After exposure of 4 minutes to gradually heated water and 2 minutes to boiling water, the rate of discharge fell from 100 divisions in 15·5 sec. to 100 divisions in 20 sec. After 5 minutes' boiling in caustic soda, the rate of discharge fell to 100 in 27 sec. After 10 minutes' exposure to strong hot nitric acid, the rate of discharge was cut down to one-half its previous value. Dilute sulphuric acid reduced the rate of discharge to one-half in 10 sec and one-quarter in 60 sec. Both hydrochloric and sulphuric acids are more powerful in destroying the radioactive power than the other solutions examined. In the case of water and caustic soda, the small diminution of intensity appears to be due as much to the mechanical action ofthe bubbling as to the chemical action on the surface. The question now arises, whether the loss of radioactivity of the active plate by immersion in solutions is due to the destruction of the radioactive power of the partic1es or their removal from the plate to the solution. A fine platinum wire, very strongly active, was placed in a few drops of dilute sulphuric acid for several minutes. The wire lost a large proportion of its

Rucliouclirity Produced hy Actioll

(~l Thorium

Compounds

257

radioactivity. The dilute acid was then evaporated down to dryness in a sand bath, and on examination it was found that the residue on the glass surface was strongly active. We may conc1ude from this experiment that the radioactivity of the particles is not destroyed, but that they pass into solution, and that on evaporating the solvent the substance still remains. Some experiments were tried to see whether a plate preserved its radioactive power when a layer of copper was eJectrolytically deposited upon it. A radioactive platinum wire was made a cathode in a copper-sulphate solution, and a current of about half an ampere passed through for one minute. The rad ioactivity was diminished to about 0·7 of its value when tested in the usual way. After washing the wire in water, it was allowed to stand some time in air, and the rate of diminution of the radioactivity observed. The intensity diminished more rapidJy at first than for an unacted-on wire; but after 10 hours the rate of diminution became normal. The more rapid decrease at first is probably due to the dilute sulphuric acid which remained in the pores of the copper deposit. When the platinum wire was made the anode in a coppersulphate solution, the radioactivity rapidly diminished. The action in this case was probably due to the production of suJphuric acid at the surface of the anode by the passage of the current which dissolved the radioactive material on the platinum plate. Discussion

0/ the

Results

Before entering on the question of the cause and nature of induced radioactivity, abriefreview may be given ofthe results obtained: (I) All thorium compounds examined produce radioactivity in substances in their neighbourhood, if the bodies are all uncharged. With charged conductors the radioactivity is produced on the - charged body. In strong electric fields, the radioactivity can be concentrated on the surface of thin wires. Thorium oxide is the most active of the thorium compounds in causing radioactivity, but loses its power if it is heated for several hours at a high temperature. (2) The power of producing radioactivity is closely connected with the presence of the 'emanation' from thorium compounds, and is in some way dependent upon it. (3) The radiation excited in bodies is homogeneous, and of a more penetrating character than the radiations from thorium or uranium. The radiation is confined to the surface of the substance, and is independent of whether the substance is a conductor or non-conductor and of the nature of its surface. (4) The intensity of the radiation emitted falls off in a geometrical progression with the time, decreasing to half its value in about eleven hours. The decay of intensity is independent of the state of concentration of the radioactivity or the nature of the substance. I

258

Tlze Collected Papers

0/ Lord Rutlzer/ord

(5) The amount of induced radioactivity increases at first nearly proportional to the time of exposure, but soon tends to a value when the intensity ofthe radiation varies very little with increase of the time of exposure. (6) The amount of induced radioactivity produced in a given time on a conductor depends on the potential difference between the electrodes, and tends to a constant value for large E.M.F.s. (7) The amount of radioactivity is independent of the press ure of the gas, except at low pressures when the amount on the - charged conductor decreases with the pressure. The amount is not much affected whether the gas is hydrogen, air, or carbonic acid. (8) No increase of weight has been observed by making a body radioactive. The radiation from a platinum wire is not much altered by placing the wire in a flame, hot or cold water, or nitrie acid. Hydrochloric and sulphuric acids rapidly remove the radioaetivity from its surface. The solution, when evaporated, leaves the active portion behind. Three possible explanations of the phenomena of induced radioactivity naturally present themselves: (a) That the radioactivity is due to a kind of phosphorescence excited in the substance by the radiation from thorium; (b) or to the deposition of the + gaseous ions produced in the gas by the 'emanation'; (c) or to the deposition of partic1es of a radioactive material emitted by thorium compounds. Tbe hypo thesis that the radiation is a kind of phosphorescence will not explain the results observed, since substances are made radioactive outside the ineidence of the radiation, and the radioactivity can be concentrated on the electrode. The question as to whether the induced radioactivity is due to the deposition of a foreign substance on bodies, or to the action of the + ions produced in the gas, or a combination of both, is difficult to decide with certainty from the experimental evidence. The theory that the + ions produced by the emanation are responsible for the radioactivity, at first sight seems to explain many of the results. Since the radioactive particles of the emanation are very small, the intensity of the radiation must be very great near them; and, in consequence of this, ions may not only be produced, but the eharges on the ions set in violent vibration: these + ions would be carried to the negative electrode, and gradually dissipate the energy of their vibration by radiation into space. On this theory, however, it is difficult to explain the variation of radioactivity with pressure. At low pressures, the experiments show that the total radioactivity produced is much the same as at atmospheric pressure, but the - electrode receives only a small proportion of the radioactive particles. On the theory that the radioactive particles are + ions, we should expect them in a strong field to be all carried to the - electrode. Another experiment on the variation of the amount of radioactivity with distance also does not fall in readily with this view. The amount of radioactivity was found to be practically the same whether the distance from the radioactive surface was 3 mm. or 3 cm. In the latter case, the number of +

Radioarf il'ify Pmdured hy A ('( ion o( T/wrium ('o111"ound~

259

ions produced by the emanation is much greater than in the former; but the amount of radioactivity is unaffected. The theory that the radioactivity is due to adeposition of radioactive particles from the thorium compounds affords a general explanation of all the results; but the difficulty is to advance a satisfactory reason for the partic1es obtaining the + charge which they must possess in order to be moved to the - electrode in an electric field. If we suppose the radioactive particles from thorium compounds emitted at a uniform rate, independent of the nature and pressure ofthe gas. we should expect to obtain the same total amount of radioactivity spread over avesseI due to diffusion of the particles, as can be obtained by concentration of all the radioactive partic1es on the - electrode; and the amount should be independent of the pressure and nature of the gas, provided it does not act on the thorium. Some experiments seem to point to the conc1usion that the radioactive particles are not charged till they diffuse out into the gas, but that they gain a + charge in the course of time. A possible explanation is that the + charge is obtained by the diffusion of the ions to the surface of the particles. Since there is reason to believe that the - ions in most cases move faster than the + ions in an electric field, there is always an excess of + ions in the gas, and the particIes in the gas thus tend to become positively charged. On this supposition, the diminution of the amount of radioactivity on the - electrode at low pressures is due to the fact that there is not a sufficient number of ions in the gas to charge the partic1es, which thus diffuse to the sides of the vessel. As far as experiments have gone, the power of exciting radioactivity appears to be confined to thorium compounds. Neither uranium nor radium nor polonium has so far shown any trace of action; but the specimens* of radium and polonium used were not very radioactive and contained considerable amounts of impurity. A plate made radioactive is not able to excite any appreciable radioactivity in another plate near it. I have tested the + and electrodes after the passage for several hours of a strong current between them due to Röntgen rays, f1.ames, and discharge from points, but no trace of radioactivity on them has been observed. Macdonald Physics Building McGill University, Montreal November 22, 1899

* As this paper was passing through the press the Comptes Rendus of November 6 was received, which contains a paper by Curie and a note by Becquerel on the radiation excited in bodies by radium and polonium. Curie has used specimens of these substances 10,000 to 50,000 times more radioactive than uranium and the phenomena observed are, in some respects, similar to those exhibited by thorium compounds ; but there are not sufficient data on which to base any comparison. No mention is made of the effect of an electric field, or whether there is an 'emanation' from radium and polonium, as there is from thorium compounds. Curie concludes that the results obtained are duc to a kind of phosphorescence excited by the radiation; while in the case of thorium the author has shown that such a theory is inadmissible. Further experiments on the comparison of the radioactivity produced by thorium with that produced by radium and polonium will be of interest.

Energy of Röntgen and Becquerel Rays, and the Energy required to produce an Ion in Gases by E. RUTHERFORD, M.A., B.SC., Macdonald Professor of Phys;cs, and R. K. MCCLUNG, B.A., Demonstrator in Physics, McGill University, Montreal

From the Philosophical Transactions 0/ the Royal Society, 1901, sero A, vol. 196, pp. 25-59 Communicated by Professor J. J. Thomson, F.R.S. (Received June 15-Read June 21, 19(0)

THE primary object of the investigations described in this paper was the determination of the amount of energy required to produce a gaseous ion when Röntgen rays pass through agas, and to deduce from it the energy of the radiation emitted per second by uranium, thorium, and other radio-active substances. In order to determine the 'ionic energy' (as it will be termed for brevity), it has been necessary to make a special investigation to measure accurately the heating effect of X rays when the rays are absorbed in metals, and also the absorption of the rays in gases. The method employed to determine the ionic energy was briefly as follows:-The total energy of the rays emitted per second was determined by measuring the heating effect of a known proportion of the rays when absorbed in a metal. The total number of ions produced by complete absorption of the rays in the gas was deduced from measurements on the current produced by the ionization of a known volume of the gas and of the absorption of the rays in the gas, assuming the value of the ionic charge recently determined by J. J. THOMSON. On the assumption that aU the energy ofthe X rays is absorbed in producing ions in the gas, the total energy of the rays, divided by the total number of ions produced, is a measure of the energy required to produce an ion. In the course of the investigation the following subjects have been considered :(1) Measurement of the heating effect of X rays and the total energy of the rays emitted per second. (2) Efficiency of a fluorescent screen excited by X rays as a source of light. (3) Absorption of X rays in gases at different pressures. (4) Energy required to produce an ion in gases, with deductions on(a) Distance apart of the charges of ions in a moleeule. (b) Minimum potential required to produce a spark in the gases.

Dlergy

4 Riinlgell ami Becquerel Ra}'s

261

(5) Rate 01' emission of energy from the radio-aetive substanees, uranium, thorium, radium, and polonium.

Heating Effect of X Rays

Experiments on the heating effeet of Röntgen radiation have been made by DORN. * The rays were partly absorbed in metal foll plaeed in one bulb of a differential air-thermometer. The heat absorbed by the metal was eommunieated to the gas and the resulting change of volume observed. In order to obtain a measure of the heat supplied, the heating effeet due to a eurrent in a wire plaeed inside the bulb was observed. MOFFATt has dedueed the energy of X rays from photometrie eomparisons of a fluoreseent sereen with the Hefner amyllamp, assuming the effieieney of a fluoreseent sereen exeited by X rays as a souree of light. Knowing the value of the energy of the visible light of the Hefner standard, the heating effeet of the rays can be deduced. In determining the heating effeet of the rays, difficulties arise from whieh measurements of the heating effect of weak sources of visible light are free. I then first case, the inconstancy of an X ray bulb as a souree of radiation for measurements extending over long intervals is always a cause of trouble. [n the second plaee, the X rays are only slightly absorbed in thin metal foil, while light rays are completely absorbed at the surface of thin metal coated with lampblaek. Only a small portion of the energy of the rays is absorbed in passing through thin metal foil, and in consequenee a bolometer like LANGLEY'S, where the change of resistance of a very thin metal sheet, due to heat supplied by the rays, is observed, is not very suitable for measurements on the energy of X rays. Ordinary thermopiles are open to grave objeetions, as will be explained later in this paper. In order to measure the heating effect of the rays, a specially designed platinum bolometer was employed, and the heating effect was determined from the change of resistance of the platinum. Description of Bolometer (Fig. I)

A platinum strip, about 3 metres long, 0'5 centim. wide, and 0·003 centim. thiek, was wound on an open mi ca frame made as light as was compatible with rigidity. The frame was 10 eentims. square, and of a shape shown in fig. I (a). The platinum strip was wound round and round the frame, the strips on the front of the frame partly overlapping the corresponding ones at the back, but not touching them. The platinum strip (fifteen complete turns in all) was held in position by notches in the side of the mica frame, and the distanee between eaeh turn of the strip was 1 millim. Two of these grids were eonstructedas similar as possible, and mounted in the same vertieal plane on a wo oden base . ... ·Wied. Annal.', vol. 63, p. 150. t 'Roy. Soc. Edin. Proc.', 1898.

262

The Collected Papers

0/ Lord Ruther/ord

Resistance of each grid = 4·2 ohms. Area of platinum surface of grid = 92·2 sq. centims.

Fig. l(a)

Fig. l(b) The X rays incident on the grid for the most part passed through two thicknesses of platinum, but, on account of the windings not completely overlapping, the rays in some portions passed through one thickness only.

263

r:l/rl'f{l' n.f Rii1l1gm a1l(1 ßrel/l/r/"rl Ra}'s

This was c1early shown in an X-ray photograph of the grid, which is sketched m fig. 1 (h). where the shaded portions are the areas where the rays only passed through one thickness of the platinum. The absorption of the rays in the mica frame was very slight, and it was only on very careful inspection ofthe photograph that tbe outline ofthe frame could be observed. For the X rays employed, the intensity was cut down to 0·45 of its value after passing through the grid.

Focus Tube The rays were excited in an automatie focus tube of the pattern shown in fig. 2, with a platin um anode and an aluminium cathode. The tube was excited by a large coil, using a Wehnelt interrupter on a llO-volt circuit. The

A

uoVoU"s.

Interrupter. Fig.2 alternative spark gap A was always kept the same length-about 5 inches. The bulb was a very hard one, and there was generally a fairly rapid succession of sparks across A during tbe working of the bulb in order to keep the vacuum constant. The constancy of the length of the spark A is of great importance in these experiments, in order to obtain rays of the same degree of penetration. A diminution of the spark length lowers the vacuum of the gas in the tube and produces rays of lower penetrating power. After working the coil for lO seconds, the platinum plate became red-hot and remained fairly constant during the next 30 seconds. Tt was generally found advisable to run the bulb at intervals for a quarter of an hour before beginning measurements, in order to get it into a steady state for the emission of rays of constant intensity. Under these conditions experiments could be

264

The ('ol/eelet! Papers

0/ Lord Rutlle/:ford

made from day to day with a maximum variation of intensity of 30 per cent., and generally with much tess. The bulb employed gave out rays of great intensity and great penetrating power. A fluorescent screen was brightly lighted at a distance of 20 feet from the bulb. With a 'soft' tube and less intense rays, it would have been difficult to measure the heating effect with accuracy.

Wehnelt Interrupter The Wehnelt interrupter was of a simple pattern. One lead plate was placed inside a thick glass vessel (a Leclanche cell was used) with three holes about 1 millim. in diameter bored in one side. The glass vessel was placed inside an ebonite box with a lead electrode, and was filled with dilute sulphuric acid in the usual manner. The acid was kept cool by having a water circulation through a coil of lead pipe in the ebonite box. By suitably tapering the holes in the glass the interrupter was made to work steadily at a slow speed and gave strong discharges in the coil. The E.M.F. employed was 110 volts and the current was about 15 amperes. The average number of breaks per second was 57. In the course of more than six months' work the glass vessel was only replaced once, on account of the gradual increase of the diameter of the holes.

Arrangement 0/ Apparatus Fig. 3 shows the general arrangement of the experiment. The bulb and coil were completely enclosed in a smalllead-covered room connected to earth. The rays passing through a circular hole in the lead, covered with aluminium, fell on one of the p]atinum grids A. A pencil of the rays, after traversing the B

D

EarI:h.

A

A' EarI:h. EarI:h. Fig.3

Energy 0/ Röntgen ami Becquerel Rays

265

grid, passed through a rectangular hole in the thick lead plate B, and made the air a partial conductor inside the discharge cylinder D. The vessel D merely served as a means of testing the constancy of the rays given out by the bulb by noting the current produced between the charged electrodes. The discharge apparatus will be described in detaillater. The two grids A and A' formed two arms of a Wheatstone bridge (fig. 4). The other two arms were formed by a manganin cylinder potentiometer of 22 ohms, corresponding to a length of about 25 metres of wire. A sensitive low-resistance galvanometer was employed, and the deflection read with a telescope and scale. The balance was first obtained for a momentary passage of the battery current. The rays were then turned on for a given time, generally either 30 or 45 seconds. The rays falling on the grid A were partly absorbed and heated the platinum to a slight extent, and the resistance consequently changed. The deflection from zero was noted immediately after the cessation of the rays. Inorder to obtain a measure of the heating etfect, a steady current was sent through the grid A for the same time as the rays acted. The magnitude of this current was adjusted until the deflection from zero was the same as for the rays. When this is the case, the amount of heat, H, supplied per second by the rays is eq ual to the amount of heat generated by the current: H where and

=

O· 24 ;2 R gramme calories,

i = current through the grid, R

=

resistance of the grid.

The heating effect due to the rays was sm all and consequently care had to be taken to avoid disturbancesof the balance due tooutside causes. The grids were enc10sed in a lead vessel with an aluminium window in front ofthegrid A. A thick covering of feIt completely enclosed the lead vessel. Between the bulb and the grid there was one plate of aluminium 1 millim. thick and two sheets of thin aluminium, besides the feIt covering. A lead screen in addition could be placed over the hole H. When the hole was covered thus with the lead plate, there was no disturbance of the zero, showing that the rays falling on the grid were responsible for the heating etfect and the rays alone. In practice it was found necessary to remove the sensitive astatic galvanometer employed a considerable distance away from the induction coil before the magnetic disturbances due to it were negligible. This necessitated additional leads, and in consequence more troublesome changes of the balance. For the most part, however, the changes of the balance point were gradual and, if necessary, could be accurately allowed for du ring the short time the grid was exposed to the rays. Two observers were required, one to start and stop the rays and to note the electrometerdeflections, and the other to observethegalvanometer deflections. With the aid of a simple system of signals, the experiments presented no serious difficulty. 1*

266

The Collected Papers of Lord Rutherford

In order that the rays should be, as far as possible, of eonstant intensity when falling on the grid, the hole in the lead plate between the bulb and the grid was eovered with a lead screen, operated from a distance, during the first 15 seconds after the rays were turned on. By means of a cord the screen was suddenly removed and the rays stopped after a definite time by breaking the current. The following table is an example of a succession of observations extending over several hours:Time of exposure to the rays = 30 seconds. Testing current through the grids = o· 04 ampere. Deflection of electrometer in scale divisions per second

Deflection of galvanometer in millims

17·0

5·17

18·6 18·4

5·53

5·33 5·27 Mean deflection = 5·32.

17·8 Mean value = 17· 9.

In order to measure the amount of heat corresponding to this deflection of the galvanometer, a steady current from aseparate battery was passed through the grid previously exposed to the rays and for the same time. Fig. 4 shows the connections. As it was necessary to determine the change of zero immediately after the passage of the eurrent for a definite time, the arms of the bridge were undisturbed, and consequently a portion of the heating current passed through the grid R 2 •

R

R,

R, G 3

Fig.4

A

r:nel:f?Y

ol Riill1gen amI Becquerel Rays

267

Resistance of grid R 1 exposed to rays :-.:" 4· 2 ohms. .. second grid R 2 = 4· 28 ohms. Total resistance of the other two arms and leads, S = 23· 04 ohms. If i be the current from the battery supplied for heating purposes then, Current through R I = " " R2 = Heating effect on R 1 = "R2 = ..

0·867 i. O· 133 i. 0·752 j2 gramme calories. 0·018 j2 "

Difference in amount of heat supplied to R I and R 2 = O' 734 ;2. We thus see that most of the heating effect is confined to the grid R I . From a special series of experiments it was found that the deflection from zero of the galvanometer in a given time due to the heating by the current was very closely proportional to the square of the current. It was therefore not necessary to find experimentally the exact value of the current to give the same deflection as the rays, but from observations on one known current the results could be obtained by interpolation. It was found that under the same conditions as the table given above, a current i = 0·0200 ampere gave a mean deflection from zero in 30 seconds of 37· 4 divisions. The mean deflection due to the rays was 17· 9 divisions. Thus the amount of heat supplied to the grid per second by the rays ==

17·9 '37.4

x

(0'02)2

x

O' 734

=

0·000141 gramme calorie.

In the first stage ofthe investigation a null method was employed to measure the heat of the rays, but unexpected difficulties arose, and the method was abandoned. The battery current was kept steadily flowing through the grids, and the balance obtained. During the time the rays were on, a portion of the current through the grid was shunted through a resistance of known value. The value of the shunt resistance was adjusted until there was no change of the balance immediately after the rays were stopped and the shunt circuit broken. It was difficult, however, to obtain satisfactory results, partlyon account of the inconstancy of the rays, but chiefly on account of the slight difference of heating effect of the two grids for equal currents. The strength of current through the grid was generally 0·04 of an ampere, and with this current the inequality of the grids was immediately seen by a change of balance, when the current was applied for some time. The addition of a shunt to one grid caused a variation of current through both grids, and the change of temperature, due to inequality of the grids, introduced an error which was not negligible compared with the small heating effect of the rays. The method was not so rapid or certain as the one finally employed.

268

The Collecled Papers of Lord RU/her/orel Measurement of Heating Effect h)' Thermopile

Some experiments were made to see if a thermopile was suitable for a measure ofthe heating effect ofX rays. The only thermopiles in the laboratory were of the ordinary solid type of 65 bismuth antimony couples. The thermopile was placed inside a metal tube covered with aluminium at one end and a rock-salt plate at the other. With a sensitive low-resistance galvanometer a defiection of 15 millims. could be obtained in 30 seconds. The rate of supply of heat was standardised by using a standard Hefner amyl-acetate lamp, the total radiation from which has been determined in absolute measure by TUMLIRZ. * It was observed that the thermopile, when exposed to the X rays, took up its final temperature very much more slowly than when exposed to the radiation from the lamp, and that the results obtained differed considerably from the bolometer method. The cause of the discrepancy lies in the unsuitability of asolid thermopile for measurements on X rays. The radiation from the lamp falling on the lampblack coating of the thermopile is absorbed at the surface of the metal, while the X rays penetrate a distance of the order of 1 millim. before much of the energy is absorbed. On account of this, the maximum rise of temperature near the surface of the junction on which the E.M.F. depends is less with X rays than with light for equal intensities of radiation. The defiection of the galvanometer is thus less for X rays than for light-waves of equal energy. From these considerations it is obvious that the solid type of thermopile is most unsuitable for such work; but a modified thermopile of thin plane sheets of metal, e.g., iron and constantan, would probably give better results and be simpler to manipulate than the bolometer. With thin sheets the heat would be equally distributed over the cross-section due to diffusion, and no appreciable error would arise. The method, however, has the objection that the amount of heat must be standardised by a known lamp or source of radiant energy. Total Energy of the Rays emitted per Second

When X rays fall on a metal plate, the plate is heated, and the question at once arises whether we are justified in assuming that the energy of the rays stopped by the metal plate is transformed into heat in the plate. The experiments of PERRIN, SAGNAC, and J. J. THOMSON have c1early shown that when X rays strike asolid body, secondary rays are set up which ionize the gas and act on a photographic plate. These secondary rays are of a far less penetrating character than the rays that excited them; but on account of the ease with which they are absorbed in the gas, the amount of ionization per cub. centim. in the gas near the surface of the body may be greater than that due to the direct rays. The total number of ions produced by the scattered rays depends to a great extent on the density of the metal as weil as on the .. 'Wied. Annal.', vol. 38, p. 640.

Ellel'gy 0/ Röntgen and Becquerel Rays

269

intensity of the incident rays. The total number of ions produced by complete absorption of the scattered rays is generally only a small proportion of the number produced by complete absorption of the direct rays. Assuming that an ion in both cases requires the same expenditure of energy to produce it, the energy of the scattered rays is thus only a small proportion of the total energy of the incident rays. The secondary rays are set up both at the points of incidence and emergence of the rays falling on the grid. The heating effect on the grid is thus less than the heat equivalent of the energy of the rays stopped by the grid by the portion of the energy used up in exciting secondary rays. The correction is probably smalI, and has been neglected in these experiments, but it is hoped in a future investigation to determine its value. There is no evidence that the chemical energy of platinum is in any way altered by the passage of the rays through it, and, as far as our present knowledge goes, the energy of the rays stopped minus the energy of the scattered radiation, is transformed into heat within the platinum. In a case where there is a chemical change, e.g., when the rays fall on a photographic film, the heating effect would not be the equivalent of the energy absorbed. It has been shown by RÖNTGEN and other observers that the intensity of the rays given out from the front surface of a platinum plate of a focus tube is approximately equal in all directions. In the experiment the rays fell normallyon the centre of the grid, but on account of the size of the grid, the intensity of the rays could not be considered constant over its surface. The intensity of the rays diminishes, and the obliquity of the angle of incidence increases from the centre of the grid outwards. In consequence of this, a greater proportion of the incident radiation is absorbed at the edges than at the centre. The intensity of the rays was cut down to about 0·45 of its incident value in passing normally through the grid. It can be shown, by approximate integration over the surface of the grid, that for the distance of the grid from the source of the rays, namely, 26 centims., and the dimensions of the grid, the actual energy absorbed is about 2 per cent less than if the rays had the same intensity over the surface of the grid as at the centre, and had fallen normally at all points of the grid. For the special bulb employed, it was shown that the rate of supply of heat to the grid was equal to 0·00014 gramme calorie per second. This corresponded to a maximum rise of temperature of about 1/200° C. Distance of the cent re of the grid from the source of the rays = 26 centims. Area of grid = 92·2 sq. centims. Now 0·55 of the incident radiation was absorbed in the grid. Total energy of the rays falling on the grid is approximately = 0·00025 gramme calorie per second.

270

The Collected Papers

0/ Lord Ruther/ord

Therefore the total heating effect due to all the rays emitted from the front of the plate (omitting absorption in the glass, air, and screens)

=

277 X (26)2 277 X

X 0·98

x 0·00025 = 0·011 gramme calorie per second, or o· 046 watt.

Now the number of discharges per second in the bulb was 57, and TROUTON* has shown that the duration ofthe rays during each discharge of an induction coil is less than 10- 3 second, and probably about 10- 4 second. Assuming the average duration of the rays for eaeh diseharge is 10 - 4 seeond, the rate of emission of energy while it lasts =

1·95 gramme ealorie per seeond.

The heating effect ofthe sun's rays falling normallyon 1 sq. centim. surface is about = O· 035 gramme calorie per second. The maximum rate of emission of energy as X rays from the bulb is thus about 56 times greater than the amount of energy per sq. centim. due to the sun's rays. Efficiency

0/ a Fluorescent Screen as a Source of Light

Experiments were made to determine the efficieney of a fiuoreseent screen as a transformer of Röntgen radiation into visible light. Photometrie observations of the light emitted by a fiuoreseent screen, exeited by X rays, have been made by A. MOFFAT,t who dedueed the energy of the rays, by assuming that the coeffieient of transformation of the energy into visible light was 4 per cent., the value found by E. WIEDEMANNt for the transformation of radiant energy into lumineseenee. It was not the object of this investigation to make a eomplete photometrie eomparison, but to deduee an approximate eoeffieient of transformation for a definite experimental arrangement which could readily be reprodueed in practice. For this purpose a pieee of fiuorescent sereen was placed over one of the diffusive surfaees of a Lummer-Brodhun sereen, and the diffused light of the screen eompared with the diffused light of the amyllamp in the usual manner. The ratio of the square of the distanee of the sereen from the bulb to the square of the distanee of the lamp was taken as the ratio of the intensities of the light emitted by the X-ray bulb and lamp. In this case the amount of the light of the amyl lamp absorbed in the pIaster of Paris surface was neglected. Dr. E. SUMPNER § has shown that a piece of blotting-paper refiected over

*

'Brit. Assoe. Report', 1896.

t 'Wied. Annal.', vol. 37, p. 233.

t 'Roy. Soc. Edinburgh Proc.', 1898. ~

'Phil. Mag.', February, 1893.

FI1l'/'gy

0/ Riintgl'11 mu/ Bl'cqul'/'l'/

Rays

271

RO per cent. of thc light incident upon it, and it was found experimentally that the pIaster of Paris surface of the screen was a still better reflector. The current in the discharge vessel was determined, during the measurements of the energy, in absolute measure, and also during the comparison of the screen with the amyllamp, in order to correct for changes of intensity of the rays during the observations. In this way it was found, using a platinobarium cyanide screen, that

Intensity of light from fluorescent screen _ . 06 - 0 02 . Intensity of light from amyllamp Now if the intensity, T, of the visible light from a Hefner amyl lamp is given by 1= K/r 2 from the experiment of TUMLIRZ* the value of K for the visible light is equal to O' 00361 gramme calorie per second, and the total energy radiated by the lamp is 41·1 times the energy ofthe light radiation alone. Now in the experiments with the bolometer the heating effect of the rays incident on the grid, area 92·2 sq. centims., at a distance of 26 centims. from the source of rays, for the same strength of rays as those incident on the fluorescent screen, was O· 00032 gramme calorie per second. 1 ·c. Kjil, and the value of K, which represents the amount of energy due to the rays falling normallyon a surface of 1 sq. centim. at a distance of 1 centim. from the source of rays

= O' 0023 gramme calorie. The intensity of the X rays was thus 0·64 of the intensity of the visible light of the standard Hefner lamp. Now the efficiency of transformation of X rays into light. energy radiated as light = O' 0206 x o· 00361 = 0.044 O' 73 y 0·0023 ' energy supplied by the rays since it was found electrically that 0·73 of the rays were absorbed in the screen. The efficiency of transformation is thus 4·4 per cent. [f we assurne that 85 per cent. of the incident light is diffused from the surface of the Lummer-Brodhun screen, the efficiency of transformation is about 3·7 per cent.

* ·Wied. Annal.', vol. 38, p. 640.

272

The Collected Papers

0/ Lord Rutherford

A calcium tungstate screen, in which the absorption was O· 36, gave almost the same efficiency of transformation. The results we have obtained afford a simple means of expressing the intensity of X rays in absolute measure, assuming the coefficient of transformation of a fluorescent screen to be about 4 per cent. Two experiments would be necessary(I) The intensity of the light from the screen would be compared with a Hefner standard lamp. (2) The absorption of the rays by the screen would be measured electrically or photometrically by placing a portion of the screen to absorb the rays. Let 11 and 12 be the intensities of X rays and a Hefner standard lamp in absolute measure, disregarding absorption of rays in glass, metal screens, air, &co When there is equality of illumination let '1 and '2 be the distances of the source of rays and lamp from the Lummer-Brodhun screen. Let '1 and '3 be the distances for equal illumination when the rays pass through a piece of the screen before falling on the Lummer-Brodhun screen. Then p, the ratio of transmitted to incident rays for the fluorescent screen, is given by p = r:Jr~.

Ratio of incident energy absorbed Therefore it is readily seen that 1 -1 12

-

=

1 - p.

,~ ,~(1

- p)

100 x --4·4'

since efficiency is 4·4 per cent. Thus, since 12 = 0·00361 gramme calorie, 2

... 11 = 0·082 ,~(1r~ p) gramme calorie. Thus two simple photometric comparisons would be required to express the energy of the radiation of any particular bulb in absolute measure. For penetrating rays the absorption in the cardboard of the screen is negligible, but if necessary it can readily be allowed for. The chief SOUfce of difficulty in the comparison is the difference in colour between the light from the Hefner lamp and a fluorescent screen. The fiuorescent light appears a greenish-blue, and the amyllamp a reddish-yellow, when seen side by side in the telescope of a Lummer-Brodhun screen. Some experiments were made using a coloured glass to make the sources of light more nearly a match in colour. A greenish coloured glass was found to give a good colour match when interposed between the screen and amyl lamp. On determining by means of a thermopile the amount of the visible

Dw(~y

(Ir Rijntgell amI IJe('l/lIel'l'I Rays

273

energy of the Hefner lamp which was allowed to go through the gl ass, it was found to be so small (less than 2 per cent.) that a special investigation was required to find the coefficient of transmission with accuracy. The experiments were not continued. owing to lack oftime, but the evidence showed that by using a coloured screen in the path of the amyl lamp to give the same tint as the fluorescent screen, the efficiency of the transformation of the screen was much higher in that case than the 4 per cent. obtained by using no coloured glass or solution.

EnergJ' required 10 produce an Ion The method employed to determine the energy required to produce an ion in gases exposed to Röntgen rays depended on the measurement ofthe heating effect of the rays, and of the total number of ions produced by the radiation in the gas. If H is the number of calories given out per second by the rays, and E the energy in ergs of the rays emitted per second, then

E=JH .

•

(1).

If W is the average amount of energy required to produce an ion, then nW = E

(2),

where n is the number of ions produced per second, supposing that all the energy of the rays, absorbed in the gas, is due alone to the production of ions. In order to determine the number of ions, n, it is necessary to measure the maximum current that can be produced between two electrodes when all the ions produced by the rays in the gas reach the charged electrodes before there is any appreciable loss of their number due to recombination. If i is the maximum or saturation current through the gas, then

i = nE,

where ~ is the charge on an ion. The value of ~ has been determined by J. J. THOMSON,· and is equal to 6·5 x 10- 10 electrostatic unit. From (1) and (2) nW = JH, therefore

JHE W - -i-'

In order to determine W it is thus necessary to determine the value of Hand i. The considerations on which the method is based are: (1) When the X rays are absorbed by asolid substance, the greater proportion of the energy is given up to the substance in the form of heat. (2) The energy of the rays absorbed in passing' through a given volume or the gas is used up in producing ions. .

* ·Phil. Mag.', Dec., 1898.

274

The Collected Papers nf Lord Rutherford

(1) has been considered earlier in the paper, and it has been shown that we are probably justified in assuming that a very large proportion of the energy due to rays absorbed in a substance like platinum is transformed into heat. A small proportion of the total energy is used up in setting up secondary rays at the point of incidence of the rays on asolid conductor and also at the point of emergence. In regard to (2), one ofthe authors* has previously shown that the absorption of the rays in agas is roughly proportional to the intensity of the ionization in the gas. Gases and vapours, which are made good conductors by the rays, also strongly absorb them. The absorption of the rays in the gas has no direct connection with the molecularweight or density ofthe gas. For example, in hydrochloric acid gas the rays are far more readily absorbed than in carbonic acid, agas of greater density. PERRIN has shown that the ionization of agas is approximately proportional to the pressure. This result has been confirmed by us, and the authors have also found that the absorption ofthe rays varies directly as the pressure, i.e., as the ionization of the gas. These results point to the conclusion that the absorption of the rays in agas is closely connected with the number of ions produced. It is possible that there is a certain amount of scattering of the rays in passing through agas, but if the apparent absorption of the rays were due in any great measure to scattering, we should expect the absorption to depend chiefiy on the density of the gas, and such is not the case. RÖNTGEN and others have observed that the gas itself which has been acted on by the rays gives out aradiation which is able to light up a fiuorescent screen. This radiation may be due either to the scattering of the rays, or to the radiation caused by the recombination of the ions. In either case it is probable that the radiation is of a type similar to the secondary radiation set up at the surface of metals when X rays impinge upon them. This secondary radiation is far more readily absorbed in gases than the primary radiation, and would be absorbed in producing ions in the gas. The rate of discharge would be increased, and provided all the scattered, or secondary, radiation were used up in producing fresh ions between the electrodes, no correction for the amount of scattered radiation would be required. This of course proceeds on the assumption that the ions produced by the primary and secondary radiation are the same, and require the same amount of energy in each case to produce them. It is not practicable to measure directly the total maximum current through the gas, due to the passage of all the ions produced between charged electrodes, as the rays could pass through several hundred metres of the gas before approximately complete absorption took place. In practice the number of ions produced in a known small volume of the gas is determined, and also the coefficient of absorption of the rays by the gas. The total number of ions that would be produced, provided all the rays were absorbed, can be directly calculated. • RUTHERFORD,

'Phil. Mag.', April, 1897.

Energ}' (!{ Röntgen (11/(1 Becquerel Ra)'s

275

An account will now bc given of the experiments performed to measure the absorption of the rays in gases. Absorption 0/ the Rays in Gases

The bulb employed gave out rays of great intensity and penetrating power, and the absorption of the rays in air was smalI. About 3 per cent. of the rays were absorbed in passing through a metre of air at atmospheric pressure and temperature. In order to measure the absorption, a delicate null method was employed. No direct method can be employed on account of the smallness of the absorption and the variation of the intensity of the rays during the experiments. Fig. 5 shows the general arrangement of the apparatus. A similar method was employed by one ofus* on a previous occasion to measure the absorption of the rays in gases.

N

&rl:It.

L JI

C

A

D

C'

A'

d ~.

E.rllt. Fig.5

Two long brass tubes, A. At, 118centims.long and 3 ·4centims. in diameter, were placed horizontally at a slight angle to each other, and in such a position with regard to the bulb that the axes of the two tubes met at a point on the surface of the platin um plate of the anticathode. The ends of the tubes were covered with aluminium ca ps 1 millim. in thickness, and were made air-tight and capable of standing apressure of 3 atmospheres. The rates of discharge due to the rays after passing through the tubes, were taken between two sets • E.

RUTHERFORD,

'Phil. Mag.', April, 1897.

276

The Collected Papers ofLorel Rutherforel

of parallel plates, CD and C'D'. The plates D and C' were of the same size and cut into three portions, ofwhich the centre plates were carefully insulated. The centre plates were thus surrounded by a guard ring, and the rates of discharge to the centre plates alone were measured. The centre plates, D and C', were connected together, and to one pair of quadrants ofthe electrometer, the other pair of quadrants being connected to earth. The plates, C and D', were connected to the terminals of a battery of small storage cells of 310 volts, the middle point 0/ which was to earth. The electrometer will show no deflection if the intensity of the rays between C and D is exact1y equal to the intensity between C' and D', since the current between C and D is equal and opposite to the current between C' and D/. Lead screens, L, M and N, were placed at the positions marked in the figure, in order to prevent any stray radiation from reaching the testing plates. The wires leading to the electrometer were enclosed in metal tubes, which were connected to earth in order to avoid any loss of charge due to stray radiation or disturbances by any electrostatic field. The electrometer was completely surrounded by a wire gauze. The separation of the quadrants of the electrometer was operated from a distance by means of suitable keys. Such precautions are very necessary during the very dry Canadian winter, when the slightest movement causes frictional electrification. The table and the woodwork on which the apparatus was placed was covered over with metal, to prevent the collection of charges either from frictional electrification or the action of the rays near charged conductors. The tubes were first adjusted so that the rays caused no movement of the electrometer needle. The tube, A, was then rapidly exhausted by means of a Fleuss pump. The intensity of the rays after emerging from the tube was thus greater than for the tube A' on account ofthe less absorption, and the electrometer therefore showed a deflection. If the tube A' were exhausted and the tube A filled with air, the electrometer gave a deflection in the opposite direction. If the end of one of the tubes was closed with a thick lead plate so that no rays could get through, then the rate of movement of the electrometer needle corresponded to the intensity, 1, ofthe rays after emergence from the other tube. If ,\ is the coefficient of absorption of the rays in the gas, then the intensity of the rays after passing through a distance d of the tube is e- M of its value if there had been no absorption. Since the currents between C and D and between C' and D' are proportional to the intensities of the radiations between the plates, then Difference between currents I Total current = 1= '\d,

Ie- M

I

e- M if,\d is small.

In order todetermine '\d, we thus require the ratio ofthe number ofdivisions

Energy

0/ Röntgen and Becquerel Rays

277

per second, given by the electrometer needle from the balance when one of them is exhausted, to the number per second when the end of the tube containing the gas is covered with a thick lead plate. The following tabte gives the results for air at pressures in one tube ranging from 0·5 of an atmosphereto 3 atmospheres, the other tube being exhausted:Difference of pressures in atmospheres

0'5 1 2 3

I Number of divisions in 5 sees. with one tube screened

-

Deftection from balance in 20 sees.

-

160 169 172

21·8 48·0 70·0

M

0·0187 0·034 0·071 0·102

The above results are the mean values of aseries of measurements. The results for 0·5 of an atmosphere were obtained at a different time from the others and with a different sensitiveness of the electrometer. The table shows that the absorption of the rays in the gas is approximately proportional to the pressure. The value of d was 118 centims. The value of Afor different pressures is thus given by the following tabIe:Pressure in atmospheres

0·5 1 2 3

Value of"

0·000158 0·000288 0·00060 0·00086

The value of A at atmospheric pressure and teJ;l1perature obtained for the same bulb after daily use for two months was found to be O· 000270. The value of A at atmospheric pressure and temperature in the calculations is taken as the mean of these two values, and is thus given by A = 0·000279. It is probable that the rays were not homogeneous, and the value of A must be considered as the mean value for the different kinds of rays. On account of the very small absorption of the rays it was difficult to determine with accuracy the absorption for pressures lower than half an atmosphere.

278

The Collected Papers of Lord Rutherford

The results, however,indicated that the absorption was, roughly,proportional to the pressure for stilliower pressures. The value of A was determined for carbonic acid gas at normal pressure and temperature. The absorption was 1· 59 times that of air, and the value of A was found to be 0·000457. The results were confirmed by varying the pressure of the air in one tube until there was no disturbance of the electrometer zero. The results agreed with the value obtained above, assuming the absorption in air is proportional to the pressure. We see from the results given above that the radiation is reduced to half its value with no absorption after passing through a length of 24· 7 metres of air at ordinary pressure and temperature. The value of Afor uranium rays* is 1 . 6, or the absorption is 6000 times as great for uranium rays as for the X rays employed. The value of A obtained some years ago for a much 'softer' bulb was 0,001, or about four times the absorption of the bulb employed in these experiments.

M easurement

0/ the Current through the Gas

In order to determine the amount of ionization in a known volume of gas, the apparatus shown in fig. 6 was employed. The rays passed into a brass cylinder, 12 centims. in diameter and 30 centims. in length, through a rectangular orifice, 0, at one end covered with an aluminium window, 1 millim. in thickness. Inside the cylinder two parallel rectangular plates, A and BCD, were fixed on a light wooden frame. The plate opposite to A was cut into three parts, B, C, D, and C was insulated from Band D. The plate A was connected to one pole of a battery of 310 volts, the other pole of which was connected to earth. The plate C was connected to one pair of quadrants of the electrometer, the other pair of which was connected to earth. The plates Band D were in connection with the cylinder, which was also connected to earth. The plates Band D thus corresponded to a partial guard ring for the plate C, and served two purposes. The electric field was rendered uniform from C to A, and most of the secondary radiation set up at the two ends of the cylinder was absorbed between A and B and between A and D, and thus did not produce appreciable ionization between A and C. A large lead plate, L, with a rectangular orifice, was so placed that the rays from the source passed into the cylinder, and did not fall on the parallel plates. This avoided the presence of secondary radiation. The two ends of the cylinders were covered inside with cardboard in order to make the amount of secondary radiation as small as possible. The amount of radiation set up at the surface of air and cardboard is very small. The amount of insulating material inside the cylinder was reduced as far as possible in order to avoid the collection of free charges on them, and consequent disturbance of the electric field. For this reason the plates were mounted on a wooden

* E. RUTHERFORD, 'Phi!. Mag.', January, 1899.

EnerKY

4

Röntgen ami BeclJuerel Rays

279

frame instead of an ebonite one. The wood was a sufficiently good conductor to quickly discharge any electrification that reached its surface. The current between C and A is thus due only to the ions which were produced by the passage of the rays between them. The length of the plate C was 12·06 centims., measuring from the centre of the air spaces. The distance between the plates was 4· 16 centims. The rays, before entering the cylinder, passed through a rectangular orifice in a thick lead plate. Knowing the distance from the source of the rays and the area of the opening, the area of a seetion of the cone of rays at any point in the cylinder could be determined. L

A

A

A

A

Eareh.

A

E4J'th.

&.reh. Fig.6 The results calculated in this way were compared with the area of the impression on a photographic plate at different distances from the orifice, and it was found that the correction to apply for the source of the rays, not being a point source, was practically negligible. Let 12 be the intensity of the rays at the beginning of the plate C, and S the area of the cross-section of the cone of rays at that point. The energy crossiug the surface per second is 12S. If there was no absorption of energy in the gas, the energy crossing the cross-section of the cone of rays at the further end of the plate C would be the same as at the beginning. But in consequence of absorption the enel'gy crossing the surface per second is

I2Se- ',1,

280

The Collected Papers of Lord Rutherford

where 1= length of the plate C, and A= the coefficient of absorption of the rays in the gas. The energy absorbed in the gas per second is equal to 12S(I - e-.)") = 12SM if M is small, as was the case in the experiments. The current between the plates A and C was determined for a voltage sufficient to move all the ions to the electrodes before any appreciable recombination could take place. Let n = total number of ions produced per second. i = current between the plates through the gas. e = charge on an ion. W = average energy required to produce an ion. Then

Wn .'.

=

W =

12SAl and

i

= ne-;

12S~IE. l

The value of 12 was determined from the heating effect of the rays, as cxplained in the earlier part of the paper. Let I I = intensity of the rays at the surface of the bolometer. PI = transmission ratio of the rays when passing through the platin um grid. Then energy absorbed in the grid per second = I1A(l - PI), where A = area of grid. Assuming the value of the intensity uniform over the surface of the grid, and equal to the value at the centre, the total energy absorbed in the grid is slightly less than the above amount, and it has been shown earlier in the paper that very approximately the energy absorbed in the grid = 0'98I I A(l - PI)' In practice the current in the discharge cylinder was observed at the same time as the heating effect. The cylinder was placed behind the platinum grid in such a position that the rays entering the cylinder passed slightly to one side of the centre of the grid, thus avoiding the mica frame of the bolometer. The rays before entering the cylinder were cut down in intensity by their passage through the grid, by the enclosing envelope and the aluminium window in the discharge cylinder.

+

Let P2 = transmission ratio of the rays through the platinum grid the feIt cover the aluminium window, &c. d 1 = distance of grid from source of rays. Let d2 = distance of the beginning of the plate C in the discharge cylinder from the source of the rays.

+

DIl'/XY (!I' IWntgell amI Becquerel Rays

2Rl

Then it can ea .. ily he seen that

4

2 --I ~ ,- - P2d e -i,(d\ -- d~J I1 d2

• ( 1).

The factor e-i,(d\-dz) is nearly equal to unity, and is the correction fOT the absorption of the rays in the gas between the grid and the discharge cylinder. If H is the number of heat units communicated to the bolometer per second, then (2), 0·98AI I (I - PI) = JH I 2 SAfE W=-.-

,

and

(3).

Dividing (3) by (2) and substituting the value ofT 2/I I from (1), we obtain W

=

JH Q.98A(I _

di

SAfE

p)' -,-. . P2 . d~ e

Determination

i,(d\- tI2)

0/ i

The value of i was determined by an electrometer with an additional capacity of 0·00248 of a microfarad in parallel. The heating effect on the bolometer and the quantity of electricity discharged between the plates of the cylinder were observed at the same time. A lead screen cut off the rays from the platinum grid and the discharge cylinder for 15 seconds after the bulb had started, for it was found that the rays gradually increased in intensity for the first 10 or 15 seconds. At the end of 15 seconds the lead screen was suddenly removed by a cord operated from a distance. After the passage of the rays for 30 or 45 seconds, the rays were stopped. The deflection from the zero of the bolometer was taken by one observer, while the deflection of the electrometer was taken by another. In the later experiments the capacity of the electrometer and connections was 1.17 of the capacity added, and the total capacity of the circuit was 0·00257 of a microfarad. The following is an example of the determination of i in electrostatic units:1 Clark cell of 1·434 volts E.M.F. gave 57·0 divisions on the electrometer scale. The deflection of the electrometer due to the passage of the rays for 30 seconds was 160 divisions.

i = quantity of electricity per second _ ~:002~7 -- 106

X

160 30

x

1·434 v v 9 57 ,,3., 10

:-: 1· 03 electrostatic units.

282

The lollected Papers of Lord Rulherfol'd

Determination of the Absorption of fhe Rays in the Bolometer, the Aluminium Window, &c.

The values of PI and P2 were determined electrically by utilising the discharge cylinder of fig. 6. The rate of discharge in the cylinder was observed with the grid before the hole in the lead plate and then without the grid. The ratio of the currents in the two cases is proportional to the ratio of the intensities, since the ionization is proponional to the intensity of radiation. A mean value of the ratio for different portions of the grid was taken, and it was found that PI = 0·453. It is thus seen that the intensity of the radiation was cut down to a little more than half in passing through the platinum grid. The value of P2 was determined in a similar manner. The intensity of the rays in this case was cut down more in consequence of passing through a thickness of feit, an aluminium window of o· I centim. thickness, and a thin layer of aluminium, as well as the platinum grid. The value found for this ratio was P2 = 0·31. The absorption of the rays in the mica frame appears in the values of PI and P2' The absorption, however, was small and practically negligible in any case. Dimensions of the Apparatus and Values of Constanls

The area of the rectangular hole through which the rays passed into the discharge cylinder, and from which S was calculated, was 7·1 sq. centims. The length of the centre plate in the discharge cylinder was 12·06 centims. For most of the experiments, the distance d of the grid from the source of rays was 26· 0 centims. and the distance of the hole in the lead plate from the source was 45· 2 centims. Area of platinum grid was 92· 6 sq. centims. Mean value of'\ = 0·000279. Value of E" = 6·5 X 10- 10 electrostatic unit. The correction for the absorption in air between the grid and the discharge cylinder was negligible. The following table (p. 289) gives the values of i, H, and W for different times of the exposure to the rays. The mean value of W, the energy required to produce an ion in air at atmospheric pressure and temperature, is given by W

=

1·90

X

10- 10 erg.

The energy required to produce a positive and a negative ion from a neutral moleeule is twice this amount, and since one ion cannot be produced without the other, then 3·8 x 10- 10 erg is the smallest amount of energy that wi1l produce ions in air.

283

Energy of Röntgen and Becquerel Rays

The mean intensity of the rays in absolute measure at the surfaee of the bolometer is given by JH = 0·98Al 1(1 - PI)' Taking the value of H as I· 5 x 10- 4 ealorie, we find that 11

=

127 ergs.

On aeeount of the very short duration of the rays from eaeh diseharge, the maximum intensity of the radiation at any time is probably over a thousand times greater than the above value. Time of exposure to rays

i in electrostatic

units

H in calories

45 sees.

0·894 0·976 1·045 1·115 0·996 1·00 1·07 1·03 1·09

1·49 x 10- 4 1·56 X 10- 4 1·47 X 10- 4 1·57 X 10- 4 1·38 X 10- 4 1·34 X 10- 4 1·47 X 10- 4 1·45 X 10- 4 1·41 X 10- 4

30 sees.

W in ergs

2·22 2·13 1·87 1·82 1·84 1·79 1·83 1·87 1·72

X X X X X X X X X

10- 10 10- 10 10- 10 10- 10 10- 10 10- 10 10- 10 10- 10 10- 10

The energy absorbed per second in producing ions in the eylinder

~W=

€

O' 29 erg, taking i

=

=

1 E.S. unit.

This absorption of energy is spread throughout a volume of over 100 cub. centims. of the gas, so that the absorption of energy per cub. eentim. in the air is very smalI. The value of W, the ionic energy, is seen to depend on the measurement of the current through the gas, the coefficient of absorption, and the heating effect of the rays. The absorption of the rays in the gas has to be determined separately from the eurrent and heating effect, and an uncertainty consequently arises on account of the variation of the penetrating power of the rays as the bulb varies. The value of ,\ determined for the rays, under different conditions as regards the frequency of the Wehnelt interrupter, was found to be approximately the same, after the bulb had been in constant use for several months. It is probable that the type of rays does not on an average vary much from day to day, but the greatest source of error is probably due to the assumption that the rays are homogeneous in character. RÖNTGEN and others have shown, from experiments on the absorption of successive thicknesses of metal, that rays are not simple in character, but contain rays

284

The Collected Papers of Lord Rutherford

of widely different order of penetrating power, so that the value of A is the mean value for the different types of rays. The value of W also depends upon the value of €, the charge on an ion, and if future investigations should assign a different value to €, the value of W would be altered in a like ratio.

Energy required to produce an Ion in other Gases When the energy required to produce an ion in one gas is known, the energy required to produce an ion in another gas can be determined from the ratio of the absorptions of the rays and the intensity of ionization in the gases. Let nl and n2 be the number of ions produced per cub. centim. in two gases. Let Al and A2 be the coefficients of absorption. Let W land W 2 be the energies required to produce ions in the two gases. Let i l and i2 be the maximum currents through the gases. Then for the same intensity of rays, absorption of energy in gas2 A21 n2W 2 absorption of energy in gas I = All = nl W 1 d . an 12 = n2€, assuming charges on the ions are equal.

.• = i.2W-2 smce '1 = 11 W I

nl€

W 2 A2 i l W 1 = AI . J;,'

Therefore

The ratios A2/AI and itli2 can be readily determined, and if W I is known, then W 2 can be calculated without recourse to experiments on the heating effect of the rays in each case. The value of A2/ Ab the ratio of the absorption coefficient of carbonic acid gas to that of air, was found to be 1· 59 for the rays employed. The ratio izli, of the current in air and carbonic acid gas for a potential difference of 300 volts was found to be 1· 43. Therefore the energy required to produce an ion in carbon dioxide 1·59

= 1'43W1 = l·11W I = 2·11 X 10- 10 erg.

This value is a little higher than in the case of air. The measured amount of i2 , the current in carbon dioxide, was somewhat less than the maximum, since the electromotive force applied was not sufficient to move all the ions to the plates before recombination. A correction for this would make the values for air and carbon dioxide more nearly equal. Taking the value 1 . 53, found by J. J. THOMSON, * for the relative ionization in carbonic acid and air, the ionic energies are nearly the same. ,. J. J.

THOMSON,

'Camb. Phil. Soc. Proc.', vol. 10, Part I.

FII('/,gy of Riilllg('11 amI B('cqu('/,el Rays

285

Thc results for air and carbon dioxide show that the energy required to produce ions in the two gases is not very different. The results of a previous paper* showed that the absorption of X rays in gases was roughly proportional to the ionization produced. From this it follows that the energy required to produce ions in the gases examined was, roughly, the same. The results obtained with uranium radiationt showed that the total number of ions produced by complete absorption of the radiation in air, oxygen, hydrogen, carbonic acid gas, hydrochloric acid gas, and ammonia were approximately the same. The results in that case were more readily obtained as the radiation was almost completely absorbed in a few centims. of the gas, and the maximum current through the gas was a measure of the total number of ions produced. The recent results of McLENNANt also point strongly in the same direction. In his experiments, cathode rays were passed out of the discharge tube into another vessel, and the maximum current produced by the cathode rays was found for different gases. Using a constant supply of cathode rays, the current, i.e., the total number of ions produced, was independent of the nature of the gas (provided the press ure of the gas was adjusted to give the same absorption of the rays in each case). The gases examined were air, hydrogen, oxygen, nitrogen, carbonic acid, nitrous oxide, and the total number of ions produced in them was nearly the same. Assuming that the same proportion of energy of the cathode rays was used up in producing ions in the gases, it follows that the energy required to produce an ion in all the gases is the same. The results on the ionization of different gases by the agency of Röntgen, Becquerel, and cathode rays all strongly point to the conclusion that the same energy is required to produce an ion whatever the gas. Variation o/Ionic Energy with Pressure

It has been shown earlier in the paper that from half an atmosphere to three atmospheres' pressure the absorption is proportional to the pressure. A special investigation has shown that for the same range the intensity of the ionization is also approximately proportional to the pressure. This shows that for the pressures examined the ionic energy is independent of the press ure. The results on the variation of absorption with pressure for uranium§ and thorium 11 radiation also point to the same conclusion. In order to fully establish such a generallaw that the energy to produce an ion is independent of the gas and its press ure, a large number of careful experiments will be required. The results so far obtained can only be considered to show that such a law is approximately true. It is intended to continue • E. RUTHERFORD, 'Phil. Mag.', April, 1897. t E. RUTHERFORD, 'Phil. Mag.', January, 1899. t 'Roy. Soc. Proc.', 1900; 'Phi!. Trans.', A, vol. 195. ~ E. RUTHERFORD, 'Phil. Mag.', January, 1899. R. B. OWENS, 'Phi!. Mag.', October, 1899.

286

The Collected Papers of Lord Rutherford

these investigations on ionic energy for other gases besides air and carbonic acid. Deductions from the Resu/ts

If ions of the same kind are produced in agas by different agencies, it is probable that the same amount of energy has been absorbed to produce the ions in the different cases. The only test we have at present for equality is to cornpare the velocity of the ions in the gas for a potential gradient of 1 volt per centim. J. J. THOMSON has shown that the charge on an ion produced by Röntgen rays is probably the same for the gases hydrogen, air, oxygen, and carbonic acid, and TOWNSEND* that it is equal to the charge on a hydrogen ion in the electrolysis of water. The velocity of the ions in a given electric field depends upon the ratio €/m of the charge to the mass of the ion, and thus if the velocities of ions produced in the same gas by different agencies are the same, the masses must be the same, since the charges are equal. It has recentlyt been shown that the ions in the 'electric wind' travel in air with the same velocity as the ions produced by rays. The energy used up in producing the ions can thus be immediately calculated. Let i = the current through the gas due to the electric discharge from a wire or point. Energy absorbed in producing ions

=

nW

~€ W,

=

where n is the number of ions produced per second and Therefore, neglecting recornbination of the ions, Energy required to produce ions Totalloss of energy

~W €

=

Vi

=

the charge.

€

Wl -;- V'

where V is the potential of the discharging wire. If V = 6000 volts = 20 electrostatic units, then the proportion of the total loss of energy used up in the production of ions =

1·90 X 10- 10 1 1 6·5 X 10- 10 • 20 = 69 approx.

Thus quite an appreciable proportion of the total energy supplied is absorbed in producing ions. The proportion decreases with the increase of voltage. Distance between the Ions in a Mo/eeu/e

If we suppose that most of the energy required to produce a positive and a negative ion from a neutral molecule is due to the work done in separating the ions from each other against the forces of electrical attraction, we can at • ·Phi1. Trans.', A, 1899.

t

CHATIOCK.

'Phi). Mag.', October, 1899.

t:ncrgy of RiilltgCIl emd Bccqucrel Rays

287

once form an approximate estimate of the distance apart of the charges in the molecule. The work done in se pa rating acharge + E from acharge - E, both charges supposed concentrated at points from a distance r to an infinite distance, is equal to E2fr. If this is equal to the energy required to produce two ions, then E2

-

r

smce

=

E =

3·8

X

10- 10

'

6'5 X 10- 10

r = 1·1 X 10- 9 centim. approximately.

The average diameter of an atom, calculated from various methods, is about 3 X 10- 8 centim. This is a very much greater distance than the value found for the distance apart of the charges on the ions in a molecule. The results support the theory advanced by J. J. THoMsoN, that ionization is produced by the removal of a negative ion from the molecule, and that the negative ion is only a sma11 portion of the mass of the atom. The positive ion is supposed to remain attached to the rest of the molecule. It is to be expected from the theory that the distance of the charges from each other would be less than the diameter of an atom. The energy required to produce an ion in air is very much greater than the energy required to produce an ion in the electrolysis ofwater. IfV (1,46 volts) is the least E.M.F. required to dissociate water, the work done in moving a quantity of electricity E is VE. The work done in producing an ion thus is 1·46 VE or 300 x 6·5 X 10-- 10 = 3 ·16 X 10- 12 erg, or about 6]Ö of the energy required to produce an ion in air by the agency of X rays, so that in water the ions are about two atoms apart. Least Potential required to produce a Spark in Air

The ions in the 'electric wind' in air have been shown to move with the same velocity as the ions produced by X rays. It is probable that the passage of a spark between two electrodes is heralded by the production of ions in the gas, and that these ions are of the same kind as the ions in air produced by X rays. Let V be the difference of potential between two electrodes in air, one electrode being connected to earth. Suppose a pair of ions to be produced and to travel to the electrodes. A quantity of energy 3·8 X 10- 10 erg is absorbed in their production, while the energy of the electric system is diminished by an amount VE. The energy required to produce the ions must be derived from the electric energy of the system. In order for an ion to be produced consistent with the conservation of energy, we must have V of

288

The Collected Papers oI Lord Rutlzel/orc/

such value that V€ is greater than the energy req uired to produce a pair of ions. V> V

3·8 X 10-- 10 ~ 10- 10 electrostatic unit. F

> 175 volts.

Now PEACE (J. J. THOMSON, 'Recent Researches,' p. 89) has shown that it is impossible to produce a spark in air below about 300 volts, however elose the electrodes are together. This is a somewhat greater value than the one found above, but is of the same order. STRUTT* has recently shown that the minimum potential difference for the passage of a spark in pure nitrogen is about 251 volts. As most of the ions in air are probably produced from the nitrogen molecules, this value makes the agreement still eloser. The results obtained would indicate that it would be impossible to produce an ion, and therefore an electric spark, below 175 volts. If the energy required to produce an ion were the same at all pressures, the minimum sparking potential according to the above theory would be unaltered. This is borne out by PEACE'S results (loe. eit., p. 86), where it is shown that the minimum potential difference for a spark between spherical electrodes 0·001 centim. apart is approximately the same for pressures from 300 to 50 millims. of mercury. The minimum potential rises below 50 millims., indicating that the energy required to produce an ion may possibly increase below that press ure. This theory would suggest that the minimum potential required to produce a spark conversely might be used as a means of determining the energy to produce an ion. The phenomenon, however, is more complex than this would indicate. The minimum sparking potential is to a small extent influenced by the metal used for the electrodes and also by the gas, and moreover it would leave unexplained the remarkable fact that when the electrodes are a small distance apart the spark does not follow the shortest path (loe. eit.) between them. Energy of Radiation

0/ Radio-aetive Substanees

In a previous paper it has been shown that the ions produced in air by uranium radiation have the same velocity as the ions produced by Röntgen rays. On the assumption that the same amount of energy is required to produce the ions, whether the agency is Röntgen or uranium rays, the energy of the radiation given out into the gas can be at once determined. Fig. 7 shows the arrangement of the apparatus to determine the current through air produced by the uranium rays. Two large parallel plates, A and B, 4· 1 centims. apart, were insulated from each other. A was connected to the electrometer in the usual manner, and B was connected to one pole of a battery of small cells of 310 volts. The uranium oxide employed was placed in a square shallow hole cut in a lead

* OPhit. Trans.', A, 1900.

See: 'The New Gas from Radium' (H. T. Brooks), page 303.

DIFFUSION APPARATUS FOR DETERMINING THE MOLECULAR WEIGHT OF RADIUM EMANATION

On Zeit: Lead cover provided with thin aluminium windows. On right: Bolometer grids of platinum strip, wound on mica frames. Each grid forms one arm of a Wheatstone bridge. See : 'Energy of Röntgen and Becquerel Rays, and the Energy required to produce an Ion in Gases' (McClung), page 266.

PLATINUM GRID BOLOMETER FOR STUDYING THE ENERGY OF RÖNTGEN AND BECQUEREL RAYS

This page intentionally left blank

J:;I/agy oI Riilllgel/ Clml !Jen/liere/ Ra)'.\'

2~9

plate placed on the plate B. The current between A and B was determined for a potential difference of 310 volts, an amount sufficient to practically remove all the ions before recombination. Since the area of the uranium surface is small compared with the area of the plates between which the uranium was placed, the total energy per second emitted by the surface S of uranium is approximately equal to IS, where I is the intensity of the radiation at the surface of the uranium.

A

Elirth.

B

E4rth.

larth. Fig.7

If ,\ is the coefficient of absorption of the rays in air, and W is the energy required to produce an ion, the energy absorbed per second between the plates at a distance d apart is equal to i IS{1 - e- Ad} = nW =-W, E'

where n is the number of ions produced per second, i is the current, and E' the charge on an ion. The total energy emitted per second is equal to

iW E'(l -

e -MY

In the case of uranium it has been shown (Ioc. cil.) that apparently two types of radiation are emitted, one of which is readily absorbed in air. The ionization due to the more penetrating rays is in general a small part of the total, especially for thin layers of uranium; so that in the present calculation we will only consider the energy given out by uranium in the production of the more absorbed type of radiation. The intensity of the radiation emitted from uranium falls to half its value after passing through 4· 3 millims. (Ioc. eit., p. 128). Therefore ,\ = 1· 6. K

290

77le

Collected Papers ofLord Rutherforel

Since d = 4,1, therefore e- Ad is small and may be neglected. Thus the energy given out into the air is !W. Now for a thick layer of E

uranium oxide (3' 6 grammes spread over a surface of 38 sq. centims.) the current i = 0·0515 electrostatic unit. Thus the energy emitted per unit area of uranium surface per second 6.5

=

X

0·0515 10- 10

x

38

x

. 1 90

x

10

-10 _

•

- 0 0004 erg

10 - l l calorie per second, approximately.

This amount of energy would suffice to raise 1 cub. centim of water 10 C., assuming no radiation of heat, in about 3000 years. It is a difficult matter to determine the total energy given off in the radiation by a given weight of uranium on account of the ease with which the radiation is probably absorbed by the heavy metal uranium itself in its passage through it. Some experiments were made on the current due to a given surface of uranium oxide when different depths of the active material were spread over it. The following are some of the resuIts:SURFACE of Uranium Oxide = 38 sq. centims. Weight of uranium oxide in grammes

Current in E.S. units per second

0·138 0·365 0·718 1·33 3·63

0·0201 0·0365 0·0471 0·0515 0·0560

The uranium oxide in the form of a fine powder was dusted on uniformly by means of a fine wire gauze. The results show that the current per gramme of uranium oxide is greater for small than for large thicknesses. Even with a very thin layer of uranium oxide in the form of powder it is probable that a large proportion of the energy emitted (supposed produced throughout the volume ofthe substance) is absorbed in the substance itself. An approximate determination ofthe total energy per second that can be radiated by 1 gramme ofuranium could be determined by dissolving a few crystals, say, ofuranium nitrate in water and pouring the solution over a large surface area. On evaporating the water a very thin film ofthe nitrate would be left on its surface. In such a case the rays produced throughout the volume of the film should reach the surface without much loss due to absorption, and the maximum

T:nclKl' o(

Riintgen and Becquerel Ra)'s

291

current through the gas would be proportional to the total energy radiated. In this case half of the total energy would be absorbed in the substance on which the film was placed and only half would be efficient in producing ions. In order to obtain an approximate value of the total energy of radiation, uranium oxide in the form of a very fine dust was spread over a surface of 38 sq. centims. Weight of uranium oxide = 0·138 gramme. Current = 0·02 electrostatic unit. Total energy radiated into the gas per second = 1·4 x 10- 10 calorie. Energy per gramme of uranium oxide radiated into the air = =

10-- 9 calorie per second. O· 032 calorie per year.

In our present state of knowledge it is uncertain whether the radiating power is confined to the surface of the uranium or is given out uniformly throughout the mass. In any case, the total energy radiated is probably greater than the value above on account of absorption of the radiation in the uranium itself, and also on account of the existence of a more penetrating type of radiation, the energy of which has been neglected in the above calculations. Energy of Thorium Radiation

The apparatus employed was the same as that for uranium. Thorium oxide was employed and the following results were obtained. Area of surface = 38 sq. centims.:Weight of thorium oxide

Current

gramme

0·339 0·665

O' 0445 electrostatic unit per second. 0·0622

"

"

"

In previous papers by OWENS* and RUTHERFORD, t the behaviour of thorium oxide as a radio-active substance has been carefully examined. It has been shown that thorium compounds give out a material emission of some kind, which possesses temporary radio-active properties. This emanation is most apparent with thick layers of thorium oxide. In the present case the layer was not thick enough to give out much emanation, and the rate of discharge was due to the radiation alone. OWENS has shown that the radiation from thorium is approximately homogeneous.

* 'Phil. Mag.', October, 1899. t 'Phil. Mag.'. January and February, 1900.

292

The Collected Papers o[ Lord Rutl1er[ol'd

The value of A, the coefficient of absorption of thorium radiation in air, is 0·69 and d = 4·1 .'. 1 - e- Ad = 0·96. Thus for a weight of 0·665 gramme the total energy radiated into the gas . iW d h . co • per umt area = O. 96eA erg, an , on t e same assumpttons as lor uramum, the energy radiated into the air per second = 1· 2 X 10- 11 calorie, a somewhat greater value than for an equal weight of uranium oxide. Excited Radio-activity due to Thorium

Thorium compounds, in addItion to the property of giving out a radioactive emanation, possess the power of exciting temporary radio-activity on a11 substances in their neighbourhood. The excited radiation is homogeneous in character, and is of a more penetrating type than the radiation from either uranium or thorium. The intensity of the excited radio-activity can be greatly increased by concentration on the negative electrode of small area by means of a strong electric field. On the assumption that the energy of the radiation excited on the electrode is dissipated in producing ions, an estimate can be formed of the energy stored up on the electrode. In a particular experiment a fine platinum wire, 0·018 centim. in diameter and 1 centim. long, caused the separation of about 10 coulombs of electricity before the radio-active power was lost. This corresponds to an emission of 2 x 10 - 5 calorie. This by no means inconsiderable quantity of energy is in some way derived from the surface of a platinum wire 0·056 sq. centim. in area, without the slightest appreciable change either in the weight or appearance of the wire. Radium and Polonium

The question of the equality of the velocity of the ions, produced by thorium radiation and the rays from the powerful radio-active substances radium and polonium, with the velocity of the ions produced by X rays, has not been specially investigated, but from the very elose similarity of the types of these radiations, it seems very probable that the ions produced by a11 are the same. In one respect, however, some of the radio-active substances, notably radium, differ in their type of radiation from X rays. BECQUEREL, CURIE, and others, in aseries of papers in the 'Comptes Rendus,' have shown that radium gives out a type of rays which are easily deflected by a magnet. This emission of rays similar in character to cathode rays of low velocity is very remarkable, but does not seem to be a necessary accompaniment of a radio-active substance. For example, GIESEL found polonium gave out rays deflected by a magnet, while BECQUEREL could obtain no magnetic action for the same substance. The rays which are deflected by a magnet seem to be present or absent according to the mode of preparation of the substance, and depend possibly on the age of the specimens. Two impure and not very sensitive

293

t:nergy of Röntgen and Becquerel Rays

specimens of radium and polonium obtained from pitchblende have been tested by one of us, but no magnetic action has been observed. BECQUEREL has found no trace of magnetic action in uranium radiation, and one of the authors has tested both uranium and thorium radiations in a magnetic field at atmospheric press ure and obtained negative results. The experiments of CURIE and BECQUEREL have shown that, in radium, two types of rays are present, one of which is deflected by a magnetic field and the other is not. The non-deflected type is similar in character to secondary X rays. and the deflected ones similar to low velocity cathode rays. We thus see that the phenomena exhibited by the radio-active substances are not simple, and that they differ from one to the other. I t is still possible, however, to form an approximate estimate of the energy of the radiation whatever its kind, provided the energy is all completely absorbed in ionizing the gas, and produces ions of the same kind. It seems probable that the radium rays acted on by a magnetic field are a type of cathode rays, and that they ionize the gas in their passage through it. The results of McLENNAN* c1early show that the energy of the cathode rays is lost in its passage through the gas, due partly to the work done in ionizing the gas in its path. Provided the ions produced by the deflected and undeflected rays of the radio-active substances are the same, and absorb the same amount of energy in their production, the relative energies of the radiations emitted can be compared by noting the total maximum current produced by the rays when completely absorbed between the electrodes. In n = the ratio of the currents between parallel plates for equal areas and thicknesses of the test substances and uranium oxide when the plates are at a sufficient distance apart to approximately absorb all the rays, then Energy radiated out by the test substances into the gas radiated by an equal area of uranium oxide.

=

n

X

the energy

This will probably apply roughly to the conductivity produced by the deflected and undeflected rays. In some experiments CURIE mentions using a specimen of radium 100,000 times more active than uranium. If this applies to measurement for equal weights and areas of radium and uranium, then the total energy radiated out into the gas by 1 gramme weight of radium is not less than 10- 4 calorie per second, or 3200 calories per year. It is evident that, unless energy is supplied from external sources, the substance cannot continue emitting energy at such a rate for many years, even supposing a considerable amount of energy may possibly be derived from rearrangements of the components of the molecule. In the light of the results on the amount of energy given out by radio-active substances, it is of interest to consider some speculations as to the origin of the rays, and of the supply of energy required for a continuous emission of the radiation.

*

'Roy. Soc. Proc.', vol. 66, 1900; ·Phil. Trans.', A, vol. 195.

294

The Collected Papers

0/ Lord Ruthelford

We will first briefly review the state of our present knowledge of the radioactive substances. Uranium, the first ofthe radio-active substances discovered, has been closely investigated. BECQUEREL has shown that it gives out radiation constantly from year to year, even when placed in the dark. The radioactivity is preserved in solution, and persists if the substance is recrystallised in the dark. The radiation given out is independent of the gas around it, and of the pressure of the gas, and is not much affected by considerable changes of temperature. The same radiation is given out by all the uranium compounds. The radio-activity appears to depend on the uranium molecule alone, and not what it is combined with. Pitchblende and other uranium minerals are active, and, as far as experiments have gone, continue radiating indefinitely. In considering the question of the emission of energy per unit weight of uranium, an important point arises which it is difficult to decide satisfactorily by experiment, viz., whether the radio-activity is confined to the surface or possessed by the whole mass of the substance. At first sight the radio-activity appears to be superficial, since the intensity of the radiation does not increase with increase of thickness of uranium. Such an action, however, is to be expected, even though there is volume radio-activity, since the radiation can only penetrate to the surface from a very short depth below the surface. The increase of the intensity of the radiation with increase of thickness for thin layers and the action of solutions support, as far as they go, the supposition that the activity is throughout the volume. The energy given out in the interior of the substance would most probably be dissipated as heat in the material. If the radio-active power is possessed by the whole volume, it follows from the above supposition that the mineral pitchblende must have been radiating energy since its formation as amineral. If we suppose the radiation has been going on constantly at its present rate in the course of 10,000,000 years, each gramme of uranium has radiated at least 300,000 calories. It is difficult to suppose that such a quantity of energy can be derived from regrouping of the atoms or molecular recombinations on the ordinary chemical theory. This difficulty is still further increased when we consider the emission of energy from radium, a substance 100,000 times more active than uranium. The emission of energy in that case is, at least, 3000 calories per year. If future experiments should show that radium, as weIl as uranium, gives out radiation at a constant rate from year to year, in order to account for such a rapid emission of energy, it would be necessary to suppose that the radioactive substance in some way acts as a transformer of energy. Such a supposition does not seem probable, and leads us into many difficulties. On the view, however, advanced recently by J. J. 1'HOMSON, that an atom is not simple, but composed of a large number of positively and negatively charged electrons, the possible energy to be derived from the eloser aggregation or regrouping of the components of a molecule is very much greater than on the atomic theory, as ordinarily understood. The energy required to completely dissociate a molecule into its component electrons would be

Dle/:~Y

nf Rjjll(~el1 mut Becquerel Ra.1's

295

many thousand limes greatcr than the energy required to dissociate a moleeule into Its atom~. The energy that might be derived from a greater concentration or c10seness of aggregation of the components of such a complex molecule would possibly be sufficient in the case of uranium to supply the energy for the emission of radiation for long periods of time. The sudden movements of electrons would set their charges in oscillation, and give rise to aseries of electromagnetic pulses corresponding to X rays. The remarkable property of some of the radio-active substances in naturally emitting a kind of cathode rays shows that the present views of molecular actions require alteration or extension in order to explain such phenomena. The energy that might possibly be derived from regrouping of the constituents of the atom would not, however, suffice to keep up a constant emission of energy from a strong radio-active substance, like radium, for many years. It is of importance that experiments to test the constancy of the radiations of a powerful radio-active substance, like radium, should be carried out at definite intervals. If the radiation should keep constant from year to year, it would be strong evidence that the energy of the radiation was not derived at the expense ofthe chemical energy of the radio-active substance.

Einfluss der Temperatur auf die "Emanationen" radioaktiver Substanzen von

E. RUTHERFORD

From Physikalische Zeitschrift, 2, 1901, pp. 429-31

IN einer früheren Arbeit*) habe ich gezeigt, dass Thoriumverbindungen dauernd eine Art von radioaktiven Teilchen aussenden, die ihr Strahlungsvermögen und die Fähigkeit, die Luft zu ionisieren, einige Minuten lang behalten. Diese "Emanation" (wie ich sie der Kürze halber genannt habe) vermag durch Wattepfropfen und Lösungen hindurchzugehen ohne merklichen Verlust ihres Strahlungsvermögens. Im Falle der Thoriumverbindungen nimmt die von der Emanation ausgehende Strahlung rapide ab, so dass sie in etwa einer Minute auf den halben Wert heruntersinkt. Dorn t) hat gefunden, dass sowohl Radium, als auch Polonium ähnliche Eigenschaften wie Thorium zeigen; doch nehmen die Emanationen mit verschiedener Geschwindigkeit ab. Ferner hat Dorn gezeigt, dass die Anwesenheit von Feuchtigkeit in der Luft das Emanationsvermögen von Thorium, Radium und Polonium vermehrt; am deutlichsten ist letzteres der Fall beim Radium. Verfasser hat gezeigt, dass die induzierte Aktivität beim Thorium in gewissem Grade von der Emanation abhängt, und Dorn hat dasselbe für Radium nachgewiesen. Um den Einfluss der Temperatur auf das Emanationsvermögen der radioaktiven Substanzen zu prüfen, wurde folgende Anordnung benutzt (Figur): Ertl(!

B

B

B

B B

Ertl(!

* Diese Zeitsehr., 1, 347, 1900. (This volume, p. 220.)

t Naturf.-Ges., HaHe, 1900.

t."in,tluss der Tempcratllr au./ die .,Emallationell" radioaktil'er Substall=ell 297 Die radioaktive Substanz befand sich in der Mitte einer langen Platinröhre (T) von 40 cm Länge und 0,65 cm Weite. Die Röhre befand sich in einem

Asbestofen und wurde mittels einer grossen Gebläseflamme geheizt. Ein Strom ziemlich trockner Luft aus einem Gasbehälter bewegte sich langsam durch ein Trockengefäss mit konzentrierter Schwefelsäure, dann durch einen Watte pfropf (B), um Staubteilehen und andere Kerne möglichst zu beseitigen. Jenseits der Platinröhre passierte die Luft einen Pfropfen von Glaswolle (C), um mitgeführte Staubteilehen und Ionen zu beseitigen. Dann trat die Luft mit den in ihr enthaltenen Emanationen in ein grosses cylindrisches Metallgefäss (D). Ein in dem Cylinder befindlicher isolierter Messingstab (E) war in üblicher Weise mit den Quadranten eines empfindlichen Elektrometers verbunden. Ein Pol einer Batterie von 300 Volt war mit dem Gefässe (D), der andere mit der Erde verbunden. Als ein Beispiel der erhaltenen Resultate betrachten wir zuerst die Wirkung des Thoriumoxydes. Etwa 1 g Thoriumoxyd befand sich in der Mitte der Röhre. Ein Luftstrom von etwa 2 ccm pro Sekunde wurde durch den Apparat getrieben, bis die von der Emanation herrührende Entladungsgeschwindigkeit in dem Prüfgefäss konstant war. Die Elektrometerablenkung betrug 10,6 Skalenteile pro Sekunde. Eine kleine Gasflamme bewirkte nach einiger Zeit ein schwaches Anwachsen der Ablenkung auf I 1.8 Teile p. sec. Bei Anwendung der vollen Flamme stieg die Ablenkung auf 18.4 Teile p. sec., und blieb konstant, solange die Flamme konstant blieb. Bei Erhitzung der Röhre auf Rotglut mittels des Gebläses stieg die Ablenkung rapide bis zu einem Maximum von 36 Teilen p. sec. und verminderte sich dann stetig bis auf 12 T.p. sec. Wurde darauf die Röhre zur Weissglut erhitzt, so stieg die Ablenkung wieder auf 19.5 T. p. sec. und fiel dann stetig im Laufe von IO Minuten bis auf 1.6 T. p. sec. Liess man nun den Apparat sich abkühlen, so gab die Emanation im kalten Zustand eine Wirkung von etwa 1 T. p. sec. Eine Wiederholung der Versuche am nächsten Tage ergab, dass das Emanationsvermögen sich nicht wieder erneuert hatte, und nach einer 20 Minuten langen Erhitzung auf Weissglut verminderte sich die von der Emanation herrührende Ablenkung auf 0.5 T. p. sec, Die allgemeine Wirkung der Temperatur auf Thoriumoxyd besteht also darin, dass bei Anwendung einer hohen Temperatur das Emanationsvermögen zuerst auf den dreifachen Betrag vermehrt, dann aber fast gänzlich zerstört wird. Dagegen ergab sich, dass bei Temperaturen unterhalb der Rotglut das Emanationsvermögen des Thoriumoxydes nicht nachliess. Ein über 3 Stunden dauernder Versuch bei der durch eine volle Gasflamme hervorgebrachten Temperatur zeigte, dass das Emanations-vermögen (etwa doppelt so stark, als bei 20°) während dieser Zeit konstant blieb. Nach Abkühlung auf Zimmertemperatur war der Betrag der Emanation derselbe wie zuvor. Wurde dagegen die Temperatur einmal aufhelle Rotglut gesteigert, K*

298

Tlle Collected Papers

0/ Lord Ruthelford

so war das Emanationsvermögen verloren und konnte nicht wieder hergestellt werden. Andere Thoriumverbindungen, wie z. B. das Oxalat, Nitrat und Sulphat wurden auch untersucht und ergaben dieselben allgemeinen Resultate, wie das Oxyd. Doch ist für diese Substanzen der Betrag der Emanation viel geringer als für das Oxyd. Beim Sulphat ist sehr langdauernde starke Erhitzung nötig, um einen grossen Teil des Emanationsvermögens zu zerstören. Uranoxyd giebt weder kalt noch warm eine merkliche Emanation. Emanation des Radiums: (Erhitzung hat einen äusserst gros sen Einfluss auf das Emanationsvermögen des Radiums.) Die untersuchten Radiumpräparate stammten von P. de Haen in Hannover. Eine kleine Menge der Substanz (Radium-bromid) wurde in die Platinröhre gebracht und in derselben Weise geprüft wie Thoriumoxyd. Liess man einen Luftstrom über die Substanz streichen, so nahm der Betrag der Emanation rapide ab und war in 15 Minuten auf etwa 1/3 des Anfangswertes gefallen, worauf er nahezu konstant blieb. Auf gleiche Materialmengen bezogen und am Ionisationsvermögen gemessen emittierte das Radium weniger Emanation als Thoriumoxyd. Bei Heizung mit einer kleinen Gasflamme stieg der Betrag der Emanation mit grosser Rapidität. Als der Emanationsstrom im Cylinder einen 300 mal stärkeren Wert erreicht hatte, als bei Zimmertemperatur, wurde die Flamme entfernt und der Luftstrom unterbrochen. Dann zeigte sich, dass der Strom im Prüfcylinder nur langsam im Laufe der Zeit sich veränderte, indem er in 10 Minuten um wenige Prozent fiel. Diese Verminderung entstand wahrscheinlich zum Teil durch Diffusion der Emanation in die freie Luft, durch die Öffnung des Prüfcylinders hindurch. Durch Ausblasen der Luft aus dem Cylinder fiel der Strom auf 1/20 seines Wertes, woraus ersichtlich, dass 19/20 des Stromes von der im Gefäss vorhandenen Emanation herrührte, durch die die Luft ionisiert wurde, während der Rest von in den Wänden des Gefässes induzierter Radioaktivität stammte. Das Experiment wurde sodann mit einer in halber Grösse brennenden Flamme wiederholt, die so lange brannte, bis der Strom einen ziemlich konstanten Wert erreichte, der 650 mal grösser war, als bei Zimmertemperatur. Bei Anwendung der vollen Flamme stieg der Strom auf mehr als das 1800 fache des Anfangswertes. Bei heller Rotglut der Platinröhre stieg die Wirkung auf mehr als das 5000 fache. Erhitzung auf Weissglut bewirkte keine weitere Erhöhung mehr. In diesem Stadium des Experimentes wurde die Flamme abgestellt, der Luftstrom unterbrochen und die Emanation wiederum ausgeblasen. Hierdurch fiel der Strom auf etwa 1/4 seines Wertes, woraus folgt, das 3/4 von der Emanation herrühren. Bei einer Wiederholung des Versuches am nächsten Tage mit demselben Material konnte der Strom durch Erhitzen auf Rotglut bloss bis auf das 65 fache des Anfangswertes vom vorhergehenden Tage gesteigert werden. Sodann wurde die Röhre 5 Minuten lang auf Weissglut

F:inf!lIs.\· de'I" TemperatuI" auf die .. Emanatiollen" radioaktil'el" Suhstan=ell

299

erhitzt. Nach der Abkühlung wurde' der Versuch wiederholt: bei Rotglut stieg die Emanation wieder etwa auf das 65 fache. Diese Versuche zeigen, dass die Emanation des Radiums mit zunehmender Temperatur enorm anwächst; nach einmaliger Erhitzung auf Rotglut ist jedoch diese anfängliche grosse Steigerung mit der Temperatur zum grössten Teil zerstört und kann nicht wieder hergestellt werden. Ähnliche Resultate wurden mit einem anderen, als "einfach" bezeichneten Radiumpräparat erhalten. In diesem Falle war die Zunahme des Emanationsvermögens gross, doch nicht so gross, wie bei dem ersten Präparat. Emanation des Radiums und induzierte Radioaktivität: Die Emanation des Radiums zeigt einige sehr interessante Eigenschaften. Mit Thorium und Radium bei Atmosphärendruck und Zimmertemperatur wird die induzierte Aktivität stets an der negativen Elektrode erzeugt. Wurde dagegen die Emanation durch Erhitzung erzeugt, so ergab sich, dass die induzierte Aktivität nicht auf die Kathode beschränkt war, sondern manchmal sich über heide Elektroden verbreitete, selbst bei Potentialdifferenzen von 300 Volt. Der Betrag an jeder Elektrode war veränderlich. Bei einem Versuch, bei dem der Betrag der Emanation sehr gross war, war die induzierte Aktivität ziemlich gleichmässig zwischen Anode und Kathode verteilt; bei den meisten Versuchen befand sich jedoch der grösste Teil an der Kathode. Noch eine andere merkwürdige Wirkung wurde an der Radiumemanation beobachtet. Die unter der Wirkung der vollen Gasflamme erzeugte Emanation wurde in den Prüfcylinder geleitet, bis der Strom zwischen den Elektroden konstant war. Dann wurde der Luftstrom abgestellt und alle Öffnungen verschlossen. Der Strom zwischen den Elektroden wurde von Zeit zu Zeit beobachtet. Im Laufe von 3,5 Stunden stieg der Strom auf das 1,31-fache seines Anfangswertes. Der Apparat wurde dann über Nacht stehen gelassen und bei einer Prüfung am nächsten Tage (d. h. nach 16 Stunden) fand man, dass der Strom auf das I, I-fache des Anfangswertes gesunken war. Nach nahezu 20 Stunden war der Strom noch immer grösser als der Anfangswert. In diesem Zustande wurde die Luft ausgeblasen, wodurch der Strom ungefähr auf den halben Wert reduziert wurde. Etwa die Hälfte des Stromes rührte somit von Emanation her, die sich noch im Apparate befand, die andere Hälfte von induzierter Aktivität. In diesem Falle beschränkte sich fast die ganze induzierte Strahlung auf die Innenfläche des äusseren Cylinders. Da die Potentialdifferenz fast während der ganzen 20 Stunden angelegt war, so musste die induzierte Aktivität von Diffusion an die Wände herrühren. Der Versuch zeigt, dass, wie auch Dorn beobachtete, die Emanation des Radiums viel länger wirksam bleibt, als die des Thoriums. In einer früheren Arbeit habe ich gezeigt, dass die Emanation des Thoriums in etwa einer Minute die Hälfte ihrer Wirksamkeit verliert. Um einen grossen Betrag induzierter Aktivität zu erhalten, verfuhr man vorteilhaft so, dass man aus zwei voneinander isolierten parallelen Platinplatten ein geschlossenes Gefäss

300

The Collecfed Papers of Lord RUfherford

herstellte. Die obere Platte wurde negativ geladen und das Radium auf die untere Platte gestreut. Nach einer Minute langer Erhitzung mit einer Flamme ergab sich die obere Platte stark radioaktiv. Es zeigte sich, dass diese Aktivität nicht sehr regelmässig abnahm, aber die mittlere Zeit, die zur Abnahme auf die Hälfte nötig war, war viel geringer, als es bei der induzierten Aktivität des Thoriums der Fall war. Beim Thorium sank die induzierte Aktivität auf die Hälfte in etwa 12 Stunden. Die Emanation des Radiums verliert also ihre Wirkung langsamer als die des Thoriums, während für die induzierte Aktivität das umgekehrte gilt. Eine detailliertere Prüfung der Radiumemanation soll in einer späteren Arbeit gegeben werden. Der Effekt der Temperatur auf das Emanationsvermögen des Thoriums und Radiums führt zu dem Schlusse, dass die Emanation wahrscheinlich von einem chemischen Vorgang im Material herrührt. Solange die Temperatur eine gewisse Grenze nicht überschreitet, ist das Emanationsvermögen vermehrt, und bleibt konstant, solange die Temperatur konstant bleibt. Wenn dagegen die Temperatur über einen bestimmten Wert steigt, so wird das Emanationsvermögen grossenteils zerstört und kann nicht wieder hergestellt werden. Die Beobachtung Dorns, dass das Emanationsvermögen in feuchter Luft vermehrt ist, kann auch als Stütze des obigen Schlusses dienen. Wenn die Emanation von einer anderen im Radium oder Thorium vorhandenen Substanz herrührte, so sollte man eine grosse Zunahme der Emanation bei der Temperatur erwarten, wo diese Substanz ausgetrieben wird. Tatsächlich ist der abgegebene Betrag sehr klein, wenn man ihn mit dem durch stetige Erhitzung des Thoriums oder Radiums, unterhalb der zur Zerstörung nötigen Temperatur, erhaltenen vergleicht. Zum Schlusse spreche ich für freundliche Unterstützung bei den Versuchen Herrn Dr. H. T. Barnes und Fr!. H. Brooks meinen Dank aus. McGill Universität. Montreal, 15. März 1901 (Aus dem Englischen übersetzt von W. Kaufmann) (Eingegangen 31. März 1901)

The New Gas from Radium by

Macdonald Professor of Physics, McGill University, Montreal and MISS H. T. BROOKS, M.A.

E. RUTHERFORD, M.A., D.se.,

Frorn Transactions of the Royal Society 0/ Canada, 1901, section iii, sero ii, vol. 7, pp. 21-25 (Read May 23, 1901)

IN arecent number of the Comptes Rendus, an account was given by Curie of the evidence of the existence of a new gas from radium, which possesses remarkable physical properties. A specimen of very radioactive radium was placed in a glass vessel connected with a mercury pump. On exhausting to a low vacuum and allowing the apparatus to stand, the pressure steadily increased. When the very small volume of gas thus collected flowed along the glass tubes, it rendered them phosphorescent, and if left in for some time, rapidly blackened them. The gas itself was powerfully radioactive, i.e. it continuously gave out a type of Röntgen rays, which made gases partial conductors of electricity and rapidly acted on a photographic plate. This gas preserves its radioactive power for several weeks. For some time past, one of the authors had been independently investigating one of the most remarkable properties of radioactive substances, namely, the power of continuously emitting radioactive particles of some kind. The term 'emanation' was applied to the substance thus emitted, as there was no evidence at the time whether the material emission was a vapour of the substance, a radioactive gas, or partic1es of matter each containing a large number of molecules. The 'emanation' from thorium compounds was shown to retain its radioactivity for several minutes and possessed the remarkable property of causing every substance in the neighbourhood of the thorium to become itself radioactive for several days. The specimens of impure radium then in the possession of the author, did not possess the power of emitting such an emanation; but Dorn, using a later and more active preparation of radium, showed that it possessed the same emanating power as thorium. One of the most interesting properties of excited radioactivity is that it can be concentrated in an electric field on the kathode, so that a very fine wire of any metal can be made to act like a powerfully radioactive substance for several days. A short time ago, one of the authors published an account of the effect of temperature on the emanating power of radioactive substances, in the

302

The Collected Papers ofLord Ruthel:{ord

Physikalische Zeitschrift. In the paper it was shown that the emanating power of thorium increased with rise of temperature to about a red heat, but on heating to a white heat the emanating power was destroyed and could not be recovered. An examination of a specimen of radium obtained from De Haen, Hanover, showed that the effect of temperature on its emanating power was very large. When the substance was heated below a red heat, its emanating power increased over 10,000 times, but was to a large extent destroyed by heating to a higher temperature. The emanation, obtained by passing a slow current of air over heated radium, was found to preserve its radioactive powers for weeks, when kept in a c10sed metal vessel. The radioactivity slowly decayed, but was still quite appreciable after a month's interval. The question now arose, if any physical experiments could be devised to settle the problem as to whether the emanation was in reality a radioactive gas, driven off from the substance, or a vapour of the substance, or a material emission of particles much larger than molecules. Experiments on thorium showed that no appreciable volume of gas could be collected by leaving thorium oxide in a vacuum tube connected with a mercury pump. No new lines were observed in the spectrum of the gas. The amount of the emanation thus given off was thus too small to detect by its volume in this way, but the electrical conductivity produced by the emanation in the gas, with which it is mixed, is often very large and can be used as a measure of the amount of emanation present. The emanation gives out a type of radiation which ionizes the surrounding gas. When a strong electric field is applied, the current through the gas reaches a maximum value, and is then a measure of the total number of ions produced per second, multiplied by the charge on the ion. By determining the rate of diffusion of the emanation into air or other gases, using the electrical method, it is possible to obtain an approximate estimate of its molecular weight. The coefficient of interdiffusion of most of the simple gases have long been known, and the results show the coefficient of diffusion of one gas into another is inversely proportional to the square root of the product of the molecular weights. lf therefore the coefficients of diffusion of the emanation into air is found to have a value lying between that of two gases A and B, we can conc1ude that the molecular weight of the emanation lies between the molecular weights of A and B. The apparatus employed was similar to that used by Loschmidt 1 in his experiments on the coefficient of interdiffusion of gases in the year 1871. Fig. (1) shows the general arrangement. A long brass cylinder AB 6 cms. in diameter, 73 cms. long, was divided into two equal parts by a movable metal slide S. The ends of the cylinder were c10sed with ebonite stoppers. Two insulated brass rods a and b, each half the length of the tube, passed through the ebonite stoppers and were supported centrally in the tube. The cylinder was insulated and connected to one pole of a battery of 300 volts, 1 Wiener Akad., 1871.

Thc

,I\'CII'

Gas/rom Radium

303

thc othcr pole 01' which was to carth. Thc central rods could be connected to a sensitive quadrant electrometer, The cylinder was covered with a thick layer of feit, and placed inside a metal box filled with cotton wooI, in order to keep temperature conditions as steady as possible. In order to carry a sufficient quantity of emanation into the half cylinder A, it was necessary to slightly heat the radium. The slide S was c10sed and the side tubes opened. A slow current of dry air from agas bag, passed through a platin um tube, in which a small quantity of a radium compound was placed. The emanation was carried with the air into the cylinder A. When a sufficient quantity had been introduced, as tested by the electrometer, the current of air was stopped. The side tubes were c10sed by fine capillary

earth

elect 'b

Tram,

gQ.Someter

B

s

A

elect

a

gQ.Someter gQ.Someter

gQ.Someter

Fig. 1 tubes. These prevented any appreciable loss of gas due to diffusion, but served to keep pressure of gas inside A at pressure of outside air. The three entrance tubes into the cylinder, shown in the figure, were for the purpose of initially mixing the emanation and gas as uniformIy as possible. After standing for several hours to make temperature conditions steady, the slide was opened, and the emanation began to diffuse into the tube B. The current through the tubes A and B was measured by an electrometer, with suitable capacity in parallel, at regular intervals. Initially there is no current in B, but after the opening of the slide, the amount in A decreased and the amount in B steadily increased. After several hours the amount in each half is nearly the same, showing that the emanation is nearIy uniformly diffused throughout the cylinder. It can be readily shown that if K = coefficient of diffusion of the emanation into air. 1 -= duration of diffusion experiments in sees. (/ ---- total length of cylinder.

304 S1

Tlte Collected Papers of Lord Rutherfol'd

=

amount of emanation in tube A at end of diffusion.

S2 = amount of emanation in tube B at end of diffusion, then S1 - S2

--=----=- =

- n 2 Kt 8 rI --;2

+

- 9n2 Kt

1 --- 02 --

+

1

*

- -< e -e etc. ~ SI S2 7T 2 L 9 j See Stefan and Loschmidt, Berichte Wien. Akad. 63 1871.

+

From this equation K can be determined, if SI and S2 are known. An uncertainty however arises in estimating SI and S2 for the rate of leak in A and B is made up of the current due to emanation alone and the current produced in the gas by the excited radioactivity on the electrodes. As the amount of excited radioactivity increases with the time, the ratio of the current due to the emanation and to the excited radiation varies with the time allowed for diffusion. The ratio of the current due to the excited radiation can be determined by removing the central electrode and finding the amount of current immediately after the introduction of a new electrode. When the emanation is allowed to diffuse for half an hour, the current due to excited radioactivity was about 0·4 of the whole. The calculated value of K was found to be about 20 per cent greater when the correction for the amount of excited radioactivity was applied. The value of K deduced from the experiments was found to be between 0·08 and 0·15. All the later observations gave a value about 0·08. This variation in the value of K deduced from the experiments is not altogether due to errors of experiment, the values obtained at first with a new specimen of radium were in all cases higher than when it had been laid by for several months. It appears as if the emanation were not simple in character, and that part of the emanation first given off was of 10wer molecular weight than that emitted after several months exposure to the air. Further experiments are now being carried out to see if the radium emanation undergoes a progressive change with time. For the purpose of comparison, we will now give a few of the coefficients of interdiffusion of gases, compiled from Landolt and Börnstein's tables. Coefficient of diffusion into air

Molecular weight

Water vapour

0·198

18

Carbonic acid gas

0·142

44

Alcohol

0·101

46

Ether

0·077

74

Gas or vapour

• Editor's Footnote : In the original paper are certain misprints which have been corrected.

The lI/eil' Gas/rom Radium

305

In the above table we see that the coefficient of interdiffusion follows the inverse order of the molecular weights. In cases of the simpler gases it has been shown experimentally that the coefficient of interdiffusion is approximately inversely proportional to the square root of the product of the molecular weight. If we apply these considerations to the emanations (K = 0·08 to 0,15) we see that it is agas or a vapour of molecular weight (allowing a wide margin) probably lying between 40 and 100. These numbers exclude the possibility of the substance being a vapour of radium, for it has already been shown by M. and Mme. Curie that the atomic weight of radium is greater than that of barium. We must therefore conclude that the emanation is in reality a heavy radioactive vapour or gas. On account of the rapid decay of the radiating power of thorium emanations, it is not possible to determine its coefficients of diffusion in the same way; but special experiments show that it diffuses rapidly, and is also probably gaseous in character. The physical properties of these emanations or gases are most remarkable. The radium emanation not only continues for long intervals to be a source of radiation which is apparently similar in character to easily absorbed Röntgen rays, but in some way manufactures from itself a positively charged substance, which travels to the negative electrode and becomes a source of secondary radioactivity.

Emanations from Radio-active Substances From Nature, 64, 1901, pp. 157-8

IN arecent number of the Comptes rendus of the Paris Academy (March 25) an account appeared by MM. P. Curie and A. Debierne of the production of a radio-active gas from radium. In their experiments some radium was placed in a glass vessel and the air exhausted by means of a mercury pump. It was found that the vacuum steadily decreased, due to the giving off of a gaseous substance from the radium. A small amount of the gas thus collected was found to be strongly radio-active. It caused phosphorescence in the glass tubes over which it passed, and in course of time blackened them. Substances exposed in the gas became themselves temporarily radio-active. Some time ago (Phi!. Mag., January and February 1900) I showed that thorium compounds continuously emitted radio-active particles of some kind, which preserved their radio-activity for several minutes. This emanation possessed the remarkable property of causing all bodies, in contact with it, to become themselves radio-active. In an electric field the excited radioactivity could be concentrated and confined to the negative electrode. In this way I was able to make a fine platinum wire become a very powerful source of radiation. The excited radio-activity gradually diminished, falling to half its value in about twelve hours. The specimen of impure radium then in my possession gave out no emanation and caused no excited radio-activity. Later, Dorn, using the same methods, showed that apreparation of radium from P. de Haen, Hanover, gave out an emanation similar in properties to thorium. With a specimen of radium obtained from the same source I have found that the emanation given off is small at atmospheric temperature, but can be enormously increased by slightly heating the radium. In this way I have obtained ten thousand times the amount of emanation given off at ordinary temperatures. An account of these experiments is given in the Physikalische Zeitschrift (April 20). By passing the emanation with a current of air into a c10sed vessel, and then c10sing the openings, the emanation remains radio-active for a long time. The radio-activity decreases slowly, but is still quite appreciable after an interval of one month. M. and Mme. Curie, some time ago, stated that they had obtained a radio-active gas which preserved its activity for several weeks; this is possibly identical with the emanation. Up to this point I had been unable to obtain any definite evidence whether the so-called emanations were vapours of the radio-active substances, radioactive gases, or radiating partic1es large compared with a molecule. The

!:.illanatiollS ./i"om Rudio-uctil'e Substanc'es

307

radium and thorium, when placed in an exhausted tube, gave no appreciable lowering of the vacuum, and no new spectral Hnes could be observed. The quantity of substance emitted was too small to examine by chemical methods. Quite recently, however, some light has been thrown on the question of the nature of these emanations by examining their rate of diffusion by an electrical method. In these experiments I have been assisted by Miss H. T. Brooks, and the results point to the conclusion that the emanation from radium is in reality a radio-active gas, with a molecular weight probably lying between 40 and 100. There is one distinct feature which distinguishes the emanations from radium and thorium. The thorium emanation loses its radio-activity in a few minutes, while the excited radio-activity due to it lasts several days. The radium emanation, on the other hand, preserved its radiating power for several weeks, while the excited radio-activity due to it disappears in a few hours. In the following experiments it was only possible to experiment with radium emanation, on account of the rapid decay of radio-activity of the thorium emanation. The diffusion apparatus was similar to that wh ich had been employed by Loschmidt in 1870 in his determinations of the coefficients of interdiffusion of gases. A brass cylinder, 73 cm. Iong, 6 cm. in diameter, was divided into two equal parts bya metal slide, which could be opened or closed. The ends were closed by insulating ebonite stoppers, through which passed central rods half the length of the tube. In order to introduce the emanation into one half of the cylinder the slide was closed, and a slow current of air, which had pas~ed over slightly heated radium and thus carried the emanation with it, was passed through the cylinder. When a sufficient amount had been introduced the current of air was stopped and the openings closed. After standing for an hour or more the slide was opened, and the radio-active emanation slowly diffused into the other half of the cylinder. The amount of emanation in each half of the cylinder after any interval was tested by observing the current through the gas, when a suitable P.D. was applied, by means of an electrometer. The current is carried by the gaseous ions wh ich are continually produced by the radiation from the emanation. From these observations the coefficient of inter-diffusion of the emanation into air at atmospheric pressure and temperature can be readily deduced. The experiments are, however, complicated by the excited radio-activity on the electrodes, which must be taken into consideration. So far as the observations have gone up to the present, the coefficient of diffusion of the emanation into air has a value between 0·10 and 0'15, and probably nearer the former. Now the coefficients of inter-diffusion of some known gases and vapours into air have been determined. The following examples have been taken from Landolt and Börnstein 's tables:-

308

The Collected Papers 0/ Lord Rutherford Gas or Vapour

Coefficient of Diffusion into Air.

Water vapour Carbonie aeid gas Aleohol Ether

0·198 0·142 0·101 0·077

Molecular Weight.

18 44 46

74

In the above we see that the eoeffieients of diffusion follow the inverse order of the moleeular weights. In eases of the simpler gases it has been shown experimentally that the eoeffieient of inter-diffusion is approximately inversely proportional to the square roots of the produet of the molecular weights. If we apply these eonsiderations to the emanations we see that it is agas or a vapour of molecular weight (allowing a wide margin) probably lying between 40 and 100. These numbers exelude the possibility of the substanee being a vapour of radium, for it has already been shown by M. and Mme. Curie that the atomie weight of radium is greater than that of barium. We must, therefore, eonc1ude that the emanation is in reality a heavy radio-aetive vapour or gas. On account of the rapid decay of the radiating power of thorium emanations it is not possible to determine its eoefficients of diffusion in the same way; but special experiments show that it diffuses rapidly, and is also probably gaseous in eharaeter. The physical properties ofthese emanations or gases are most remarkable. The radium emanation not only eontinues for long intervals to be a souree of radiation whieh is apparently similar in eharacter to easily absorbed Röntgen rays, but in some way manufaetures from itself a positively charged substance, which travels to the negative electrode and becomes a source of secondary radio-aetivity. Space is too short to enter into the interesting question of the possible explanation of these complicated phenomena. E. RUTHERFORD McGilI University, Montreal May 30

Dependence of the Current through Conducting Gases on the Direction of the Electric Field by E. R UTHERFORD, M.A. Macdonald Professor of Physics, McGill University, Montreal

From the Philosophical Magazine for August 1901, sero 6, ii, pp. 210-228 Communicated by Professor J. J. Thomson

IN the conduction of gases under the infl.uence of Röntgen and Becquerel rays, it has generally been considered that the magnitude ofthe current between the electrodes is independent of the direction of the electric field, except in the case of potential differences of the order of one volt, * when the contact differences of potential between the electrodes may cause inequalities in the currents. In the general case, however, of gas conduction, I have found that there is, in most cases, a difference between the values of the current with the reversal of the electric field. This difference is due to the inequality of the velocities of the positive and negative ions. It is only in certain special cases that the current is independent of the direction of the electric field. These cases are: ( I) When the ionization of the gas is symmetrical with regard to the

electrodes. (2) When the electric field is so great that the current is a maximum, i.e. when all the ions reach the electrodes before recombination occurs. (3) When the number of ions present is so small that their movement between the charged electrodes does not appreciably disturb the potential gradient. (4) When the positive and negative ions have equal velocities. In all other cases the positive and negative currents are unequal in value, the difference depending upon the distribution and intensity of the ionization, the distance apart and shape of the electrodes, and the potential difference applied. In many of the experimental arrangements of previous observers, by means of which the equality of the current in the two directions was observed, one or more of the above conditions was fulfilled . ... E. Rutherford, Phi!. Mag., January 1899.

310

The Collected Papers o[ Lord Ruther[ord

The essential conditions for obtaining unequal currents are: (1) Ionization unsymmetrical with regard to the electrodes. (2) Disturbance of the potential gradient due to the movement of the ions in the electric field. (3) Unequal velocity of the ions.

A few simple cases will now be considered, in which the difference between positive and negative currents is strongly marked. The difference between the currents in air is most readily shown when the gas is dry, and when the ionization between the electrodes is powerful and confined mainly to the surface of one electrode. This condition can readily be fulfi1led by allowing a thin stratum of strong Röntgen rays to pass between two parallel plate electrodes, but nearer one plate than the other, or by using a powerful radioactive substance, like radium, where the ionizing power due to the radiation is mainly confined to within a few centimetres of the active surface. Some experiments with Röntgen rays will be given later, but, on account of the inconstancy of a Röntgen tube, it was found more convenient and accurate to experiment with radioactive substances. Fig. 1 shows the general arrangement of the experiment with radioactive substances. Two parallel circular insulated plates oflead A and B, 19 cm. in diameter, were fixed horizontally inside a tin vessel C. The lower plate was covered with a layer of tin foil on which a thin layer of radium was lightly sprinkled. The upper plate was also covered with tin foH and rigidly attached to the lid of the vessel. The central portion D of the upper plate was separated from the outer

Enl:l7

A

D

c

W

A

[arM Fig. 1

Ec1rtit

Depel/dellcl' of fhe ('urrelll fhrough ("olldu('ting Gasl'!.

311

by a narrow air gap and insulated from it, so that the outer served as a guard ring. The electrometer was connected to the central plate in the usual manner; the guard ring and vessel were connected to earth; the lower plate to one terminal of a large battery of small accumulators, the other terminal of which was to earth. The crossed lines in the figure represent insulators. The current between the plates was measured by noting the deflection of the electrometer needle in a given time. The quadrants were separated, and the connection with the plate D broken after adefinite interval by two keys operated at a distance. The following results were obtained on the difference between the positive and negative currents with variation of potential difference between the plates. Diameter of centre plate D 7 cm. Air dried over P20 S for one day. Current P.D. between Ratio between Currents Plates in volts Lower Plate + Lower Plate -

6 12 26 53 104 157 211 310 460 610

I 2·1 4·9 12·5 31·7 54·7 84·7 130 171 190

1·10 2·5 6·2 15·8 41·7 70·4 109 157 190 206

1·10 1·20 1·26 1·27 1·32 1·29 1·29 1·21 1· 11 1·08

Each division of arbitrary scale is 2·3 x 10- 12 amperes. The results are plotted graphically in Fig. la, where the ordinates represent current and abscissae volts. It will be observed that the current when the 10wer plate is negative is always greater than when it is positive. Tbe maximum ratio observed is 1·32 for 104 volts. Another peculiarity is that the current increases much faster than the potential difference. For an increase of P.D. of four times from 26 to 104 volts, the currents increase 6·7 times and 6·5 times respectively. When the plates were brought closer together, the same general results were obtained for low voltages; but the currents approached to approximate equality sooner for the high voltages. For example, with the plates 3 cm. apart, the ratio of currents was 1· 20 for 27 volts, but for 300 volts the currents were nearly equal. In this case most of the ions are produced near the lower plate, and are then separated by the electric field. In consequence of this, when the current is passing, nearly all the ions near the upper plate are of one sign. A special experiment, after the manner described in a previous paper,'"

*

E. Rutherford, Phi!. Mag., January 1899.

312

The Collected Papers 01 Lord Rutherford

showed that the intensity of the ionization rapidly diminished from the surface of the radioactive plate. Using a thin layer of radium chloride, the intensity of the radiation was found to be about one-tenth of its value at the surface after passing through a distance of air of 3 cm., and diminished approximately in G.P. with the distance. 200 175

ISO

125, ... ISO

...CI!:~

~

u 75

75

251 251251

P,D.IIN VDL TI 300

ISO

450

600

Fig. la

Potential Gradient between Plates

J. J. Thomson* has worked out the mathematical theory of gas conduction,

due to distributed ions, under certain special conditions. He has shown that in the case of uniform ionization, the equation is not integrable in the practicable case when the ions move with different velocities. The present case is still more complicated than the one considered; for, in addition, the rate of production of ions is not uniform between the plates. Some light is, however, thrown on the question by examining the potential gradient between the plates when a steady P.D. is applied.

* Phil.

Mag., March 1899.

Dl'{Jelldel1c(, of tlle Cur/'l'111 t/rrougll Conducting Gases

313

In the present case, where the plate on which the radioactive substance was spread was horizontal, it was not convenient to use a water- or mercurydropper to determine the potential at any point between the plates. Recourse was had to the method used by Zeleny, in which an insulated wire connected to the electrometer takes up the potential of the gas in which it is placed. The wire W, shown in Fig. 1, was bent into a circ1e in a horizontal plane, and connected to a vertical rod inside an insulated glass tube, so that the potential to which the wire was raised was not in any way infiuenced by the vertical support. The wire and all the metal surfaces of the vessel were of tin, and in this way contact differences of potential were to a large extent avoided. The wire takes up the potential rapidly at a point where positive and negative ions are plentiful, but more slowly at points ne ar the upper plate where the ions are nearly all of one sign, and not nearly so numerous. Unless special precautions are taken, there is a danger that near the upper plate the average potential may not be indicated. Suppose, for example, in the present case, that the lower plate is positively charged and the upper plate earthed. The ions near the upper plate are nearly all of positive sign, and the wire will continue to receive a positive charge and rise in potential until it has reached the potential corresponding to the gas in the neighbourhood of the wire where the potential is highest. Since there are no negative ions present, any accidental excess of charge on the wire cannot be got rid of, and the indicated potential may be too high. A mercury- or water-dropper has the advantage that it takes up the potential at a point rapidly whether the ions of one or both signs are present. The wire, however, gives accurate results if care is taken, and is often more convenient to use than the mercury- or waterdropper.*

uM

R R

firM Fig.2 • This point has been discussed in a recent paper by C. D. Child, Phys. Rev., March 1901.

314

The Collected Papers oJ Lord RutherJord

In determining the potential gradient the method shown in Fig. 2 was employed. The insulated battery of small accumulators was connected to the terminals of a Thomson-Varley slide R of 100,000 ohms resistance. By means of the movable earth connection E the potential difference at the terminals of the resistance was divided in such a ratio that the potential ofthe air was zero elose to the potential wire. When this is the case, the electrometer shows no defleetion on separating the quadrants. This method was found in practice to be rapid and accurate. The following are results obtained for different voltages: Distance between plates 10·5 cms. Upper plate charged, lower plate earthed. Distance in cm. from Lower Upper Plate Upper plate Upper Plate Upper Plate 214107 volts + 107 voltsPlate 214 +

1 2 3

4

5

6

7

8 9

1·55

0·91

6·01 9·79 16·8 27·0

7·11 15·3

3·32

41·2 59·4

76·5

2·14 4·20

25·1

40·1

58·3

76·0

4·50 9·03 16·1

26·2 44·4

67·8 99·5 131

167

2·44

5·84 12·0

22·3

38·6

63·1

97·3 127 164

The ratio of the two currents through the gas was 1· 27 for 107 volts, and 1· 20 for 214 volts. Fig. 2a shows the results graphically. It will be observed that the potential gradient, near the radioactive surface, is very small compared with that near the upper plate, and that, near the upper plate, the potential gradients are very nearly the same whether the top plate is positive or negative. The potential gradient near the lower plate is considerably greater when the upper plate is positive than when it is negative. The results given in the table for 107 volts show that the rate of change of potential per centimetre from about 6 cm. upwards from the lower plate is numerica1ly very nearly the same for both directions of the electric field. If tbis holds accurately, it can be readily shown that the currents through the gas in the two directions for the same P.D. are direct1y proportional to the ratio of the velocities of the positive and negative ions. For let Vbe the P.D. between the plates. Let us suppose that the lower plate is positively charged, and that positive ions only are present near the upper plate. Let nt be the number of + ions per cubic centimetre at a distance x from lower plate.

f)ependellce

0/ fhe Currelll through Conducting Gases

315

200

150

150

o

VOLTS P.D.BflWEEN PL.ATES

150

R

2

3

4

/)/STANCP 8FTWEGN Pl.ATGS 5 , 7

9

10

Fig. 2a Let K 1 K2 be the velocities of the + and - ions for unit potential gradient. The current i l per square centimetre through the gas is everywhere the same between the plates, and near the upper plate is given by dV

i l = K l l1l e -d x'

where e is thc charge on an ion, and ddV the potential gradient.

x

By Poisson's equation, d2V

dx 2

=

41Tnle.

We therefore have .

K 1 dV d 2 V 41T dx dx '

'1=-'-'2

316

The Collected Papers of Lord Rutherford

In exactly the same way it may be shown that i 2 , the current when the lower plate is negatively charged, is given by .

K2 dV d 2 v

'2 - -dx ' - 'dx2' - 47T Since the potential gradient near the upper plate is the same for both directions of the electric field, dV d2 V dx and dx2 are numerically the same for both currents at the same point. Therefore i1 K 1 t;. = K2' or the currents are directly proportional to the velocities of the ions. Since the current is greater when the lower plate is negatively charged, these results are in agreement with the fact, first observed by Zeleny, that the negative ion moves faster than the positive. In the experimental case, there is always a small number of ions produced by the radioactive substance near the upper plate. For small voltages, the current near the upper plate is then not altogether carried by ions of one sign. The general effect of this is to bring the currents in the two directions more nearly to equality. When high voltages are applied, the current must also tend to equality, since the maximum current through the gas for a P.D. sufficient to remove all the ions to the electrodes, before appreciable recombination takes place, depends only on the number of ions present and not on their velocity. We should thus expect in the present case to obtain a maximum ratio of the currents for a certain P.D., and this ratio to diminish when the P.D. was raised or lowered. The potential gradient, near the radioactive surface, is much greater with the lower plate negative than positive. The reason of this is easily seen. Since the current through the gas is greater when the lower plate is negative, more positive ions have to reach the lower plate than negative ions when the lower plate is positive. The positive ions also move more slowly than the negative in the dry gas, and both causes assist in requiring a steeper potential slope to force the number of positive ions, necessary to carry the current, towards the lower plate. Effect 01 Moisture and Vapours on the Ratio 01 the Currents

Zeleny* has directly measured the velocity of ions produced by Röntgen rays, and found that the ratio of the velocity of the negative to that of the positive ion for dry air is I· 375, and for moist air 1· 10. Townsend, t by comparing

* Trans.

Roy. Soc., 1900.

t

Ibid., 1900.

Dependl'I1(,(, o{ file Cllrrl'11I fhrough Condllcting

Gases

317

the rates of diffusion of ions, determines the ratio of the velocity for dry air to be 1· 54 and 1·09 for moist air. If the ratio of the currents in the experimental case considered depends on the ratio of the velocities of the ions, we should thus expect to obtain wide differences between the ratios of the current when the air is dry or moist. This is fully confirmed by the experiments. The maximum ratio of the currents observed for air which had been left over P20S for a week was 1·45. The ratio observed for air standing over water at a temperature of 18° C. was 1·12. For intermediate stages of dryness the ratio of the currents took intermediate values. The effect of alcohol-vapour on the ratio of the currents is more marked than in the case of water-vapour. With fairly dry air in the apparatus, the ratio observed for the currents was 1· 37. On introducing alcohol through a side tube into the closed apparatus, the ratio ofthe current with the lower plate negative rapidly diminished, while the current in the other direction was not appreciably affected. After a certain stage, however. the currents in both directions decreased. The ratio observed for air saturated with alcohol-vapour at 18° C. was 1·04. We thus see that, since the currents are proportional to the velocity of the ions, the negative ion at first moves more slowly with the addition of alcoholvapour, while the velocity of the positive ion is unaffected. This is probably due to the condensation of alcohol round the negative ion. The final slow decrease of the currents in both directions is probably due to the disturbance of the current of ionization at various points, and alteration of the velocity of the ions in consequence of moving through air mixed with alcohol-vapour, rat her than to a condensation of alcohol on both positive and negative ions. If the air was only partially dry, the effect of alcohol-vapour was not nearly so marked. The ratio of the currents for air with some moisture present was observed to be 1·19. On addition of aIcohoI-vapour the ratio fell to 1·12; while if the air had been completely dried, it would have fallen to 1,04, or even less. lt thus appears that if the two vapours are present, the water condenses on the negative ion more readily than a1cohol. Since alcohol has a smaller value of surface tension than water, it was thought that possibly surface tension determined the amount of condensation in the negative ion. For this reason the vapour of ether was tried, which has a smaller surface tension than alcohol. Ether-vapour, however, was found to occupy an intermediate position between alcohol- and water-vapour. The reduction observed was from 1· 35 for fairly dry air, to 1· 23 for air saturated with ether-vapour at 18° C. The introduction of the vapour of methyliodide reduced the ratio of velo cities from 1·37 to 1· 11. These results can only be considered as approximate; but they serve to show in a simple manner the effect of these substances in altering the velocity of the negative ion.

3tH

Tlte Collected Papers 0/ Lord RlIflter/ord

It is intended to continue the investigation of the effect of vapours and other agents in affecting the velocity of the ions by this method, which is very simple and convienent. To account for the slow velocity of the ions through the gas, I have shown that, on the principle of the kinetic theory of gases, * the ion behaved as if a cluster of molecules travelled with it. This result has been confirmed by later investigations. t It has also been suggested that the slowness of motion through the gas might not be due to the size of the ions, but possibly to the fact that a small charged body moving through agas would experience a greater resistance than an uncharged body of the same size. It does not seem, however, that this explanation will suffice. On introducing water- or alcohol-vapour into the dry gas, the velocity of the negative ion decreases by gradual and not sudden stages. If the ion were molecular in size, the addition of a molecule of water-vapour would greatly alter its velocity. The slow and gradual decrease ofthe velocity ofthe negative ion with the addition of water- or a1cohol-vapour, seems to point to the conc1usion that the ion on which it is condensing is already large, unless it is assumed that only a fraction of the number of ions is affected by the presence of the vapour. The addition of a molecu1e of water-vapour to an ion of the same dimensions as a molecule would greatly lower its velocity , while it would make only a slight difference to an ion already much larger in comparison. There seems to be considerable evidence that the ion, somewhere between the pressure of an atmosphere and about 1 mm. of mercury, alters greatly in size. I have shown that the velocity of the negative carrier produced by ultraviolet light is about the same as that of an ion produced by x-rays; but the results of J. J. Thomson, Lenard, and others show that it behaves at low pressures as ifit were identical with the cathode-ray carrier, which is apparently much smaller than an atom. Townsendt has also shown that the negative ion produced by Röntgen rays at apressure of about 1 mrn. of rnercury is very srnall compared with a molecule.

Experiments with Röntgen Rays

Fig. 3 shows the general arrangement. A thin stratum of Röntgen rays was passed near the surface of the plate A but not touching it. The current between the vertical plates A and B was measured in the usual manner by an electrometer with suitable capacity in parallel with it. The plates C and D served as guard plates for the plate B in order to produce a uniform electric field between A and B. In order to keep a check on the constancy of the rays, the current between two standard plates, between which the rays passed, was observed at the same time. • Phi!. Mag. [5], xliv, p. 422. t Townsend, Trans. Roy. Soc., 1900 ; Lenard, Drude, Ann., iii, p. 298. :I: Phi!. Mag., February 1901.

/)cjJcndcllcc (?( Ifte ('/lrrCnl 1ftrough ('onducling Gases

J 19

The following results were obtained when a strong stratum of rays passed through the fairly dry air of a Canadian winter.

Earth

Earth

c

D

B

Earth

A

Earth Fig.3 Platcs 10 cm. apart. Mean width of stratum 1 cm. P.D. in volts

32·5 58 115

Current Plate A

0·63 1·84 5·7

+

Plate A-

0·82 2·35

6·4

Ratio of the Currents

1·30 1·28 1·12

Further data of this experiment are given in a later calculation (p. 323). The same general results are here observed as in the case of ionization produced by a radioactive substance. The current with the plate A negative is greater than with the plate A positive. It will also be observed that the current increases very much more rapidly than the P.D. For an increase of the voltage of 3·5 times the current with A positive increases nearly nine times, and with the plate A negative nearly eight times, i.e. the current is more nearly proportional to the square of the P.D. than directly proportional to it. A general explanation ofthese results can readily be given when the movement of charged ions between the plates is considered. Imagine a thin stratum of rays producing very powerful ionization ne ar the surface of the plate A. Suppose the plate A is positive and charged to a potential V, and the plate B to earth.

320

The Collected Papers of Lord Rutherforc/

Let d = distance between the plates; at all points between the parallel plates a short distance from the surface the ions are positive in sign. Using the same notation as on pp. 314/5, the current i l per sq. cm. through the gas is given by dV i l = Klnle dx' where nl is the number of + ions per cubic centimetre. Also d2 V

dx2 = 47Tnle.

From these two equations we obtain dV d 2 V 47Til dx' dx 2 = K l '

If xis measured from the plate A, 2 ( -dV)2 =87TilX -+A ' dx Kl

where A is the value of

(~;) at the surface ofthe plate A. Now the value of A

can be made as small as we please, by making the ionization very intense near the surface of the plate, since the potential gradient, which will remove a sufficient number of positive ions from the ionized layer to carry the current, decreases with increase of the intensity of ionization. Let Then

S7Til -K 1 =a•

dV dx = (ax

+ A2)!.

Integrating between the limits 0 and d, 2 V = -[(ad + A2)f - A3] 3a

= 3'2 a!d t

[( 1 + ad A2)A:.I

(A2)A] ad :.I •

-

Since A2 is very small compared with ad, V = ja!dt , or and

V

=

i. K1 =

~ J87Til d t 3 K1 9V2 327Td3'

'

APPARATUS FOR STUDYING THE IONIZATION PRODUCED BY RÖNTGEN AND BECQUEREL RAYS This apparatus may have been used originally at Cambridge.

Below: Same apparatus, showing insulated disc on left, mounted on insulating bridge and surrounded by a large guard ring. See: 'Dependence of the Current through Conducting Gases on the Direction of the Electric Field', page 310.

This page intentionally left blank

J)f'I' 00 :c 0

80 I-~ 60

40 20

130

140

15e-

AMPERES

160

Fig.4 F or plates 3 cms. apart a very distinct difference is observed. In Fig. 5* the results are given for P.D. ranging from 53 to 300 volts. From - 53 and to - 102 volts, the current reaches a maximum and then decreases with rise of temperature, and, in this respect resembles curves for the positive discharge. For 127 volts P.D., the current is nearly constant at its higher portions, while from 153 to 300 volts the current is found to increase continuously and rapidly with increase of temperature. The temperature of the platinum plate has not been pushed high enough to obtain sufficient ionization at the surface for the current, due to negative ions, to vary as the square of the P.D. For this reason no attempt has been made to compare the velocity of the positive and negative ions.

*

Editor's Footnote : Correclion: On bottom curve of Fig. 5 read - 53V instead of - 35V.

The Collected Papers

342

0/ Lord RU/her/orel

200 P ATES 3 OMS. J PART

80

60

:;

40

~ 20 b.

o

Z 0100

~ ..J

l:iso

c

60

40 20

-85 V -

140

150

160

AMPERES

Fig.5

Temperature Gradient Between Plates In the experimental arrangement the plates were vertical and exposed freely to the air on all sides. The average temperature of the gas between the plates is shown in Fig. 6, for distances of 2 and 3 cms., when a current of 124 amps. was passing through the plates, which corresponded to a bright red heat. The temperatures were determined by a platinum rhodium thermocouple, with a special apparatus, designed and calibrated by H. M. Tory, M.A., of McGill University, for the measurement of high temperatures. It was observed that the platinum plate was appreciably cooled by conduction at the point of contact of the junction with its surface. For this reason the indicated temperature of the plate was much too low. For a distance of 2 cms. between the plates, the indicated temperature of the platinum plate was 991 C., for a current of 134 amps., and was probably at least 1000 C. too low. The temperature fell extremely rapidly near the surface of the platinum plate, became approximately constant for some distance midway between the plates, and increased slightly near the copper plate. This is to be expected since the copper plate was heated by radiation to a higher temperature than the surrounding air. The lowest temperature 0

Oisrhargc

(~( Electricity froJII

Glml'ing Platinul11 amlthe Velodty of Ions 343

observed for the air between the plates was 169" C. for 3 ems. apart and 281 0 C. for 2 ems. With inerease of distanee the average temperature between plates steadily deereases. 1000 'C

800 IIJ

u

I-

""-'

~ 600

'"w

~ t-

Ul

m er C

() 400

.,: :::E

PART

Ul

2 CMS. A

I-

o

ART

S. AP

2 CM

200

1 OISTANCE IN eMS, 2

3

Fig.6 Velocity o[ the Ions

From the theory diseussed in the beginning of the paper, the velocity of the positive ion is given by 321Tid3 K= 9iJ2' At first sight this equation offers a simple and fairly accurate method of determination of the velocity of the ions, since i, the limiting current (which is measured by a galvanometer or eleetrometer), is independent of the intensity of the ionization at the surfaee of the platinum plate, and the quantities v and d ean be aecurately measured. The eurrent has been shown to vary very approximately as the square of the P.D., so that the veloeity dedueed is independent of the voltage employed. The following table is an example of the determination of velocities for distanees between the pI at es varying from 2 to 7 ems., by means of the theory. Area of eaeh plate surrounded by guard ring, 96 sq. ems.

344

The Collected Papers ofLord Rutherford

Distance Between Plates

P.D. in Volts

2 3 3 4 4 5 5 7

52 103 155 155 256 255 306 282

Deflection of Galvanometer

Calculated Velocities for 1 Volt per cm.

103 68 155 53

5 . 5 ems. per sec. 3·17 " 3·19 " 2· 59 " 2·69 " 2·20 " 2·20 " 1·90 "

ISO

63 90 24

Current through platinum plate about 110 amps. 1 scale division of galvanometer = 3.50 X 10- 10 amps. = 1'05 E. S. units per sec.

The eurrent observed, except for distanee of 2 ems., was a maximum. In the above table the value of i, the eurrent per sq. em., is equal to the defleetion divided by 192, sinee the eurrent was measured to two plates, 96 sq. ems. in area, at equal distanees on eaeh side of the platinum plate. In order to ealeulate the velocity, i and V are expressed in eleetrostatie units, and the value K is then given for a potential gradient of 1 E. S. unit per seeond. The velocity for 1 volt per em. is obtained by dividing this value by 300. It will be observed that the ealeulated value of the velocity is greatest when the plates are elosest together and graduaUy diminishes with the distance. For a distance of 2 ems. between the plates, the gas through whieh the ions travel is at a mueh higher average temperature than for a distance, say, of 7 ems. A very important question now arises whether we are justified in assuming that the ealeulated value of the veloeities of the ions given above represents even approximately the average velocity of the ions through the gas. The theory is based on the supposition that the ions are aU of the same kind and size and that the velocity is unaltered in passing from one plate to the other. These eonditions are, however, widely departed from in praetiee. Sinee the temperature of the gas between the plates is greatest near the platinum surfaee and gradually falls off to the copper plate, a given ion will deerease in velocity, for a given potential gradient, as it passes from the hot to the cold plate. The experiments of Mac1elland * also show the velocity of the ions drawn from flame gases diminish rapidly as the temperature of the gas, through whieh they move, deereases. This decrease of velocity is mueh greater than ean be accounted for by the lowering of the temperature of the gas and seems to be due to an aetual inerease of size of the ion. It • Proc. Roy. Soc., 1899.

Dis('JulIge

(d

IJel'trid/y fl'ol1/ (jIOll'il1g Platil1l1l11 ami/he Velo city of Ions

345

is probable that the ions from a heated platinum surface will also show similar beha\'ior. It will also he shown later that the velocity of the ions at a given point of the gas is not a constant, but varies between fairly wide limits. We have also seen that some very slow moving carriers are present between the plates, the effect of which is in some cases to greatly reduce the current. Taking all these divergences between the actual and theoretical conditions into consideration, it cannot be assumed without a special investigation that the velocities of the ions tabulated from the equation represent even approximately the true values.

Direct Determination 0/ Velocity 0/ the Ions Tnorder to settle the amount of agreement between the calculated and actual velocities, a direct method of measuring the velocity of the ions was employed. The method is identical in principle with that previously employed by the author· to determine the velocity of the negative ion produced by ultraviolet light falling on a negatively electrified surface. A direct P.D. was changed to an alternating P.D. of known frequency by means of a suitable revolving commutator. This alternating P.D. was applied to the platinum plate. Suppose the current through the platinum plate is large enough to produce a large supply of positive ions, but not large enough to discharge negative electricity. The instant the platinum plate is charged positively, the positive ions start to travel towards the copper plate. If the sign of the platinum plate is reversed before the leading ions reach the negative electrode, the positive ions travel back again and no current passes through the galvanometer. The ions are thus alternately advancing and retreating during each successive change of the electric field. Let d be the smallest distance between the plates such that for a P.D., "0' no ions reach the copper plate. The electric force between the plates

(assumed uniform) is

'1.' --J and the distance d passed over in the time T of a

half alternation is given by d = ~ K,T, where K is the velocity of the ion per unit potential gradient; therefore

d2

K--

- voT'

In the special case we are considering, the electric field is not uniform. A special series of experiments showed that tbe potential gradient between the plates was very approximately the same as that deduced from theory.

*

Proc. Camb. Phil. Soc. 1898.

346

The Collected Papers of Lord Ru/herfort!

We therefore have dv

dx

=

-

a!x!,

and Vo =

~a!di 2 ,

-

therefore

vox!

d?!

-dx -~-­ -- 2 dt· The time dt taken to pass over a distance dx is given by

dx di dt=-d =!~.dx. v dx

.n.t'oX2

K-

Therefore, if in time T of a half alternation, the ions can just travel over a distance d, by integration we obtain T

=

"3

d2 Kvo

4

d2__

4_

or

K="3 voT'

or for a given P.D. between the plates, the time taken for an ion to travel between them is l' that taken if the field were uniform. In experiments on the velocity of the negative ions produced by ultraviolet light, the alternating E.M.F. of the Cambridge town mains was employed. It is, however, much more satisfactory to commute a direct E.M.F., as in that case the period of commutation is under control and the P.D. is constant during each half alternation. The commutator employed was a two part one of diameter 6· 3 cms., driven by an electric motor. The insulation gap between each segment was O' 5 cm. so that the P.D. was applied during 0·95 of the time of alternation. The speed of rotation of the commutator was determined by means of an apparatus devised by Prof. R. B. Owens, of McGill University. The armature of a smal1 magneto-machine rotated on a continuation of the axis of the commutator. The E.M.F. generated, which was measured by a Weston voltmeter, was direct1y proportional to the speed. This apparatus is very advantageous for reading speeds accurately at a distance. In the practical determination of the velocities, it was found convenient to keep the plates at a fixed distance apart and then decrease the P.D. until the point of no appreciable current through the galvanometer for a known speed of the commutator. Theoretically, if the ions al1 have a uniform velocity, the P.D. for which no ions reach the plate should be sharply

Discharge

0./ ElectricityJrom GIOIring Platillul1l (md tlle VelocitJ'

of 10l1s 347

defined. This was observed to be the case in the measurement of the negative ions produced by ultra-violet light, but in the first experiments it was found that the P.D. could not be adjusted accurately, but that a slight current passed through the galvanometer over a considerable range of voltage. This result is not surprising when we consider that the ions are probably growing in size with time and thus move more slowly on their return journey. Some of the ions, therefore, will be unable to return to the platinum plate in the time of a half alternation. If this process continues, in the course of a large number of alternations, these ions will reach the copper plate and a current will be indicated on the galvanometer with a lower voltage than is actually necessary to carry over the original ion from the platinum to the copper plate. This difficulty, however, can be overcome very simply. In series with the alternating E.M.F. ± E a battery of accumulators is placed of E.M.F. EI less than E. The P.D. acting for one-half alternation is E + EI and for the other half E - EI' If the platinum plate is charged positively to a P.D. . .IOn moves In . an eI ' fi eld approxlmate ' ]y E EI E - E" t he returmng ectnc E+ _ EI times that for the outgoing ion. By adjustment of the values of E and E" the value E - EI can be made as small as we please compared with E + Eh so that all the ions which set out during the one-half alternation are swept back during the next half. Under these conditions it was found that the value of the P.D. necessary to give no current could be adjusted without certain well-defined limits. The following is an examp]e of the determination of velocity of ions for plates 3 cms. apart, with a current through platinum plate of 105 amps. In this case, the deftection of the galvanometer was too small to observe with accuracy, so that a one-third microfarad condenser, in parallel with the copper plate, was charged up for 30 seconds and then discharged through the galvanometer. P.D.-!-

p.n. -

46 volts 48·6 56 65

143 volts 146 153 162 173

77

Deflection

0 1· 5 8

18 60

Time

30" 30" 30" 30" 30"

Commutator made 750 revolutions per minute. Time of application of P.D. was 0·0380 second, correcting for insulation distance between the segments.

348

The Collecled Papers ofLord Rutherford

Taking 46 volts as the minimum positive potential for no current and substituting in the formula 2 - 4 d K -'3 VoT we obtain K = 7·9 cms. per second as the maximum velocity of the ions between plates 3 cms. apart. It will be observed from the above table that the p.n. sweeping back the ions to the platinum plate is between 2 and 3 times as great as the p.n. which moves them from the platinum plate. In this way it is assured that all the ions which do not reach the copper plate are swept back to the platinum surface during the succeeding half alternation. Now the steady current for + 43 volts gave a throw in the galvanometer of 156 divisions when the eondenser was charged for 30 seconds. This would correspond to adefleetion 1- x 0·95 x 156 = 74 if all the ions starting out from the platinum plate du ring each half alternation reaehed the copper electrode. Assuming the current proportional to the square of the p.n., it can be readily ealculated that for the voltages 48· 6, 56, 65, 77, the ratio of the number of ions starting from the platinum plate which reach the copper plate, is 0,015, 0,07, 0,13, 0,28, respectively. Taking as data that for 77 volts, 0·28 of the number of ions reach the copper plate, a rough average of the velocity of the ions can be deduced. F or 0·28 of the time O· 038 seeond of a half alternation, ions reach the copper plate. The time taken for ions to eross between the plates is thus 0·72 x 0·038 second, i.e., O· 027 second. The average velocity is thus 5·8 ems. per second. In a similar way it was found that the minimum voltage to carry over ions for a distance of 2 cms. for 1,000 revolutions per minute of the commutator, was 14· 5 volts. The maximum velo city for this distanee, 2 ems., is thus 13 cms. per second. This result ean, however, be eonsidered only as a rough approximation on account of the uncertainty of determining the actual positive potential of different portions of the platinum plate. It obviously varies from one end to the other on account of the P.D. sending the heating eurrent through it. The following method was employed to determine the average velocity of the ions between the plates. A steady deflection corresponding to V volts between the plates was observed. The commutator was then slowly rotated and the deflection fell slightly below one-half its value. Theoretically it should fall to ! x 0·95 of its value and this was experimentally observed to be the ease. On increasing the speed of rotation, the deflection steadily diminished. Let D I be the deflection for aperiod T of half alternation, and D the deflection for a very slow rotation of the commutator. The ratio of the number of ions reaching the copper plate is

~

of the total starting out from the platinum surface,

and if t be the time taken for the ions to cross over between the plates

O· 95T (1 - ~1) =

t.

Disclllll'!Jl' (!{ U('('lricilY /rOlli G/OIl'illg P/arillUIIl alU/lhl' Ve/ocily

0/ /OIlS 349

The velocity of the ions is then determined by substituting in the equation of veloeities and we obtain d2 K 4 -( - - V x Q·95T 1 - -D1

-jj--)

Using this method, a large number of determinations of velocities were made for distanees 2, 3 and 5 ems., for different periods of alternation and different voltages. The mean ofthese values is ineluded in the following table. Velocity in cms. per second /or I volt per cm Distance

Maximum

Average

Calculated

2 3

13

7·8 5·8

5·5

5

7·9

4·7

3·2

2·2

The maximum veloeity for a distance of 5 ems. was not determined, as the eommutator sparked aeross before suffieient voltage could be employed to make sure of sweeping baek the ions on their return journey. Tbe eurrent through the platin um plate, in a11 these cases, was kept eonstant at about 100 amperes, whieh eorresponded to the temperature of maximum eurrent at a11 distances except 2 ems. It was experimentally observed that for a given distanee the velocity of the ions inereased with inerease of temperature of the platinum plate, although the aetual eurrent observed through the galvanometer was diminished. The results obtained from the direct measurement of the veloeity of the ions may be summarized as follows = ]. The velocity of the positive ions drawn from glowing platinum is not a eonstant for given eonditions, but may vary within wide limits. 2. The veloeities of the ions inerease with inerease of the temperature of the gas through whieh they pass. 3. The veloeities caleulated from the theoretical formula are too small. For 2 ems. distanee between the plates the differenee between the theoretieal and experimental values is not large, but the difference inereases with distanee apart of the plates. For 5 ems. the caleulated veloeity is less than one-half of the average veloeity. The final eonelusion we may draw from the results is that the veloeity of the ions dedueed from measure of the limiting eurrent does not necessarily represent the average velocity and is mueh less than the maximum value of the veloeity.

The Collected Papers of Lord Rutherford

350

During the course of the investigation two papers appeared by C. D. Child, in the PHYSICAL REVIEW (February and March, 1901). In these papers the velocity of the positive and negative ions drawn from the flames and the arc have been determined by direct measurement of the current, using the theory given in the early part of the paper. His results show elearly that the current varies approximately as the square ofthe voltage, and falls off more rapidly than the inverse cube of the distance. The diminution of the current with distance in the case of flames is not, however, so rapid as in the case of glowing platinum. It will be of interest to compare the values of the velocity of the positive ions drawn from ßames and from glowing platinum. For purposes of comparison we will give the 'calculated' values of the velocities of the positive ions in the two cases: Ions Drawn flom Flame

Ions Drawn from Glowing Platinum

Distance from Edge ofFlame

Calculated Velocity of Positive Ion

Distance Between Plates

Calculated Velocity Positive Ion

1 cm. 2cms. 4 6 8

2·85 cms. per sec. 2·27 2·21 2·00 1·45

2 3 4 5 7

5·5 cms. per sec. 3·18 2·64 2·20 1·90

The calculated velocities of the positive ion in the two cases are thus not very different for distances of 4 to 7 cms. The velocity of the flame ions is, however, considerably smaller for distances of 2 to 4 cms. We have shown, however, that the calculated values given above for the platinum ions are far too small on account of the divergences between experimental and ideal conditions. The method employed by Child is a very simple and accurate one of determining velocities, provided the ions alt travel with the same velocity, i.e., are all of the same size. The presence, however, of some slow moving carriers between the electrodes may diminish the value of the current and the resultant calculated velocity may be much too small. Whether such slow moving carriers are produced in the case of ßames as in the case of glowing platinum cannot be definitely answered without a special investigation. The presence of even a small number of large nuelei, such as are produced by smoke or fumes, tend to make the calculated velocities too smalI. For these reasons the measurements of veloeity of the ions elose to the electric are must be interpreted with great eaution. Macdonald Physics Building, McGiIl University, July Ist.

Übertragung erregter Radioaktivität von

E. RUTHERFORD

From Physika/ische Zeitschrift, 3, 1902, pp. 210-14 (Der amerikanischen physik. Gesellschaft mitgeteilt am 29. Dez. 1901)

EINE der interessantesten Eigenschaften der radioaktiven Substanzen, Thorium und Radium, ist ihre Fähigkeit, allen Körpern in ihrer Nachbarschaft zeitweilige "erregte" Radioaktivität mitzuteilen. Wenn ein stark negativ geladener Draht in ein geschlossenes Metallgefäss gebracht wird, welches Thor oder Radium enthält, so ist die erregte Radioaktivität vollständig auf die negative Elektrode beschränkt. Ist der Draht positiv geladen, so bleibt er inaktiv, aber die erregte Radioaktivität tritt an den Wänden des Gefässes auf. Wenn kein elektrisches Feld wirksam ist, so wird erregte Radioaktivität auf allen Substanzen in der Nachbarschaft des radioaktiven Materials hervorgerufen. Für eine gegebene Menge radioaktiver Substanz ist der Gesamtbetrag der in einer bestimmten Zeit erzeugten Radioaktivität nicht sehr verschieden, sei es, dass die erregte Radioaktivität in einem elektrischen Felde auf der negativen Elektrode konzentriert wird, sei es, dass sie durch den Prozess der Diffusion über die Wände des einschliessenden Gefässes verstreut wird. In früheren Mitteilungen hat der Verfasser die durch Thoriumverbindungen hervorgerufene erregte Radioaktivität untersucht und gezeigt, dass sie innig verknüpft ist mit der Fähigkeit, eine radioaktive "Ausströmung" von sich zu geben. Curie und Debierne* haben im einzelnen die erregte Radioaktivität untersucht, die durch sehr aktive Proben von Radium hervorgerufen wird, wenn kein elektrisches Feld angewendet wird. Dornt fand, dass Proben von Radium (von P. De Haen in Hannover hergestellt) eine ähnliche Ausströmung von sich gaben, wie Thorium. Die erregte Radioaktivität, die von Thorium und Radium herrührt, verschwindet mit der Zeit. Für Thoriumverbindungen fällt die erregte Strahlung in ungefähr 11 Stunden auf ihren halben Wert. Der Abfall der vom Radium erregten Strahlung erfolgt viel schneller, befolgt aber kein einfaches Gesetz. Er ist zuerst schnell und weiterhin viel langsamer. Der Verfasser hat gefunden, dass verschiedene Proben von Radium, die er besitzt, eine erregte Strahlung verursachen, deren Abfall in ganz verschiedener Weise erfolgt. * C. R. 132, 548, 768, 1901. Diese Zeitschrift 2, 500, 513, 1901. t Naturwissenschaftliche Gesellschaft Halle, Juni 1900.

352

The Collected Papers

0/ Lord Rutllelford

Auf der andern Seite verliert die Ausströmung die vom Thorium ausgeht, ihr Strahlungsvermögen sehr viel schneller, als diejenige, die vom Radium ausgeht. Die erstere fällt in ungefähr einer Minute auf ihren halben Wert, während die letztere ihr Strahlungsvermögen einige Wochen beibehält. Die Ausströmungen von Thorium und Radium verhalten sich in jeder Beziehung wie radioaktive Gase oder Dämpfe. Sie diffundieren sehr schnell durch Gase, durch poröse Substanzen, wie Pappdeckel, und, im Gegensatz zu den Gasionen, die sie auf ihrem Wege erzeugen, dringen sie durch Wattepfropfe hindurch und wandern durch Lösungen, ohne Absorption zu erfahren. Der Verfasser vertritt die Anschauung, dass diese Ausströmungen in gewisser Weise die direkte Ursache der erregten Radioaktivität sind. Zur Stütze derselben seien folgende Thatsachen zusammengestellt: 1. Nur die Substanzen, welche Ausströmungen von sich geben, d. h. Thorium- und Radiumverbindungen, haben die Fähigkeit, erregte Radioaktivität hervorzurufen. 2. Wenn das Ausströmungsvermögen von Thorium und Radium durch starkes Erhitzen teilweise zerstört wird, so nimmt die Fähigkeit, Radioaktivität zu erregen, in demselben Verhältnisse ab. 3. Erregte Radioaktivität kann in Substanzen hervorgerufen werden, wenn nur die Ausströmung, nicht aber auch die radioaktive Substanz zugegen ist. Andererseits wird die Fähigkeit der radioaktiven Substanz selbst, Radioaktivität zu erregen, durch einen Gasstrom stark vermindert, der über sie hinwegstreicht und die Ausströmung mit sich fortträgt. Im Falle von Radium kann die Ausströmung in einem geschlossenen Gefässe mehrere Tage abgesperrt sein und doch noch radioaktive Erregtheit erzeugen. Die Strahlungsfähigkeit der Thoriumausströmungen lässt zu schnell nach, als dass sie ein solches Experiment gestattete. Die charakteristische Eigenschaft der erregten Radioaktivität ist die, dass sie in einem starken elektrischen Felde auf die Kathode beschränkt werden kann. Es ist daher wahrscheinlich, dass sie von einem Transport irgendwelcher positiv geladener "Träger" in dem elektrischen Felde herrührt. Die Experimente, die jetzt beschrieben werden sollen, bestätigen diese Anschauung vollkommen und zeigen, dass sich die Träger in einem elektrischen Felde ungefähr mit derselben Geschwindigkeit bewegen, wie das positive Ion. Prinzip der Methode Die Methode, die zur Bestimmung der Geschwindigkeit des Trägers verwendet wurde, ist eine Abänderung einer schon angewendeten Methode zur Bestimmung der Geschwindigkeit des negativen Ions, welches an der Oberfläche eines Metalles durch ultraviolettes Licht hervorgerufen wird. * Sie bedient sich eines wechselnden elektrischen Feldes. Eine gleichgerichtete E. M. K. wurde durch einen rotierenden Kommutator in eine wechselnde E. M. K. von bekannter Frequenz verwandelt. Wenn in dieser Weise ein • Proo. Cambr. Phil. 800., 1898.

Uberlragung erregler Radioaklirilät

353

wechselndes Feld zwischen zwei parallelen Platten erzeugt wird, zwischen denen eine radioaktive Ausströmung gleichmässig verteilt gehalten wird, so werden gleiche Beträge erregter Radioaktivität in jeder Elektrode erzeugt. Wenn hintereinander mit einer wechselnden E. M. K. E o eine Batterie von der E. M. K. EI (EI< E o) aufgestellt wird, so bewegt sich der positive Träger während der einen Hälfte des Wechsels in einem stärkeren elektrischen Felde als während der anderen. Ein Träger bewegt sich folglich während der beiden halben Wechsel um verschiedene Strecken, falls die Geschwindigkeit des Trägers der Stärke des elektrischen Feldes proportional ist, in dem er sich bewegt. Hieraus folgt, dass die erregte Radioaktivität ungleich auf die beiden Elektroden verteilt sein wird. Wenn die Frequenz des Wechsels gross genug ist, so werden die positiven Träger nur innerhalb einer gewissen kleinen Entfernung von einer Platte zu ihr übergeführt werden, der Rest wird im Verlaufe einiger folgender Wechsel zur anderen Platte getragen. .Eo

8

.EM~N~TIDN X.

.EM~N~TIDN

R .Eo

Seien A und B (Fig.) zwei parallele Platten, die radioaktiv gemacht werden sollen. Die radioaktive Ausströmung zwischen ihnen wird gleichförmig verteilt gehalten. Wenn B negativ ist, sei die Potentialdifferenz zwischen den Platten Eo-E h wenn A negativ ist, E o + EI; d sei der Plattenabstand, T die Zeit eines halben Wechsels; C das Verhältnis des Betrages der auf B erregten Radioaktivität zu der auf A und B zusammen. K sei die Geschwindigkeit des positiven Trägers für die Einheit des Potentialgefälles. Unter der Annahme, das Feld zwischen den Platten sei gleichförmig, und die Geschwindigkeit des Trägers sei proportional dem elektrischen Felde, ist dann die Geschwindigkeit des positiven Trägers nach B hin E o - EI K d und während des nächsten halben Wechsels

Eo + EI K nach A hin. M

d

354

The CoUected Papers ofLord Rutile/lord

Die grössten von einem positiven Träger während zweier aufeinander folgender Wechsel zurückgelegten Entfernungen sind _ Eo + EI

_ Eo - EI KT.

d

Xl -

d

,X2 -

Y'T' .LU.

Wir wollen annehmen, die positiven Träger entständen zeitlich gleichförmig mit einem Betrage von q in der Sekunde auf die Einheit des Plattenabstandes. Die Zahl der positiven Träger, die B während eines vollständigen Wechsels erreichen, kann in zwei Teile geteilt werden: 1. Ein Teil, welcher innerhalb der Entfernung Xl von B während der Zeit T des halben Wechsels erzeugt wurde; er hat den Betrag von lXI q T. 2. Alle diejenigen Träger, welche am Ende des voraufgegangenen Wechsels innerhalb der Entfernung Xl von B zurückgeblieben sind. Ihr Betrag ist Xl

-!XI·-q. T. X2

Nun werden alle diejenigen positiven Träger, die zwischen A und B erzeugt werden und B nicht erreichen, während einiger folgenden Wechsel nach B überführt, vorausgesetzt dass die Stärke des elektrischen Feldes die Gewissheit giebt, dass keine bemerkenswerte Wiedervereinigung der Träger auf dieser Strecke eintritt. Die Gesamtzahl der Träger, die während eines ganzen Wechsels erzeugt werden, ist 2 d q T. Das Verhältnis p der Anzahl positiver Träger, welche B erreichen, zu der Gesamtzahl ergiebt sich so zu p

Wenn man für

Xl

und

X2

= ! Xl d

• Xl

+ X2.

X2

die Werte einsetzt, erhält man

+

K= 2(Eo EI) d 2 Eo(Eo - EI) T

•

c.

Bei den Experimenten wurden die Werte von E o, Eh d 2 und T variiert, und die allgemeinen Resultate wurden in Übereinstimmung mit der Gleichung gefunden. Angewendeter Apparat Für die Experimente mit Thoriumausströmung wurde eine dicke Schicht von Thorium in eine flache Kupferschachtel innerhalb eines Hartgummikästchens von 11 qcm Grundfläche und 3 cm Tiefe gelegt, welches fest auf einen metallischen Untergrund gekittet war. Das Thorium wurde völlig mit Filtrierpapier in zwei Lagen bedeckt, welches das meiste von der direkten Strahlung auffing, der Ausströmung aber den Durchgang gestattete. Der

Uhertragullg erregter Radioaktiritüt

355

Apparat wurde durch einen Metalldeckelluftdicht gemacht, der ringsum an dem oberen Rande des Hartgummikästchens in Quecksilber tauchte. Beim Beginn des Versuches wurde ein quadratisches Stück Aluminiumfolie auf das Papier gebracht, welches das Thorium bedeckte, eine Zinkplatte oben auf das Hartgummikästchen gelegt und der Deckel in seine Lage gebracht. Das wurde so schnell als möglich gemacht, und dann das elektrische Wechselfeld angewendet. Die Ausströmung diffundierte schnell durch das Papier und die Aluminiumfolie und verteilte sich zwischen den Platten in dem elektrischen Felde. Nach einiger Zeit, die zwischen 20 und 90 Minuten varüerte, wurde das Aluminium und das Zink weggenommen und ihre Radioaktivität auf dem gewöhnlichen Wege mit Hilfe eines empfindlichen Quadrantenelektrometers geprüft. So wurde das Verhältnis der erregten Radioaktivität auf den beiden exponierten Platten bestimmt. Dieses Verhältnis fand sich unabhängig von der Zeit, die man bis zur Prüfung vergehen liess, so dass die Radioaktivität jeder Platte in demselben Verhältnis abnimmt. Die Mengen von Thorium, die bei den Versuchen verwendet wurden, variierten zwischen 25 und 100 g. Der Betrag der erregten Radioaktivität in einer bestimmten Zeit schwankte mit der Menge des verwendeten Thoriums, aber das Verhältnis auf beiden Platten wurde nicht davon berührt. Im Verlaufe der Versuche ergab sich, dass eine Platte, welche kurze Zeit der Thoriumausströmung ausgesetzt wurde, nach dem Wegnehmen derselben noch einige Stunden eine allmähliche Steigerung seiner radioaktiven Kraft erfuhr. Der Betrag dieses Anwachsens schwankte mit der Zeitdauer der Ausströmungswirkung, erreichte aber bei kurzen Wirkungszeiten den dreioder vierfachen Betrag des Anfangswertes. Für Wirkungszeiten von einigen Stunden ist die Erscheinung nicht so ausgeprägt, nach einer Einwirkung von einem Tage ist sie nur schwer zu beobachten. Derselbe Apparat und die nämliche Methode wurden auch bei einigen Radiumexperimenten angewendet. Das Radium, welches ich besitze, strömte bei atmosphärischer Temperatur sehr schwach aus. Deshalb wurde von einer früher vom Verfasser beobachteten Erscheinung* Gebrauch gemacht, dass der Betrag der Radiumausströmung mehrere tausendmal wächst, wenn man das Radium etwa bis zur Rotglut erhitzt. Die Ausströmung des erhitzten Radiums wurde zunächst durch einen Luftstrom in einen kleinen metallischen Cylinder überfUhrt. Dann wurden dessen Öffnungen geschlossen. Die so gesammelte Ausströmung reichte einige Tage für die Versuche aus. Beim Beginn des Versuches wurden die beiden Platten in eine Ebonitschachtel gebracht und das Wechselfeld angewendet. Durch zwei seitliche Röhren an dem Ebonitkästchen wurde mit einem schwachen Luftstrom ein kleiner Betrag der Ausströmung aus dem Cylinder zwischen die Platten gebracht. Dann wurden die seitlichen Röhren geschlossen. Nach einer Exposition von etwa einer halben Stunde wurde ein I.uftstrom durch die Schachtel getrieben, um sie von der Ausströmung zu • Diese Zeitschrift 2, 429. 190 I.

356

The Collected Papers o/Lord Ruther/ord

reinigen. Die Platten wurden dann entfernt, und ihre Radioaktivität geprüft. Mit Rücksicht auf den anfänglich schnellen Abfall der vom Radium erregten Radioaktivität war es schwer, befriedigende Vergleichungen der Platten vor Ablauf von 15 Minuten zu machen, innerhalb deren der Abfall langsam genug wurde, um eine exakte Bestimmung des Verhältnisses zu gestatten. Alle diese Versuche zeigten, dass dieses Verhältnis unabhängig ist von der Zeitdauer, die man bis zur Untersuchung hat verstreichen lassen. Ziemlich die meisten Experimente wurden mit der Ausströmung von Thorium gemacht. Vergleichungen der Geschwindigkeit des Trägers wurden über ein weites Gebiet der Wechselzahl und der Spannung ausgedehnt. Die allgemeinen Ergebnisse waren mit der oben entwickelten Theorie in Übereinstimmung. Es fand sich, dass bei konstanter Spannung der Wert von p mit abnehmender Wechselzahl sich verminderte. Bei konstanter Wechselzahl nahm er mit der Spannung zu. Obwohl genügend hohe Spannungen angewendet wurden, ergab sich, dass die gemessenen Werte der Geschwindigkeit zu hoch waren. Dies rührt zum Teil her von der Wiedervereinigung von Ionen zwischen den Platten. Wenn eine E. M. K. angewendet wird, die nicht genügt, die Ionen vor der Wiedervereinigung an die Elektroden zu führen, so wird die erregte Radioaktivität sowohl auf die positive, wie negative Elektrode verteilt. In der Theorie haben wir das Potentialgefälle zwischen den Platten als gleichförmig angenommen. In Wirklichkeit sind wir davon weit entfernt, besonders, wenn die Ionisation zwischen den Platten gross ist. Die Versuche von Child und Zeleny haben nachgewiesen, dass ein plötzlicher Potentialfall dicht an jeder Elektrode vorhanden ist, so dass das elektrische Feld in der Nähe der Platten grösser ist, als in der Mitte. Nach dem weiter unten hin entwickelten Gesichtspunkte ist es auch möglich, dass die positiven Träger bei ihrer Entstehung eine grosse Anfangsgeschwindigkeit haben, die sie einen kurzen Weg durch das Gas, unabhängig von dem äusseren elektrischen Felde, tragen kann. Aus diesen Gründen erreicht, wenn die Wechselzahl sehr gross oder das elektrische Feld klein ist, eine grössere Zahl von positiven Trägern die Platte B, als man nach der einfachen Theorie erwarten würde. Die berechneten Werte der Geschwindigkeit sind folglich in diesen Fällen zu gross. Die folgende Tabelle ist ein Beispiel für einige Resultate, die bei verschiedenen Spannungen und Plattenentfernungen erhalten wurden. Temperatur 18°, die Luft fast trocken. Plattenabstand 1,30 cm.

Eo

+ El

Volt

Eo - El

Wechselzahl in der sec

p

75 152 225 300

50 101 150 200

57 57 57 57

0,17 0,27 0,38 0,44

K ern/sec

1,6 1,25 1,17

1,24

Ubertragung erregter Radioaktil'ität

357

Der Wert von K ist in ern/sec für ein Potentialgefälle von ein Volt/ern angegeben. Für das letzte Beispiel, bei dem der Träger sich während jedes halben Feldwechsels über eine Entfernung von mehr als 1,3 cm bewegte, war eine abgeänderte Form der Gleichung notwendig, um die Geschwindigkeit zu berechnen. Der Wert von 1,6 ern/sec bei 50 Volt ist aus den oben entwickelten Gründen zu hoch. Plattenabstand 2 cm. Eo+ EI

237* 300

Eo - EI

207 200

Wechselzahl in der sec

fJ

K

44 53

0,37 0,286

1,47 1,45

Versuche über die Geschwindigkeit des Trägers der vom Radium erregten Radioaktivität sind noch nicht vollendet. Indessen sind sie weit genug vorgeschritten, um zu zeigen, dass die Wirkungen einer Änderung von Wechselzahl und Spannung den beim Thorium erhaltenen durchaus ähnlich sind. Der Wert der Geschwindigkeit des Trägers ist sicherlich nicht sehr verschieden von dem beim Thorium beobachteten. Die Ergebnisse werden bei Radium verwickelt durch eine Verteilung der vom Radium erregten Aktivität, welche immer an der positiven Elektrode in einem starken elektrischen Felde auftritt, sobald die Ausströmung vollständig von den Platten weggeblasen wird. Es sind Versuche im Gange, wenn möglich, die Ursache dieser Wirkung aufzufinden und sie aus den Experimenten auszuschalten. In einer kürzlich erschienenen Arbeit fand Zelenyt die Geschwindigkeit des positiven Ions zu 1,36 ern/sec für ein Potentialgefälle von 1 Volt/ern bei atmosphärischem Druck und Zimmertemperatur. Es scheint, als ob die Geschwindigkeit des positiven Trägers der erregten Radioaktivität dieselbe oder wenigstens keine sehr verschiedene ist von derjenigen des positiven Ions, welches durch Röntgen- oder Becquerel-Strahlen erzeugt wird. Bei den vorstehenden Versuchen haben wir die Übertragung der Radioaktivität in einem elektrischen Felde betrachtet. Dieselbe Entwickelung findet auch ihre Anwendung, wenn kein elektrisches Feld wirkt. In diesem Falle entsteht die erregte Radioaktivität an den Elektroden durch Diffusion des Trägers zu ihrer Oberfläche hin. Der Betrag von erregter Radioaktivität an einem gegebenen Körpersystem wird also von dem Betrage der radioaktiven Ausströmung in ihrer unmittelbaren Nachbarschaft abhängen. Überlegen wir, in welcher Weise der positive Träger zum Überführungsmittel der erregten Strahlung wird, so bieten sich zwei Erklärungen dar. Die • This value should probably be 273 as given in a later paper in Phil. Mag., January 1903 [Ed.]. t Phil. Trans. Roy. Soc., 1900.

358

The Collected Papers ofLord Ruthe({ord

erste ist die in einer früheren Mitteilung aufgestellte (1. c.), nämlich, dass das durch die Ausströmung erregte positive Ion die Fähigkeit hat, radioaktives Material der Ausströmung an seiner Oberfläche zu verdichten, ähnlich wie sich Wasserdampf in einem feuchten Gase an dem negativen Ion kondensiert. Jeder Träger würde so eine Spur von radioaktiver Substanz an die negative Elektrode tragen. Die andere Erklärung, die mir von Professor J. J. T horn s 0 n nahe gelegt wurde, ist die, dass die Moleküle der Ausströmung die Fähigkeit haben, negativ geladene "Korpuskeln" oder Elektronen auszusenden, ähnlich wie das Radium im festen Zustande. Jedes Molekül, welches ein negatives Korpuskel ausgesandt hat, behält eine positive Ladung zurück und wird darum an die negative Elektrode übergeführt. Beide Erklärungen würden genügen, um die Ablagerung radioaktiver Substanz irgend welcher Art an der negativen Elektrode anschaulich zu machen. Die Ansicht, dass erregte Radioaktivität durch eine strahlende Substanz verursacht wird, die sich an Körpern ablagert, ist bis zu einem hohen Grade von Wahrscheinlichkeit bestätigt. Ich brauche nur zwei der zwingendsten Thatsachen zur Bestätigung dieser Anschauung zu erwähnen. Bei der Untersuchung der vom Thorium erregten Aktivität habe ich gezeigt (I. c.), dass der Betrag der an Körpern erregten Radioaktivität von der chemischen Natur der Substanz völlig unabhängig ist. Derselbe Betrag wird auf Glimmer, Papier oder Metallen unter gleichen Bedingungen abgelagert. Ich habe auch gezeigt, dass die auf einem Metalle, z. B. Platin, erregte Radioaktivität teilweise in Säure gelöst werden kann, und in der Lösung zurückbleibt. Dampft man die Lösung trocken ein, so bleibt die Radioaktivität im Rückstand. Hieraus erhellt, dass die Radioaktivität von einem Niederschlag radioaktiver Substanz herrührt, welche ein bestimmt definiertes chemisches Verhalten zeigt. Dieser Ansicht steht es nicht im Wege, dass an einem stark radioaktiven Körper eine Gewichtsänderung nicht nachgewiesen werden kann; denn aller Wahrscheinlichkeit nach ist die Strahlungsfähigkeit dieser Substanz für ein gegebenes Gewicht ungeheuer viel grösser, als bei den aktivsten Proben von Radium, die man bisher hergestellt hat. Was wir bisher wissen, genügt noch nicht, um endgültig zwischen den beiden Anschauungen zu entscheiden, die bestimmt sind, die Entstehungsweise des positiven Trägers zu erklären, der das Überführungsmittel ist. Die von J. J. Thomson angeregte Elektronenthorie scheint die einfachste Erklärung der Erscheinungen zu sein; aber ehe man sie endgültig annimmt, sind noch gewisse Dissonanzen zwischen Theorie und Experiment aufzulösen. Ich habe (1. c.) gezeigt, dass bei Drucken von der Grössenordnung I mm Quecksilber die erregte Radioaktivität in einem elektrischen Felde nicht auf die negative Elektrode beschränkt ist, sondern sich über die ganze Wand des Gefässes verteilt. Nach der Elektronentheorie würde die Radioaktivität in einem starken elektrischen Felde bei allen Drucken völlig auf die Kathode beschränkt sein. Bei den erwähnten Versuchen war das elektrische Feld nicht sehr stark, und es ist möglich, dass der positive Träger eine hohe Anfangsgeschwindigkeit in dem Momente erhält, wo das Elektron von dem Molekül

Ch(,l'lragUng (,1'1'('g!('/, Radioak!il'iliil

359

weggeschleudert wird. Wenn das Elektron eine Anfangsgeschwindigkeit von 10 10 ern/sec hat, würde dies bestimmt der Fall sein. Trotzdem das elektrische Feld sehr gross ist, ist es so möglich, dass einige der positiven Träger eine genügend grosse Geschwindigkeit haben, um zu entrinnen und an die Anode zu gelangen. Gewisse Erscheinungen, die man beobachtet hat, unterstützen diese Auffassung. Es sind jetzt Versuche im Gange, die Verteilung der Radioaktivität bei den tiefsten Drucken, die man erhalten kann, im starken elektrischen Felde zu untersuchen. Ich hoffe, dass diese Experimente in die vorgetragenen Anschauungen noch mehr Licht hineinbringen werden. McGiIl Universität. Montreal, 15. Dez. 1901 (Aus dem Englischen übersetzt von H. Th. Simon) (Eingegangen am 22. Januar 1902)

Erregte Radioaktivität und in der Atmosphäre hervorgerufene Ionisation von

E. RUTHERFORD

und

S. I. ALLEN

From Physikalische Zeitschrift, 3, 1902, pp. 225-30

01or der amerikanischen Physikalischen Gesellschaft vorgetragen am 27. Dez. 1901)

DIE Versuche von Elster und Geitel* und C. T. R. Wilsont haben endgültig gezeigt, dass ein gut isolierter, geladener Leiter innerhalb eines geschlossenen Gefässes allmählich seine Ladung verliert, und dass dieser Ladungsverlust von einer geringen, von selbst auftretenden Ionisation des Gasvolumens im Innern des Gefässes herrührt. Wilson berechnete aus diesen Versuchen, dass ungefähr 19 Ionen in der Sekunde auf den cm3 des Gases erzeugt werden. Ganz kürzlich haben Elster und Geitel gezeigtt, dass ein negativ geladener Leiter in offener Luft zeitweilige Radioaktivität erhält. Diese Radioaktivität entweicht innerhalb weniger Stunden und ist ganz ähnlich der erregten Radioaktivität, die unter der Wirkung von Thorium und Radium in Substanzen auftritt. Sie kann in derselben Weise, wie das der eine von uns§ für die von Thorium erregte Radioaktivität gezeigt hat, durch Auflösung in Säuren teilweise entfernt werden. Dampft man die Lösung ein, so wird die Aktivität auf die Wände des Gefässes übertragen. Bei den Versuchen von Elster und Geitel und Wilson wurde der Betrag der Ionisation der atmosphärischen Luft bestimmt durch Beobachtung der Zeit, innerhalb deren die Blättchen eines Elektroskopes besonderer Art zusammenfielen. Diese Bestimmungsmethode ist im allgemeinen langsam und gestattet in manchen Fällen nicht, die Versuchsbedingung genügend zu variieren. Bei den vorliegenden Versuchen benutzten die Verfasser ein empfindliches Quadrantenelektrometer zur Untersuchung der Ionisation der Luft und der durch Luft hervorgerufenen erregten Radioaktivität. Das verwendete Elektrometer ist eine Abänderung des von D 0 I e z ale k beschriebenen Instrumentes (Verh. d. D. Physik. Ges. 3, 18-72, 1901). Es ist von dem gewöhnlichen Quadrantentypus mit einer leichten Nadel von Silberpapier, die an einem feinen Quarzfaden aufgehängt ist. Der Apparat, wie er von der Firma Ge 0 r g Bar tel s in Göttingen konstruiert wird, ist für

* Diese Zeitschrift 2,590, 1901. t Diese Zeitschrift 3, 76, 1901.

t

Proc. Roy. Soc., 1901. § Phil. Mag. Febr., 1900.

Erregte Radioaktil'ität und in der Atmosphäre hervorgerufene Ionisation

361

die Bestimmung kleiner Potentialdifferenzen bei elektrochemischen Arbeiten bestimmt. Für unsere Zwecke war es notwendig, die Isolation und Verbindungsweise der Quadranten vollständig zu verändern. Bei unseren Versuchen wurde die Nadel alle zwei Tage durch leichtes Berühren mit einem an eine Batterie von 200 Volt angeschlossenen dünnen Draht geladen. Es fand sich, dass die Nadel nicht mehr als 10% ihrer Ladung in 24 Stunden verlor. Die Dämpfung der Nadel war infolge ihrer Leichtigkeit sehr gross und es war kein weiterer Dämpfungsflügel nötig. Die Ablenkung wurde mit Spiegel und Skala bei einem Abstande von 2 m abgelesen. Der Nullpunkt blieb sehr konstant, und die Ablesungen konnten, wenn nötig, auf /0 mm genau gemacht werden. Für die erste Aufhängung, die verwendet wurde, gab das Elektrometer einen Ausschlag von 2000 mm der Skala bei einem Volt Potentialdifferenz zwischen den Quadranten und einer Ladung der Nadel auf 200 Volt. Diese Aufhängung wurde zufällig im Laufe der Untersuchungen zerbrochen, und der neue Quarzfaden gab nur mehr etwa ein Viertel dieses Ausschlages für dieselbe Spannung. In Anbetracht des kleinen Betrages der Entladung, der durch die freiwillige Ionisation der Luft eintritt, ist es sehr wesentlich, dass jede Vorsicht getroffen wurde, um vor äusseren elektrostatischen Störungen sicher zu sein. Das Elektrometer und alle Zuführungsdrähte wurden in Metallcylinder eingeschlossen, die zur Erde abgeleitet waren. Der Boden und das Holzwerk in der Nachbarschaft des Prüfungsapparates wurden mit Metall bedeckt und an Erde gelegt, die Schaltung der Quadranten wurde mit Hilfe eines besonderen Quecksilberschlüssels bewirkt, der aus der Entfernung mittels eines Bindfadens bethätigt wurde. Die isolierenden Substanzen, die bei der Anordnung notwendig waren, wurden mit Hilfe von Flammen vollständig entelektrisiert. Ionisation der atmosphärischen Luft Vorversuche zeigten, dass der mit dem Elektrometer zwischen Cylindern beobachtete Strom nur von dem Volumen des zwischen den bei den Cylindern eingeschlossenen Gases und nicht von der Natur der Elektroden abhing. Folgende Versuchsanordnung wurde getroffen, um die Zahl der im cm3 Luft pro Sekunde erzeugten Ionen und die Änderung des Ionisationsstromes mit der Potentialdifferenz zwischen den Elektroden zu bestimmen. Der Ionisationsstrom wurde zwischen zwei konzentrischen Zinkcylindern von 124 cm Länge, 25,5 und 7,5 cm Durchmesser beobachtet. Die Cylinder waren senkrecht aufgestellt, und ihr Boden war geschlossen. Der weite Cylinder war oben durch eine Zinkplatte geschlossen, in deren Mitte eine Kreisöffnung von etwas grösserem Durchmesser angebracht war, als dem inneren Cylinder entsprach. Ein oben an dem inneren Cylinder ringsum befestigter Verschlussring grenzte an einen Ebonitring. Zwischen dem Ebonit und der Zinkplaue befand sich ein dünner Metallring, der an Erde gelegt war, und dieser stiess an einen Ring aus einem Halbleiter, wie Pappe. Der dünne, an Erde gelegte Metallring diente als Schutz, so dass bei keiner M*

362

The Collected Papers of Lord Rutherford

noch so hohen Potentialdifferenz zwischen den Cylindern ein Strom über den Isolator zum inneren Cylinder fliessen konnte. Der innere Cylinder war in der bekannten Weise mit dem Elektrometer verbunden. Der äussere Cylinder war an den einen Pol einer Batterie angeschlossen, der andere Pol derselben lag an Erde. Die Elektrometernadel zeigte bald eine von dem Ionisationsstrome zwischen den Elektroden herrührende lebhafte Bewegung, bei einer Potentialdifferenz von wenigen Volt zwischen den Cylindern. 80· 7S·

I

10 65 65

II

65

80·

tbwn. derl'-itfäkigkeiJ./'el

7S·

cona/IIt,.ÜMM CalfiaJen".

65 65

10 65 65 65

s

Jeder Theil-o,ooisvolt. Jeder Theil-o,ooisvolt.

65

Jeder Theil-o,ooisvolt.

10

s

tbwn. derl'-itfäkigkeiJ./'el der 10

'$

20

2$

3D

40

lfS

50

:u

Fig. I Der Cylinder wurde ziemlich luftdicht gemacht und ohne Störung stehen gelassen. Über einen Zeitraum von mehr als einem Monat wurden Beobachtungen des Ionisationsstromes zwischen den Cylindern gemacht. Um Korrektionen wegen der Empfindlichkeitsänderung des Elektrometers zu vermeiden, wurde zu gleicher Zeit der Ionisationsstrom zwischen zwei isolierten parallelen Platten gemessen, der von einer Einheitsprobe Uraniumoxyds erzeugt wurde. Die Kurven der Fig. I zeigen die Beziehung zwischen dem Strome im Gase und der verwendeten Spannung. Kurve I wurde aufgenommen, nachdem das Gas einen Monat lang ohne Störung in dem Cylinder gewesen war, Kurve 11 einige Stunden, nachdem gewöhnliche Zimmerluft in den Apparat eingeführt war. Der Strom ist für 50 Volt in beiden Fällen fast derselbe; die allgemeine Gestalt der Ionisationskurven ist derjenigen sehr ähnlich, die man bei Ionisation der Luft durch Röntgen- und Becquerel-strahlen beobachtet hat. Mit Rücksicht auf den sehr kleinen Betrag der Ionisation in dem Gase und der infolgedessen langsamen Wiedervereinigung wird der maximale

D'I'l'gte Radioak I il'iriit /ll/d in der A fll1osl'hiil'l' hl'1'\'OI:~l'I'/lfene Ionisatioll

363

Strom bei einer sehr niedrigen Spannung erreicht. Der Unterschied in Kurve I und 11 rührt wahrscheinlich von der Anwesenheit von Staubteilehen im letzteren Falle her. Einige der Ionen geben bei ihrer langsamen Wanderung zwischen den Elektroden ihre Ladungen an die Staubwolken ab, und dadurch wächst offenbar die Schnelligkeit der Wiedervereinigung der Ionen in dem Luftvolumen. Es muss beachtet werden, dass in Kurve I der maximale Strom nahe bei einer Potentialdifferenz von 5 Volt erreicht wird. Die Kapazität des Elektrometers, des Cylinders und der Isolatoren war 150 E. S.-Einheiten, während I mm Teilstrich am Elektrometer 0,00182 Volt entsprach. Der Durchnittswert der Elektrometerablenkung während der länger als einen Monat dauernden Beobachtungen war 100 Teile in 132 Sekunden, bei 50 Volt zwischen den Cylindern. Der Ionisationsstrom zwischen den Cylindern war so 6,9' 10-- 4 E. S.Einheiten oder 2,3' 10- 13 Ampere. Das Volumen zwischen den Cylindern war 71 200 cm 3 • Nimmt man den Wert von 6,5' 10- 10 E. S.-Einheiten als die Ladung auf einem lon*, so ist die Zahl der im cm 3 pro Sekunde erzeugten Ionen 15. Dieser Wert ist nicht sehr verschieden von der Zahl 19, die von Wilson mit Hilfe der elektroskopischen Methode gefunden wurde. Ein Unterschied in dem Ionisationsstrome zwischen den Cylindern wurde innerhalb eines Zeitraumes von mehr als einem Monat nicht beobachtet. Die Erzeugung von Radioaktivität in der Luft legte die Anschauung nahe, dass möglicherweise eine radioaktive Ausströmung in der Luft vorhanden war. Wenn das so ist, so nimmt deren Strahlungsfähigkeit sehr viel langsamer ab, als die vom Radium herrührende Ausströmung. Um zu prüfen, ob zeitweilige Ionisation ausser in Luft auch in anderen Gasen erregt würde, wurde der grosse Cylinder durch Verdrängung der Luft mit Kohlensäure gefüllt. Die Kohlensäure wurde aus einer Bombe mit käuflicher flüssiger Kohlensäure entnommen. Die Ionisation war zuerst grösser als die in Luft, aber nach einigen Stunden ging sie allmählich auf einen von dem bei Luft gefundenen nicht sehr verschiedenen Wert herab. Dieses Resultat scheint zu zeigen, dass in Kohlensäure eine zeitweilige Ionisation von etwa derselben Grössenordnung auftritt, wie in Luft. Nach der Natur dieser Experimente indessen musste ein kleiner Bruchteil sowohl von Luft wie von anderen Verunreinigungen zugegen sein, und es ist möglich, dass eine solche Beimengung das Resultat stark beeinflusst. Erzeugung von erregter Radioaktivität Wie EIs t e rund Gei tel zuerst hervorgehoben haben, ist die Erscheinung der von Luft erregten Radioaktivität derjenigen sehr ähnlich, die von Thorium und Radium hervorgerufen wird. Die Radioaktivität wird in einem starken elektrischen Felde in beiden Fällen allein an der Kathode erzeugt. Noch nie ist Radioaktivität an einem positiv geladenen und der Luft ausgesetzten • J. J. Thomson, Phi!. Mag. Decbr. 1898.

364

The Collected Papers of Lord Rutherford

Drahte beobachtet worden. Eine besondere Reihe von Versuchen wurde gemacht, um die zeitliche Abnahme der erregten Radioaktivität zu ermitteln, die an einer negativ geladenen Oberfläche entsteht. Bei einem Versuche wurde eine isolierte, 8 Fuss lange Messingstange, die ausserhalb des Fensters angebracht war, mit Hilfe einer grossen Reibungselektrisiermaschine auf einen Potential von ungefähr 100000 Volt gehalten. Nach einer Exposition von einer Stunde wurde der Stab weggenommen, innerhalb eines Prüfcylinders aufgestellt, und die durch die erregte Radioaktivität hervorgerufene Ionisierung zwischen den Cylindern in regelmässigen Zeiträumen gemessen. Wenn der Stab positiv geladen war, wurde keine erregte Radioaktivität erzeugt. .",

30 30

30

16' .",

30 30

.8

Jeder Theil-o,ooisvolt.

.",

30

Jeder Theil-o,ooisvolt.

30

Jeder Theil-o,ooisvolt. Jeder Theil-o,ooisvolt.

30

Cun-en. du .Ahm. lu für durch Luh erreste Radiom;.(;f.y;,tiit.

.",

30

.",

.",

30

30

30 30 30 30

Zeit;

in- Minflt;eAv

OM~~~~.~~~~~~~~~*~~~~m~

Fig.2 Bei einem andern Versuche wurde ein langer Kupfer- oder Bleidraht in dem grossen Dachgeschoss des Laboratoriums aufgehängt, wo keine Gelegenheit war, dass sich die Luft durch die in dem Laboratorium verwendeten radioaktiven Substanzen verunreinigte. Der Draht wurde einige Stunden mit einer durch einen Motor getriebenen Wimshurstmaschine auf einen Potential von 20 000 bis 30 000 Volt gehalten. Der Draht wurde dann vollständig auf ein eisernes Gestell aufgewickelt und in einen Prüfcylinder gebracht. Die Abnahme der Radioaktivität in der Zeiteinheit fand sich unabhängig von dem Material des Drahtes oder Stabes, und, innerhalb der Versuchsgrenzen, nicht sehr von der Spannung und der Expositionszeit des Drahtes beeinflusst. Der Grad der erregten Radioaktivität wächst bei einem gegebenen Drahte zuerst regelmässig mit der Zeit, aber nach einigen Stunden sehr viel langsamer. In den Kurven der Figur 2 ist die Abnahme der Strahlungsfähigkeit mit

Erregte Radioaktirität und in der A tlllosphäre herrvrgerufene Ionisation

365

der Zeit dargestellt. Die Ordinaten bedeuten Teile der Elektrometerskala, die in der Sekunde vorüberwandern, Kurve I gilt für einen Kupferdraht vom Durchmesser 0,01 cm, 20 m lang und 2 Stunden einer Potentialdifferenz von -29000 Volt ausgesetzt. Um den Draht zu prüfen, wurde er auf ein Eisengestell von 121 cm Länge aufgewickelt und in einen Cylinder von Eisengaze gebracht. Die Potentialdifferenz zwischen den Cylindern war 50 Volt. Die natürliche Zerstreuung, die von der zeitweiligen Ionisation der Luft herrührte, war 2,5 Teile in der Sekunde. Nachdem die Korrektur für die natürliche Zerstreuung angebracht ist, ergiebt sich, dass der Ionisationsstrom (welcher ein Mass für die Intensität der Strahlung ist), in 52 Minuten auf die Hälfte seines Wertes sinkt. Kurve 11 gilt für einen Bleidraht von 10m Länge und 0,125 Durchmesser, der 190 Minuten auf-30000 Volt gehalten wurde. Der Draht wurde in Form einer flachen Schneckenlinie gewunden, und der zwischen zwei Metallplatten hervorgerufene Ionisationsstrom gemessen. Die natürliche Zerstreuung des Apparates in diesem Falle war 0,14 Sk. Teile in der Sekunde. Hier fällt die erregte Radioaktivität in ungefähr 45 Minuten auf ihren halben Wert. Diese beiden Beobachtungen liegen um zwei Monate auseinander und wurden unter sehr verschiedenen atmosphärischen Bedingungen angestellt. Der Abfall in der Zeiteinheit der durch Luft erregten Radioaktivität ist sehr viel schneller als der durch Thorium, welche innerhalb 11 Stunden auf die Hälfte ihres Wertes sinkt. Endgültige Vergleichungen mit der durch Radium erregten Radioaktivität konnten nicht gemacht werden, da deren Abfall unregelmässig ist und von der Besonderheit des gerade verwendeten Radiums abhängt. Durchdringungsfähigkeit der erregten Strahlung Frühere Versuche hatten gezeigt, dass die Durchdringungsfähigkeit der von Thorium und Radium erregten Strahlungen gleich war. Es war von Interesse, die von Luft erregte Strahlung damit zu vergleichen. Bei diesen Versuchen wurden Bleidrähte angewendet, um sie leicht in die Form von flachen Schneckenlinien bringen zu können. Der Draht wurde durch eine Exposition von zwei bis drei Stunden bei -30000 Volt erregt. Er wurde dann in die Form von flachen Schneckenlinien gewunden und in einen Apparat aus parallelen Platten gebracht. Der Ionisationsstrom zwischen diesen Platten wurde beobachtet, indem verschiedene Lagen dünner Aluminiumfolie aufgelegt wurden. Die durchschnittliche Dicke der Folie war 0,00034 cm. Die Resultate sieht man in Kurve I der Figur 3, wo die Durchdringungsfähigkeit anderer bekannter Formen von Strahlung zum Vergleich beigefügt ist. Der Ladungsverlust in der Zeiteinheit ist für die unbedeckte radioaktive Oberfläche in jedem Falle mit 100 angenommen. Die durch Luft erregte Strahlung hat ein grösseres Durchdringungsvermögen wie jede andere Form der Strahlungen, die nicht in einem magnetischen

366

The Collected Papers

0/ Lord Rutherford

Felde abgelenkt werden, also der von den radioaktiven Substanzen Uranium, Thorium,Aktinium und Radium ausgehenden; ebenso ist sie durchdringender als die von Radium und Thorium erregte Strahlung. Es sind jetzt Versuche im Gange, um die Änderung der auf 1 cm 3 Luft entfallenden Ionenanzahl zu verschiedenen Zeiten zu bestimmen. Hierzu wird die Luft von ausserhalb des Gebäudes her mit Hilfe eines Ventilators durch einen 30 cm weiten Metallcylinder geleitet. Die Luft passiert bei ihrem Laufe zwei parallele im Abstand von 2 cm isoliert angebrachte Drahtmarken. Der Draht nächst dem Ende ist mit dem Elektrometer, der andere mit einer grossen Batterie verbunden.

tao 90

Yon"JUumirUumfolie

70 60 SO

40 30-

20

'''' •

Jeder Theil-o,ooisvolt. svosvo . .

So

'0

Lasen Yon"JUumirUumfolie •

.2.

3

'-'

oS

6

Fig.3 Für einen bestimmten Luftstrom wächst der am Elektrometer beobachtete Strom mit der Spannung, bis ein Punkt erreicht ist, bei dem eine Zunahme der Spannung keine Stromzunahme mehr bewirkt. Wenn die zweite Marke positiv geladen ist, so wandern die positiven Ionen dem Luftstrom entgegen. Wenn die Ionengeschwindigkeit in dem elektrischen Felde wesentlich grösser ist, als die des Luftstromes, so erreichen alle positiven Ionen die erste Marke und der Elektrometerstrom ist ein Maximum. Aus derartigen Beobachtungen, bei denen die Geschwindigkeit des Luftstrornes zwischen 100 und 250 ern/sec verändert wurde, ergab sich die Geschwindigkeit des positiven Ions zu ungefähr 1,5 ern/sec für 1 Volt/ern Potentialgefälle. Das ist nicht sehr verschieden von dem von Zeleny* gefundenen Werte 1,36, der sich auf Ionen bezieht, die durch Röntgenstrahlen in Luft von atmosphärischem Drucke und Zimmertemperatur erregt werden. Wegen der fortdauernden Änderungen der Leitfähigkeit der bei diesen

*

Phil. Trans. Roy. Soc., 1900.

EI'I'l'!{te Radioaktil'itüt

IIlld

in eier Atmnspliiil'l'

1il'/'\'(l1:~t?l'l~fc'Il(,

Ionisation

367

Versuchen durch den Cylinder gesaugten Luft konnten wir bisher die Geschwindigkeit des negativen Ions noch nicht mit Genauigkeit bestimmen. Oie Ergebnisse zeigen indes jedenfalls soviel, dass diese Geschwindigkeit wesentJich grösser ist, wie diejenige des positiven Ions. Wenn man den maximalen Ladungsverlust zwischen den Marken für einen bestimmten Luftstrom bestimmt hat, kann man daraus die Zahl von positiven und negativen Ionen ermitteln, die in der durch den Cylinder geführten Luft vorhanden sind. Die Versuche sind schon im Gange, und die allgemeinen Ergebnisse zeigen, dass diese Zahl sich beständig ändert, von Stunde zu Stunde, von Tag zu Tag. Klare sonnige Tage haben ziemlich die höchsten Werte ergeben, während Tage mit Schneefall tiefe Werte liefern. Bei den meisten dieser Versuche, die wir während des trockenen kanadischen Winters machten, wechselte die Aussentemperatur zwischen -3 und . -18 0 C. Diskussion der Ergebnisse Bei diesem Stande unserer Kenntnis über die Radioaktivität dürfte es nicht überflüssig erscheinen, einige Vorstellungen über die mögliche Ursache der zeitweiligen Ionisation der Luft und die von Luft erregte Radioaktivität zu entwickeln. In der vorhergehenden Nummer dieser Zeitschrift* hat der eine von uns nachgewiesen, dass die von Thorium- und Radiumverbindungen erregte Radioaktivität direkt von der radioaktiven Ausströmung herrührt, die von diesen Substanzen ausgeht. Es wurde gezeigt, dass die erregte Radioaktivität von der Ablagerung einer strahlenden Substanz an der negativen Elektrode herrührt, vermittelt durch positive "Träger", welche in einem elektrischen. Felde mit Geschwindigkeiten wandern, die von derjenigen des positiven, durch Röntgen- und Becquerelstrahlen in Luft erregten Ions sehr wenig verschieden sind. Wenn kein elektrisches Feld vorhanden ist, so werden diese radioaktiven Träger durch Diffusion auf alle Körper in ihrer Nachbarschaft zerstreut. In einem starken elektrischen Felde werden sie alle zur Kathode getrieben, auf die dann die erregte Radioaktivität beschränkt ist. Zwei mögliche Erklärungen der Entstehungsweise dieser positiven Strahlungsträger wurden vorgebracht, entweder. ]. dass die radioaktive Substanz der Ausströmung auf dem positiven Ion verdichtet wird, welches von der Ausströmung durch Strahlung erzeugt wird (ähnlich, wie sich Wasserdampf in einem ionisierten Gase auf dem negativen Ion niederschlägt), oder. 2. dass das Molekül der Ausströmung (nach den Anschauungen J. J. Thomsons) die Fähigkeit besitzt, ab und zu ein negativ geladenes Korpuskel oder ein Elektron mit grosser Geschwindigkeit von sich wegzuschleudern. Als Folge hiervon würde das Molekül eine gleichgrosse positive Ladung zurückbehalten lind in einern starken elektrischen Felde zu der Kathode

*

Diese Zeitsehr. 3, 210. 1902.

368

The Collecled Papers 0/ Lord Ruther/orel

wandern. Von dieser Anschauung aus würde die erregte Strahlung von inneren Schwingungen herrühren, die in dem Molekül infolge der Ausstossung des Elektrons erregt würden. Obschon beide Anschauungen die experimentellen Ergebnisse hinreichend zu erklären vermögen, scheint doch die zweite die wahrscheinlichere zu sein. Wenn wir die Elektronenhypothese annehmen, so muss vorausgesetzt werden, dass die Fähigkeit, Elektronen auszuschleudern, bei bestimmten Formen der Materie sehr ausgesprochen ist, wie bei der Ausströmung von Radium- und Thoriumverbindungen und bei festen Körpern, wie Uranium, Radium und Aktinium. (Siehe die Resultate von Becquerel, Curie, Debierne und anderen.) Es ist indessen möglich, dass diese Fähigkeit in viel geringerem Grade auch bei anderen bekannten Formen der Materie vorhanden ist. Wenn einer oder mehrere der Gasbestandteile unserer Atmosphäre die Eigenschaft hätte, ab und zu ein Elektron abzustossen, so würde die zeitweilige Ionisation der Luft und die durch sie erregte Radioaktivität auf einmal erklärt sein. Die zeitweilige Ionisation der Luft würde so von den Ionen herrühren, die in dem Gase durch die Bewegung des weggeschleuderten Elektron hervorgerufen würden, gerade so, wie ein Kathodenstrahlträger auf seinem Wege Gasionen erzeugt. Die positiven Strahlungsträger, die nach der Ausstossung des Elektrons übrigbleiben, würden an die Kathode überführt werden und dort zu den Erscheinungen der erregten Radioaktivität Anlass geben. Da es unwahrscheinlich ist, dass innere Schwingungen von Molekülen verschiedener chemischer Natur sowohl nach Charakter wie nach Dauer dieselben sind, so ist zu erwarten, dass von verschiedenen Substanzen erregte Radioaktivität sowohl in Bezug auf die Durchdringungsfähigkeit, wie auch den Ladungsabfall pro Zeiteinheit verschieden sein wird. Die Versuchsergebnisse dieser Mitteilung zeigen, dass die erregte Strahlung für Thorium, Radium und Luft mit sehr verschiedenem Betrage abfällt. Ebenso ist die Durchdringungsfähigkeit der von Luft erregten Strahlung grösser als die vom Thorium und Radium. Vorstehende Anschauung ist früher von dem einen von uns in einem Briefe an die "Nature" entwickelt worden, wo auch gezeigt ist, dass die Ausströmung vom Radium sich wie ein radioaktives Gas verhält und direkt der erregten Radioaktivität entspricht. Elster und Geitel haben diese Anschauung aufgenommen und die Ansicht ausgesprochen, dass die zeitweilige Ionisation und erregte Radioaktivität der Luft von einer radioaktiven Ausströmung oder einem radioaktiven Gase in unserer Atmosphäre herrühren möchte. Diese Ausströmung würde zur Ionisation und erregten Radioaktivität in derselben Weise Anlass geben, wie Thorium-und Radiumausströmung. Zwei von uns beobachtete Thatsachen, welche mit der Elektronhypothese in gutem Einklange stehen, sind nicht leicht nach der Ausströmungsanschauung zu erklären. Es wurde gezeigt, dass der Betrag der zeitweiligen

Erregte Radioaktil'ität und in der Atmosphäre hervorgerufene Ionisation

369

Ionisation in einer begrenzten Luftmasse immer derselbe blieb, wenn man dieselbe einen Monat unverändert liess, und dass die zeitweilige Ionisation in Kohlensäure von der in Luft nicht sehr verschieden war. Will man diese Ergebnisse aus der "Ausströmungsanschauung" erklären, so müsste man annehmen, dass die Strahlung von der Ausströmung weg mit äusserster Langsamkeit erfolge, und dass der Betrag der mit Kohlensäuregas vermischten Ausströmung ungefähr derselbe wie in Luft wäre. Das letztere erscheint besonders unwahrscheinlich. Gegenwärtig sind Versuche im Werke, ob eine oder mehrere Bestandteile der Atmosphäre, chemisch dargestellt, eine ausreichende zeitweilige Ionisation und erregte Radioaktivität zeigen, um von der in der Atmosphäre beobachteten Wirkung Rechenschaft zu geben. McGill Universität, Montreal, Physikalisches Laboratorium, 20. Dez. 1901 (Aus dem Englischen übersetzt von H. Th. Simon) (Eingegangen am 25. Januar 1902)

Versuche über erregte Radioaktivität von

E. RUTHERFORD

From Physikalische Zeitschrift, 3, 1902, pp. 254-7

In einer früheren Mitteilung*) habe ich gezeigt, dass die von Thoriumverbindungen erregte Radioaktivität in einem starken elektrischen Felde auf die negative Elektrode konzentriert werden kann. Der Betrag der erregten Radioaktivität, die in einer gewissen Zeit unter gleichen Bedingungen hervorgerufen wird, ist unabhängig von der chemischen Beschaffenheit der Elektrode, und ebenso von der Grösse der Fläche, auf der die erregte Radioaktivität hervorgerufen wird. Die Abnahme der erregten Radioaktivität mit der Zeit ist unabhängig von der Natur der Substanz, auf der sie hervorgerufen worden ist, und von dem Drucke und der Natur des umgebenden Gases. Es wurde gezeigt, dass die Intensität der erregten Strahlung mit der Zeit abnimmt und in etwa 11 Stunden auf die Hälfte ihres Anfangswertes sinkt. Bei diesen Versuchen wurde der radioaktiv zu machende Körper stets rur Zeiten von zwei Stunden bis zu mehreren Tagen in Gegenwart von Thorium exponiert. Bei neueren Arbeiten stellte es sich als notwendig heraus, den Betrag der Radioaktivität an einem Körper zu untersuchen, der nur eine kurze Zeit in Gegenwart des Thoriums exponiert war. Für diesen Zweck wurde ein besonders empfindliches Elektrometer angewandt, um mit Genauigkeit den geringen Betrag der erregten Strahlung zu messen. Es fand sich, dass der Betrag der erregten Strahlung, am Elektrometer gemessen durch Beobachtungen des Ionisationsstromes zwischen parallelen Platten oder konzentrischen Cylindern, noch einige Stunden hindurch beständig wuchs, nachdem das Thorium entfernt war. Die folgenden Tabellen genügen, um den allgemeinen Verlauf der Erscheinung zu zeigen. 1. Platindraht als Kathode 15 Minuten in einem Cylinder exponiert, der Thorium enthielt. Potentialdifferenz zwischen den Elektroden 110 Volt; dann wurde der Draht entfernt und die Radioaktivität in bestimmten Intervallen geprüft. Erste Beobachtung 5 Minuten nach der Entfernung.

* Phil. Mag., Februar 1900.

371

Versuche über erregte Radioaktil'ität Zeit

o

7,5 24 43 58 78 99

Bewegung der Elektrometernadel in Skalenteilen pro Sekunde

1,9 2,8 4,0 4,6 5,2 5,9 6,5

In diesem Falle wuchs die Radioaktivität um das mehr als Dreifache und hatte auch nach einer Zeit von 99 Minuten ihren grössten Wert noch nicht erreicht. 11. Aluminiumblatt als Kathode in einem Apparat mit parallelen Platten, der Thorium enthielt. Exposition 41 Minuten. Erste Beobachtung 6 Minuten nach der Entfernung. Zeit

0 21 Minuten 31

57 70 91 120 160 ]80 22 Stunden 49

Skalen teile pro Sekunde

1

1,6 1,8 2,0 2,2 2,5 2,9 2,9 2,9 1,0 0,21

In diesem Falle wurde, um vergleichen zu können, der Ionisationsstrom beim Beginne der Beobachtungen als Einheit genommen. Man bemerkt, dass die Radioaktivität gerade wie in der ersten Tabelle, mit der Zeit sehr schnell wächst und nach 2 Stunden ein Maximum von 2,9 erreicht. Dann bleibt sie einige Stunden lang annähernd konstant, und vermindert sich schliesslich allmählich im normalen Verhältnis. Ahnliche Ergebnisse wurden erhalten, wenn der Körper durch kurzdauernde Exposition in Gegenwart von Thorium radioaktiv gemacht wurde, ohne dass ein elektrisches Feld wirksam war. Dieses Wachsen der Radioaktivität mit der Zeit ist unabhängig von der Natur der Elektrode und von dem Grade dei" Konzentration der Radioaktivität. Wenn die Platte oder der Draht einige Stunden ausgesetzt wird, ehe man ihn entfernt, ist die nachfolgende Steigerung der Radioaktivität sehr klein; für eine noch längere Expositionszeit fällt nachher die Radioaktivität

372

The Collected Papers 0/ Lord Ruther/ord

beständig. Dieses Fallen nach langen Expositionen darf man erwarten, wenn die Steigerung des Strahlungsvermögens der radioaktiven Substanz, die in den letzten paar Stunden der Exposition niedergeschlagen wird, mehr als kompensiert wird durch die Abnahme der Strahlung des Restes. Diese Ergebnisse erklären eine auffällige Anomalie, die früher (loc. cit. S. 178) in dem Wachsen der von Thorium erregten Radioaktivität mit der Expositionszeit beobachtet worden war. Es war dort gezeigt worden, dass unter der Annahme eines gleichförmigen Niederschlages von radioaktiver Substanz, die an der erregten Strahlung schuld ist, und einer regelmässigen Verminderung der Strahlungsintensität mit der Zeit, das Anwachsen der erregten Radioaktivität mit der Expositionszeit dem Anwachsen eines elektrischen Stromes gleich ist, der in einem Stromkreise von konstanter Selbstinduktion entsteht. Während dies zu einer Erklärung der experimentellen Ergebnisse im grossen und ganzen ausreichte, wurde beobachtet, dass der Betrag der erregten Radioaktivität in den ersten paar Stunden der Exposition sehr viel kleiner war, als es der Theorie entsprach. Diese Abweichung erklärt sich indessen, wenn jeder Anteil der radioaktiven Substanz einige Stunden braucht, um seine grösste Strahlungsfahigkeit zu erreichen. Wenn wir die Anschauung zu Grunde legen, dass erregte Radioaktivität von dem Niederschlag einer irgendwie beschaffenen radioaktiven Substanz auf den Körpern herrührt, so hat es den Anschein, dass entweder 1. die Strahlung einer allmählichen molekularen Umlagerung oder chemischen Kombination zugeschrieben werden muss, welche einige Stunden brauchen, um ihre maximale Intensität zu erreichen, oder dass 2. die niedergeschlagene radioaktive Substanz in dem Drahte oder der Platte erregte Radioaktivität veranlasst, die sich zu ihrer eigenen ursprünglichen Radioaktivität hinzu addiert. Es wurden auch Versuche gemacht, indem die strahlende Elektrode ungefähr auf Rotglut erhitzt wurde; indessen war es nicht möglich, in dieser Weise die Zeit zu verkürzen oder zu verlängern, die bis zur Erreichung des Maximums des Strahlungsvermögens notwendig war, noch auch, das endgültige Maximum merklich zu beeinflussen. Von Radium erregte Radioaktivität Es wurden einige Versuche gemacht, um zu erfahren, ob das nachträgliche Wachsen des Strahlungsvermögens auch für die von Radiumverbindungen erregte Radioaktivität beobachtet werden kann. Ein Platindraht wurde in einem Gefässe mit Radiumausströmung 10 Minuten lang zur Kathode gemacht. Dann wurde er entfernt und die Änderung seiner Radioaktivität mit der Zeit geprüft. Die zwei Proben von Radium, die von P. de Haen, Hannover stammten, gaben verschiedene Abfallkurven. Die Abfallkurve der als "konzentriert" bezeichneten Probe (Fig. 1) war sehr unregelmässig, kann aber in drei Teile geteilt werden:

Versuche über erregte Radioaktivität

373

I. Ein anfänglich sehr schneller Abfall der erregten Aktivität für ungefähr

10 Minuten. 2. Eine sehr langsame Änderung für die nächsten 30 Minuten etwa. 3. Eine schnellere Abnahme, bis die Radioaktivität verschwunden ist. Die Abfallkurve für das als "einfach" bezeichnete Radium war nicht so unregelmässig, aber der allgemeine Verlauf wurde ebenso gefunden, nur in einer weniger ausgeprägten Form. Für die ersten paar Minuten fiel die Radioaktivität sehr schnell, dann langsamer und schliesslich wieder schneller.

IVO

80 7011

~ ~ ~

7011 ~

·i

6011":3

so "

~

40

30

ZD ID

Yon"JUumirUumfolie ID

30

30

30

30

30

I~O

Fig. 1 Um die Gestalt der in Fig. 1 wiedergegebenen Kurve des Abfalles der von Radium erregten Strahlung zu erklären, scheint es nötig, anzunehmen, dass die radioaktive Substanz, die auf den erregten Körper übertragen wird, wenigstens zwei Arten von Strahlungen von sich giebt. Die eine derselben nimmt sehr schnell mit der Zeit ab. Die Intensität der anderen wächst einige Zeit, nachdem die radioaktive Substanz niedergeschlagen ist, ähnlich wie das bei Thorium beobachtet wurde, nachher aber fällt sie in regelmässiger Weise. Eine solche Hypothese würde die allgemeine Form der Abfallkurve für die von beiden Proben des Radiums erregte Radioaktivität erklären. Der relative Betrag derjenigen Strahlung, die schnell abnimmt, verglichen mit der anderen, scheint bei verschiedenen Proben von Radium verschieden zu sein.

374

The Collected Papers

0/ Lord Ruther/ord

Wirkung von Lösungen Es war früher gezeigt worden (1. c.), dass Schwefel- und Salzsäure einen Teil der erregten Radioaktivität wegnehmen, die an einem Thorium ausgesetzten Platindrahte erregt wurde. Die so entfernte Radioaktivität bleibt in der Lösung. Wenn man die Lösung bis zur Trockenheit eindampft, so wird die Radioaktivität auf die Wand des Gefässes übertragen. Keine anderen Substanzen haben sich finden lassen, die so lebhaft die erregte Radioaktivität auflösten, wie Schwefelsäure, Salzsäure und Fluorwasserstoffsäure, obwohl eine grosse Zahl organischer und unorganischer Säuren und Lösungen geprüft worden ist. Bei einigen der früheren, zwei Jahre zurückliegenden Experimente war gefunden worden, dass 1/ 10 normale reine Schwefelsäure die gesamte auf einem Platindrahte erregte Radioaktivität bis auf 8 Proz. in wenigen Minuten entfernte. Verdünnte Salzsäure des Handels entfernte sie in wenigen Sekunden bis auf 10 Proz. Unsere Versuche mit verschiedenen Proben reiner und käuflicher Säure gaben sehr verschiedene Ergebnisse. Beispielsweise wurden mit einer anderen Probe von reiner Schwefelsäure ungefähr 50 Proz. sehr schnell entfernt, der Rest aber wurde äusserst langsam gelöst. Käufliche Salzsäure vermindert die Radioaktivität bis auf 29 Proz. Die entfernte Menge wurde durch beträchtliche Verdünnung der Säure nicht wesentlich verändert. Käufliche Salz- und Schwefelsäure findet sich sehr viel wirksamer, die Radioaktivität zu entfernen, als die ganz reinen Säuren. Die Bedingungen, unter denen der Platindraht radioaktiv gemacht worden war, hatten keinen besonderen Einfluss auf das Ergebnis. Die Expositionszeit, die Gegenwart oder Abwesenheit von Wasserdampf bedingten keinerlei Unterschied. Die grosse Verschiedenheit der Fähigkeit, erregte Radioaktivität zu entfernen, wie sie bei verschiedenen Proben reiner Schwefelsäure und bei käuflicher und reiner Säure auftritt, scheint zu dem Schlusse zu drängen, dass die Entfernung der erregten Radioaktivität von dem Platin von einer geringen Verunreinigung herrührt, die in verschiedenen Proben der Säure mit verschiedenem Betrage vorhanden ist. Abnahme der erregten Radioaktivität in Schwefelsäure Es wurden Versuche unternommen, um zu erkennen, ob die einer verdünnten Lösung von Schwefelsäure mitgeteilte Radioaktivität mit derselben Geschwindigkeit wie in Luft abnimmt. Ein radioaktiv gemachter Platindraht wurde für wenige Minuten in 1/10 normale Schwefelsäure Lösung getaucht, die in einer Bürette von 12 cm 3 enthalten war. Nach der Entfernung des Platindrahtes wurde die Säure tüchtig geschüttelt und einige Tage sich selbst überlassen. Zu verschiedener Zeit wurden 2 cm 3 derselben Lösung entfernt und in einer Platinschale eingedampft. Die Radioaktivität in der Schale wurde in der gewöhnlichen

Versuche über erregte Radioaklirilät

375

Weise mit Hilfe eines Elektrometers geprüft. So wurden die Beträge der Radioaktivität, die in gleichem Volumen der Lösung zurückblieben, nach verschiedenen Zeitintervallen bestimmt. Die Ergebnisse zeigten, dass die Abnahme der erregten Strahlung in der Lösung nahezu dieselbe war, als wenn sie an dem Platindrahte der Luft ausgesetzt worden wäre. Dieses Ergebnis in Verbindung mit anderen ähnlicher Art scheint zu zeigen, dass die Abfallgeschwindigkeit der erregten Radioaktivität einem in der radioaktiven Substanz selbst verlaufenden Vorgange zuzuschreiben ist, der von der Substanz, mit der sie in Berührung ist, nicht beeinflusst wird. Einfluss der Temperatur Die an einem Platindrahte von Thorium erregte Strahlung wird durch eine Steigerung der Temperatur bis zur Rotglut wenig beeinflusst. Ein grosser Betrag der Radioaktivität aber wird rasch entfernt, wenn man bis zur Weissglut erhitzt. Eine lang andauernde Hitze auf hohe Temperatur ist notwendig, wenn man auch den Rest entfernen will. Der Teil, der durch Hitze schwer entfernt wird, wird auch durch Lösungen von Salz- oder Schwefelsäure sehr wenig angegriffen. Keine bemerkenswerte "radioaktive Ausströmung" wurde von einem erregten Platindrahte erhalten, während er auf Weissglut erhitzt wurde, welche ihr radioaktives Strahlungsvermögen sehr bald zerstörte. Es sind Versuche im Gange, um zu zeigen, ob hierbei die erregte Radioaktivität zerstört oder nur auf die umgebenden Körper übertragen wird. McGiII Univ., Montreal, den 14. Jan. 1902. (Aus dem Englischen übersetzt von H. Th. Simon) (Eingegangen 14. Februar 1902)

The Radioactivity of Thorium Compounds I

AN INVESTIGATION OF THE RADIOACTIVE EMANATION

by E. R UTHERFORD, M.A., D.SC., Macdonald Professor of Physics, and FREDERICK SODDY, B.A. (OXON), Demonstrator in Chemistry, M cGill University, M ontreal From the Transactions ofthe Chemical Society, 1902,81, pp. 321-350

THE following paper contains a preliminary account of an investigation into

the property possessed by the compounds of thorium of giving a radioactive emanation, and also into the nature of the emanation itself. It was shown by one of us (Phi!. Mag., 1900 [v], 49, 1, 161) that the compounds of thorium, besides being radioactive in the same sense as the uranium compounds, also continuously emit into the surrounding atmosphere, under ordinary conditions, something which, whatever its real nature may be, behaves in all respects like a radioactive gas. This 'emanation', as it has been named, is the source of rays, whieh, like the Röntgen and uranium rays, and the ordinary well-recognized type of thorium radiation, will darken a photographie plate, and will render agas capable of eonducting an electric current (that is, will 'ionize' it), but is sharply distinguished from them by the following considerations. It can be moved from the neighbourhood of the thorium compound by a current of air passing over it, or even by the process of ordinary gaseous diffusion, and transported long distances, so that the characteristic photographie and ionization effeets appear in the air far away from the original source of radioactivity. The Röntgen and uranium rays, as is weIl known, travel in straight lines from their source, and any object opaque to them interposed in their path will sharply screen the space behind. But in the case of the thorium radiation there is no such screening effect, because here we have a case of a substance emitting, not only straight line radiation, but also particles of agas, itself radioactive, capable of diffusing through the surrounding atmosphere around obstac1es placed in its direct path, and so arriving and producing its effects at points completely screened from rays travelling from the thorium in straight lines. It was shown in the original communication that these effects could not be ascribed to minute particles of thoria dust carried off mechanically, and all the subsequent work on the subject shows that the hypothesis that the

The Radioaclil'ily 41'l/O/'ium Compoullds. 1

377

compounds of thorium emit a radioactive gas is not merely the only one which will explain the facts, but that it does so in every observed case in a completely satisfactory manner.

Present State of the Subject from a Physical Standpoint In the papers referred to, the general character of the phenomena in question was presented, and a short resume will perhaps not be out of place here. It was shown that the radiation from the emanation decays rapidly, but at a perfect1y defined rate, that is, the effects it produces diminish with the lapse of time, falling to about one half the original value at the end of one minute. This 'rate of decay', as will be shown later, is of great value in identifying and distinguishing between different types of emanation. The emanation passes unchanged through cotton wool, weak and strong sulphuric acid, and aluminium and other metals in the form of foi!, but not through an extremely thin sheet of mica. The emanating power of thoria is independent of the surrounding atmosphere, but is destroyed to a large extent by intense ignition, and does not return when the substance is kept. One of the most striking properties of the thorium emanation is its power of exciting radioactivity on all surfaces with which it comes in contact, that is, a substance after being exposed for some time in the presence of the emanation behaves as if it were covered with an invisible layer of an intensely radioaetive material. If the thoria is exposed in a strong electric field, the excited radioaetivity is entirely confined to the negatively charged surface. In this way, it is possible to concentrate the exeited radioactivity on a very small area. The excited radioaetivity itself has a regular rate of decay, but different from that of the emanation, its effeet falling to half value in about eleven hours. There is a very elose conneetion between the excited radioaetivity and the emanation. It was shown that the amount of the former produced under various conditions was proportional to the amount of the latter, and if the emanating power of thoria be destroyed by intense ignition, its power to exeite radioactivity eorrespondingly disappears. Some apparent diserepancies whieh at first stood in the way of too elose a conneetion being inferred have resolved themselves by reeent work into strong confirmation of the view that the two are related to each other as eause and effect. Another remarkable property of the excited radioactivity is that it is soluble in sulphuric and hydrochloric acids, that is, a platinum wire, rendered radioaetive by being made the negative pole of an electric field in the neighbourhood of some thoria, will give up its radioactivity to these acids. If the acid be then evaporated, the radioactivity remains on the dish, whilst, if left to itself, the radioaetivity of the acid solution decays at a rate identical with that of the original excited radioaetivity on the platinum wue.

378

The Collected Papers o[ Lord Rutherfard

Simultaneously with the discovery of excited radioactivity due to thoria, Curie showed that radioactive barium possessed a similar property. Later, Dorn (Abh. der Natur/orsch. Ges. für Halle-a-S., 1900) repeated the work quoted for thoria, and extended it to include two preparations of radioactive barium compounds (radium) prepared by P. de Haen, and apreparation of radioactive bismuth (polonium). He found that radium gave out an emanation which was similar to that from thoria, but which retained its radioactive power much longer. The excited radioactivity from radium, on the other hand, decayed more rapidly than that from thoria. The special property of emitting an emanation is, however, confined to the compounds of radium and thorium, those of uranium and polonium do not possess it to an appreciable extent. An approximate determination of the molecular weight of the emanation produced by radium has been carried out (Rutherford and H. T. Brooks, Nature, 1901, 64, 157) by a diffusion method, taking advantage of the slow rate of decay of the radium emanation. From comparison of the rate of diffusion of gases of known molecular weight into one another, it was found that the molecular weight probably lies between 40 and 100. It seemed probable that an examination of the phenomena by chemical methods might throw light upon its nature, and the emanation produced by thoria was chosen as more suitable for the purpose than that produced by radium, on account of the obscurity still surrounding the chemistry of the latter, and the difficulty of producing material of even approximate uniformity of properties. Thoria, on the other hand, is an article of commerce, and specimens from different sources show surprising uniformity in this respect. During the progress of the work, the subject has acquired additional importance and interest through the discovery by Elster and Geitel (Phys. Zeit., 1901, 2, 590) that it is possible to produce excited radioactivity from the atmosphere, without further agency, by simply exposing a wire highly charged to a negative potential in the atmosphere for many hours, and that this also possesses the property of being dissolved off by acids, and of being left behind unchanged on the evaporation of the latter. But here again the rate of decay is different from that of the excited radioactivity produced by thoria, which is evidence for assuming that the two are probably not identical, although so strikingly analogous. However, the elose connection between excited radioactivity and the emanation established in the case of thoria renders it probable that the excited radioactivity obtained from the atmosphere is caused by the presence there of an emanation or radioactive gas analogous to, although probably different from, the thorium emanation. The discovery is likely, as Elster and Geitel pointed out, to have important bearings on the theory of atmospheric electricity and, in our opinion, renders a elose study of the thorium emanation the more imperative.

Ihe RadiollC'liI>ity

0/ Thorium Compmmds. 1

379

The Chemieal Aspect 0/ the Question The foregoing furnishes a short review of the physical side of the question at the present time. With regard to the chemical aspect, this has so far not been studied. The photographic method, almost the only one that has until now been used by chemists in the study of radioactivity, is not one which allows of the recognition and differentiation of an emanation as a component factor in producing the phenomena. The photographic method is of a qualitative rat her than a quantitative character; its effects are cumulative with time, and as a rule long exposures are necessary when the radioactivity of a feeble agent like thoria is to be demonstrated. In addition, Russell has shown that the darkening of a photographic plate is brought about also by agents of a totally different character from those under consideration, and, moreover, under very general conditions. Sir William Crookes (Proe. Roy. Soc., 1900, 66, 409) has sounded a timely note of warning against putting too much confidence in the indications of the photographic method of measuring radioactivity. The uncertainty of an effect produced by cumulative action over long periods of time quite prec1udes its use for work of anything but a qualitative character. Two or three chemists have studied the radioactivity of thoria, using thc photographic method, without, however, distinguishing between thc radioactivity due to the emanation and that due to the thoria itself. Sir William Crookes, loc. eil" who suceeeded by an elegant method in separating and isolating the radioactive constituent of uranium, also describes some experiments on thorium eompounds with the same object, but did not succeed in effecting aseparation. A method based on the fraetional precipitation of the sulphate failed completely, but another method, the fraetional crystallization of the nitrate, gave preparations showing a difference in their photographie aetions in the ratio of one to three. According to slight variations in thc method employed, as, for example, whether a glass or a card bottom was used for the cell eontaining the substanee to be tested (and both seem to have been employed), the radiation from the emanation would or would not contribute largely to the photographic aetion observed. Debierne (Comptes Rendus, 1900, 130, 906), working on a very large seale, obtained from pitehblende, by using reactions which would lead to the separation of thorium, a material different in its chemical properties from radium (barium) and polonium (bismuth), but consisting in great part of thorium. This preparation was 100,000 times more active than uranium, and he therefore assumed the existenee of a new element, 'actinium', therein. He hazarded the suggestion that the radioactivity of thoria is due to the presence of the same substance, and derived support for this view from the reeent work of one of us on the radioactivity of thoria, although on what grounds is not c1ear. In the course of their work on the atomic weight of thorium, Brauner (Trans., 1898, 73, 951) and Baskerville (J. Amer. ehem. Soc., 1901, 23, 761),

380

The Collected Papers of Lord Rutherford

have obtained evidence of the presence of a foreign substance associated with thorium. The latter noticed that the separation, as he interpreted it, of this impurity reduced the photographic action considerably, and he concluded that the pure material would be without photographic action. He employed a modification of Crookes' photographic method, but it cannot be decided with certainty from the description whether the radiation from the emanation would be eliminated or not. The present work is concerned primarily with the radioactive emanation, although, of course, frequent occasion has arisen to examine correspondingly the ordinary radiation also. The methods employed are of an electrical character, based on the property generally possessed by all radiation of the kind in question, of rendering agas capable of discharging both positive and negative electricity. These, as will be shown, are capable of great refinement and certainty. An ordinary quadrant electrometer is capable of detecting and measuring a difference of potential of at least 10-2 volts. With special instruments, this sensitiveness may be increased a hundredfold. An average value for the capacity of the electrometer and connections is 3 x 10-s microfarads, and when this is charged up to 10-2 volts, a quantity of electricity corresponding to 3 x 10-13 coulombs is stored up. Now in the electrolysis of water one gram of hydrogen carries acharge of lOS coulombs. Assuming, for the sake of example, that the conduction of electricity in gases in analogous to that in liquids, this amount of electricity corresponds to the transport of a mass of 3 x 10-18 grams of hydrogen, that is, a quantity of the order of 10-12 times that detected by the balance. For a more delicate instrument, this amount would produce an inconveniently large effect. The effects under investigation, from the nature of their manifestation, may weIl be, and probably are, produced by quantities of matter of the order of magnitude described, and therefore altogether beyond the range of the balance. But to assume on that account that the subject is beyond the pale of profitable chemical investigation is needlessly to limit the field of chemical inquiry. Although surpassing the spectroscope as a detective agent, as a quantitative instrument the electrometer is Httle inferior in accuracy to the balance. To take as an example the case of thoria mixed with zirconia, the former could be detected and accurately measured by means of its emanation with an electrometer, even although it were only present to the extent of one part in many thousands. A distinction must be made here between emanation and emanating power. The quantity of the former is what is measured by the electrometer. To express this in terms of weight, the emanating power, that is, the quantity of emanation produced by a given weight of the substance in question, must be known. As will be shown, this value varies with the previous history and present condition of the substance. The electrometer also affords the means of recognizing and differentiating between the emanations of different chemical substances. By the rate of decay, the emanation from thorium, for example, can be instantly distinguished from that produced by radium, and although a difference in the

Th('

Radioacti~'ily

of T/l0/'ium C ompolillds. I

381

rate of decay does not in itself argue a fundamental difference of nature, the identity of the rate of decay furnishes at least strong presumption of identity of nature. In the sense that has just been explained, the electrometer can be said to supply the investigation of the property of emanation with methods, so to speak, of quantitative and qualitative analysis which are simple and direct, and there is, therefore, no reason why the property in question, and even the nature of the emanation itself, should not be the subject of chemical investigation.

Scope

0/ Work

Of the great number of questions which immediately present themselves for answer in an investigation of this kind, the following are at present claiming our more immediate attention: (1) Is the power of producing an emanation a specific property of thorium, or is it to be ascribed to the presence of a foreign substance, possibly in minute amount, associated with it and amenable to chemical methods of separation? (2) Can the emanating power of 'de-emanated' thoria be regenerated by chemical means? It has been mentioned that thoria, when intensely ignited, loses, to a very great extent, its power of giving an emanation. If such deemanated thoria be subjected to aseries of chemical changes, will it regain its emanating power or not? (3) Does the emanation or radioactive gas itself possess any property which would associate it chemically with any known kind of gravitational matter? (4) Is it possible to detect, by means of the balance, any loss in weight corresponding to the continuous emission of the emanation or any gain in weight of bodies rendered radioactive thereby? (5) Does the chemistry of thorium present any peculiarity capable of being connected with its almost unique power of producing an emanation?

To interpret rightly the results obtained, a more or less complete study of the effect of chemical and physical conditions on the emanating power is necessary. The effect of the state of aggregation, the presence or absence of water, the infiuence of light, temperature, the nature of the surrounding atmosphere, the lapse of time since preparation, etc., on the emanating power, as weB as the differences in this property exhibited by different compounds, have been investigated. The present communication does not attempt a full answer to all the above questions. The results so far obtained in answer to the first three will be presented. The work on the fourth is in progress, whilst the results of the investigation of the fifth question will be most conveniently given Iater in a separate communication.

382

The Collected Papers of Lord RlIlherford

Electrometer Method 0/ Measuring Emanating Power and Radioactivity The term 'radioactive' is now generally applied to a c1ass of substances, like uranium, thorium, radium, and polonium, which have the power of spontaneously giving off radiations capable of passing through thin plates of metal. The radiations are in some cases very complex, but in the case of the substances mentioned, a portion at least of the radiation is similar in all respects to easily absorbed Röntgen rays. The characteristic and general property possessed by these radiations is to produce, from the gas through which they pass, positively and negatively charged carriers, which in an electric field, travel to the negative and positive electrodes respectively. In this way, a small current is able to pass through agas exposed to the radiations, even with a very weak electric field, and the measurement of this current by means of the electrometer affords a means of comparing the intensities of radiation. As has been mentioned, compounds of thorium (and radium), in addition to radiations which travel in straight lines, emit radioactive emanations, which behave in all respects like a temporarily radioactive gas, and diffuse rapidly through porous substances, as, for example, thick cardboard, which are completely opaque to straight line radiation. Each partic1e of the emanation behaves as if it were a radiating centre, producing charged carriers throughout the gas in its neighbourhood. The emanation passes through plugs of cotton wool and can be bubbled through liquids without appreciable loss of radioactivity, whereas the charged carriers, produced by the emanation in common with the straight line radiation from radioactive substances, on the contrary, completely disappear on passing through a plug of cotton or glass wool, or by bubbling through liquids. The means of eliminating the effects of the straight line radiation and of measuring the amount of the emanation alone thus suggest themselves. Air passed over uranium or other non-emanating radioactive substances will no longer conduct a current after passing through cotton wool. The conductivity in the case of thorium, however, will persist, and afford a measure of the amount of emanation present. Fig. 1 shows the experimental arrangement for comparing the emanating power of substances. These are placed in the form of fine powder in a shallow lead vessel inside the glass cylinder, C, 17 cm. in length and 3·25 cm. in diameter, provided with indiarubber corks. A current of air from a large gas-bag, after passing through a tube containing cotton wool to remove dust partic1es, bubbled through sulphuric acid in the vessel, A. It then passed through a bulb containing tightly packed cotton wool to prevent any spray being carried over. The emanation mixed with air was carried from the vessel C through a plug of cotton wool, D, which completely removed all the charged carriers carried with the emanation. The latter then passed into a long, brass cylinder 75 cm. in length and 6 cm. in diameter. The cylinder insulated on paraffin blocks was connected to one pole of a battery

111(' Radioaetil'ity of Thorium Compounds. I

3~3

of smalliead accumulators, the other pole of which was connccted 10 earth. Three electrodes, E, F, H, of equal length were placed along the axis of the cylinder, supported by brass rods passing through ebonite corks in the side of the cylinder. The current through the gas, due to the presence of the emanation, was measured by me ans of a Kelvin quadrant electrometer of the White pattern. The electrometer and the connections were suitably screened by me ans of wire gauze connected to earth. An insulating key was arranged so that either of the electrodes E, F, H, or all of them together, could be rapidly connected to one pair of quadrants of the electrometer, the other two being always connected to earth. en Lasen

sen

r

c:

r

N

c:

A

sen

Fig. 1 The insulation of the electrodes was first tested by sending a current of air through the apparatus without any emanating material in C. The rate of movement of the electrometer needle was accurately observed. On placing the emanating substance in C and continuing the air current for several minutes at a constant rate, the current due to the emanation reached a steady state. On separating the quadrants of the electrometer, the deflection from zero increased uniformly with time. The time taken to pass over 100 divisions of the scale was observed with a stop-watch. The number of divisions passed over per second may be taken as a measure of the current through the gas. With this apparatus, the emanation from 10 grams of ordinary thorium oxide produces a current of 3·3 X 10- 11 ampere between the three electrodes connected together and the cylinder. With the electrometer working at average sensitiveness, this corresponded to adefleetion of 100 divisions of the scale in 12 sec., so that one-hundredth part of this current could be readily measured, that is, the emanation produced by one-tenth of a gram of thorium oxide. An electrometer one hundred times more sensitive than this failed to detect the presence of an emanation or radioactivity in the oxides of tin, zirconium, and titanium, the other elements of the same group in the periodic table. Variation 0/ the Current with Va/tage. The current through the gas observed with the electrometer at first increases with the voltage, but a stage is so on reached when there is a very small increase for a large additional

384

Tlle Collected Papers o..tLord RutIleilord

voltage. This is one of the most characteristic properties of conducting gases. For small voltages, only a small proportion of the charged carriers reach the electrode, on account of their recombination throughout the volume of the gas. When the electric field is increased untiI all the carriers reach the electrode before any appreciable recombination can occur, the current is at a maximum, and remains constant for large increases of voltage, provided, of course, that the electric field is below the value necessary for a spark to pass. In the experimental case, apressure of 50 volts was found sufficient to give the maximum current between the electrodes. This property of conducting gases allows us at once to make sure that the insulation of the electrodes is perfect at all stages of a long experiment; 100 volts applied instead of 50 to the cylinder should give the same current if the insulation is unaffected. Rate of Decay of the Radiation from the Emanation. The three electrodes, E, F, H, were used to compare the 'rates of decay' of the radiations from the emanations of different substances. In the previous papers quoted, it has been shown that the radiating power of the thoria emanation falls to half its value in about a minute. In consequence of this, the current observed for the electrode E is greater than for electrode H. Knowing the velocity of the current of air along the cylinder and the respective currents to the electrodes E, F, H, the rate of decay of the radiation can be readily deduced. If, however, we merely require to compare the rate of decay of one emanation with another, it is only necessary to compare the ratio of the currents to the electrodes E, F, H in each case, keeping the current of air constant. lf the ratio of the currents is the same we may conc1ude that the radiating power of each diminishes at the same rate. The comparison of emanation is thus rendered qualitative as weIl as quantitative. In most of the experiments, the current to the electrode E, was about twice that to the electrode H; the velocity of the current of air along the cylinder was thus about 0·8 cm. a second. Comparison of Emanating Power. The experiments in all cases on the amount of emanation from different substances are comparative. The standard of comparison was usually a sampie of 10 grams of thoria as obtained from the maker, which gave out a conveniently measurable quantity of emanation. Preliminary experiments were made to find the connection between the weight of thoria and the amount of emanation as tested in the cylinder. The following numbers show that the amount of emanation is directly proportional to the weight of the substance: Weight of thoria

2 grams 4 grams 10 grams 20 grams

Divisions of scale per second

1-41 2-43

6·33 13·2

FREDERICK SODDY

This page intentionally left blank

Ihe RadioQetirity of 71lOriulIl Compouncls. I

3~5

This result shows that within the limit of accuracy desired we may take the amount of emanation as directly proportional to the weight of the substance. The determinations in the above table were made with the three electrodes connected together with the electrometer, and with a constant ftow of air. The lead vessel in which the thoria was placed was 7· 4 by 3· 5 cm. in area and 3 mm. deep. In the comparison of emanating power, the maximum current between the electrodes for the standard 10 grams of thoria was first observed. This was removed and a known weight of the specimen to be tested was substituted, and the deftections again observed after the conditions had become steady. If d, = No. of divisions per second for a weight, lt'" of thoria; d2 = No. of divisions per second for a weight, »'2, of the specimen; Emanating power of specimen d2»'1 then Emanating power of thoria = d1W2' The values d 1 and d2 are corrected, when necessary, for natural leakage, that is, the current which passes under similar circumstances when no emanating material is present. This current is chiefty made up of a leakage due to conduction over the ebonite, as weIl as the current produced by tbe excited radioactivity which has collected on the negative electrode during the course of the day's experiments. It is gene rally very smalI, and the correction is only necessary when a specimen of substance almost free of emanation is being tested. An example taken at random from the notebook will serve to illustrate the method of calculating the results, the emanating power of the comparison sampie being considered 100 per cent: Dec. 7th, 11 a.m.-Naturalleakage

5 grams comparison sampie Th02 3·6 grams Th02 ignited 24 hours over Bunsen burner in platinum crucible d, = 4· 25, corrected for nato leakage = d2 = 1·42 =

10 divisions in 50" 0·20 divisions in 1" 100 divisions in 23·5" 50 divisions in 35·2" 4·05 1·22

d2W l = 0,42, or 42 per cent. d'»'2

Comparison o/lntensity 0/ Straight Line Radiation It was frequently of interest to obtain information about the intensity of the ordinary radiation correspondingly with measurement of emanating power. In order to do this rapidly and accurately, the following method was used. Fig. 2 shows the general arrangement. 0·1 gram of the compound to be tested was reduced to fine powder and uniformly sifted over a platinum plate 36 sq. cm. in area. N

386

l1le Collected Papers

0/ Lord Ruther/ord

This plate was placed on a large metal plate connected to one pole of a battery of 300 volts, the other pole of which was earthed. An insulated parallel plate was placed about 6 cm. above it, and the whole apparatus enc10sed in a metal box connected to earth, to prevent electrostatic disturbance. The shaded portions in the figure represent insulators. A door was made in the apparatus so that the plate could be rapidly placed in position or removed. The current between the plates is observed in the usual way with the electrometer. The ratio of the currents for two substances is a comparative measure of their radioactivity. It is only possible to compare together with certainty substances of similar density and state of division -a light, floury material will tend to give lower values than a dense powder.

Earth

R~dioactive

Material Earth

Fig.2 If a substance gives off an emanation, the current between the plates increases with time. Under these conditions when the thoria is exposed in thin layers with a maximum of radiating surface, all but I or 2 per cent of the total effect is due to the straight line radiation; even when the effect due to the emanation has attained its maximum, this' constitutes a very small percentage of the whole. This effect, however, may be to a large extent eliminated by taking the current between the electrodes immediately after the material is placed in the testing apparatus, or by passing a current of air between the electrodes to remove the emanation, and prevent it accumulating. It is thus possible to compare intensity of radiations with an error not exceeding 1 or 2 per cent, and with great rapidity, and in these respects the electrical method is altogether superior to the photographic. Comparison of Emanating Power. The apparatus (Fig. 2) described for the comparison of radiations, can also be quite weIl employed for a comparison

The Radioactivity 0/ Thorium Compounds. I

387

of emanating power. In this case, a thick layer of thoria (several grams) is spread over the plate and covered with two thicknesses of ordinary paper. This has been found almost completely to stop the straight line radiation, whilst allowing the emanation to pass through unimpeded. The current is now measured when a steady state has been reached, due to the accumulati on of the emanation. This takes some time, and draughts of air must be guarded against. For this reason, it is less convenient than the method first described, but the results obtained by the two methods are almost exactly the same. Thus a sampie of 'de-emanated' thoria which gave 12 per cent of the emanating power of the comparison sampie by the first method gave 13 per cent by the second method, whilst a sampie of oxide prepared from thorium oxalate gave 37 per cent and 39 per cent by the two methods respectively. The elose agreement in the values by methods so completely different in character is a proof that the indications of the methods are worthy of a great degree of confidence.

The De-emanation

0/ Thoria and the Regeneration 0/ the Emanating Power

The emanating power of thoria, as has been stated, is destroyed to a large extent by intense ignition. A eIoser study of this is the first step in the investigation of the phenomenon. Previous experiments had not succeeded in completely de-emanating thoria, although a reduction to about 15 per cent of its original value had been accomplished. A sampie of this preparation which had been kept for two years had not altered from this value. An experiment was performed in which thoria was heated to the highest temperature which could be safely employed with platinum vessels: (1) in a thin layer in a large platinum dish, and (2) in bulk in a small platinum crucible placed inside the dish. The two were heated together by means of a powerful gasoline furnace for one hour. The temperature was such that the fireeIay walls fused, and the pipec1ay of a tri angle showed signs of having been softened. It was found that the sampie that had been heated in a thin layer in the dish retained about 16 per cent of its original emanating power, whilst the other sampie retained about 8 per cent. There is thus no advantage in heating in thin layers, in fact rather the reverse, for the sampie showing 16 per cent again heated to a slightly lower temperature for half an hour in a small crucible was reduced to 12 per cent. In another experiment, a small platinum crucible fi1led with thoria was heated for half an hour in a small furnace by a large blowpipe and powerful pair of bellows. Some asbestos wool had completely fused on the outside of the crucible, and the temperature was probably but little lower than in the previous experiment. This sampie also retained about 8 per cent of its emanating power. No further attempt has yet been made to completely destroy the emanating power.

388

The Collected Papers

0/ Lord Ruthe/ford

A smaU quantity of thoria heated in a platinum crucible in the open over an ordinary small-sized blowpipe and bellows for five minutes retained about 45 per cent of its emanating power. The effect of time as weIl as of temperature was studied by heating about equal quantities in a platinum crucible over an ordinary Bunsen burner for different periods. Heated 10 minutes Heated 1 hour Heated 24 hours

Emanating power = 61 per cent Emanating power = 59 per cent Emanating power = 42 per cent

It thus appears that there is a large and practically sudden decrease of emanating power for each temperature above a red heat, followed by a very gradual decrease with time when the temperature is maintained; thus five minutes on the blowpipe, whilst much more effective than the same time at the temperature of the Bunsen burner, produced rather less effect than 24 hours at the latter temperature. Effect 0/ Moisture. The next point to be examined was whether the loss of emanating power could be attributed to a loss of water and desiccation of the thoria by ignition. A sampIe of de-emanated thoria (retaining about 14 per cent) was placed in the middle of a Jena glass tube, one end of which was c10sed and contained water, the other end being drawn out to a jet. This was supported in a powerful tube furnace in a sloping position, and the part containing the thoria heated to the highest possible temperature, while a slow current of steam from the water at the end was passed over it, escaping by the jet. When all the water had evaporated, the jet was drawn off and the tube allowed to cool in an atmosphere of steam free from air. The thoria, on testing, was found to have been lowered in emanating power to about 7 per cent. The further heating had thus reduced the emanating power without the steam having at all regenerated it. In the next experiment, the reverse process was tried. Two exact1y parallel processes were carried out for ordinary thoria possessing the normal amount of emanating power. In the first, it was heated in a porcelain tube in the tube furnace for three hours, while about 500 C.c. of water were distilled over it from a retort. In the second, another quantity of thoria was heated in exact1y the same way for the same time, only a current of weIl dried air was substituted for the steam. The result was conclusive: each sampIe had had its emanating power reduced to exact1y the same amount, that is, about 50 per cent of the original. These experiments prove that water vapour exerts no infiuence either in de-emanating thoria or in effecting a recovery of its lost emanating power. The Regeneration 0/ the Emanating Power by Chemical Processes. The task of subjecting de-emanated thoria to aseries of chemical changes to see if it would recover its lost emanating power was then undertaken. Jt may first be mentioned that thoria which has been subjected to ignition

The Radioaclil'ity of Thorium Compounds. I

3g9

has changed very materially in chemical and physical properties. The pure white colour changes at temperatures corresponding to the first stages of de-emanation to a light brown, and after subjection to the very highest temperatures to a pure pink. At the same time, as has been observed before, the solubility of the substance in sulphuric acid is greatly diminished. Apart always obstinately refuses to dissolve, even after long and repeated boiling with the concentrated acid, although this part is diminished by each successive treatment and appears to be in no way different from the rest of the substance. No difference, however, occurs in the readiness with which chlorine attacks it when intimately mixed with carbon. The formation of the chloride by this method is the easiest way of dissolving ignited thoria. Two quantities of the same de-emanated thoria were converted, the one into chloride and the other into sulphate, by the usual methods, and from each of these the oxalate was formed by precipitation of the acid solution with oxalic acid. Parts of the oxalates were then converted into oxides by heating over the Bunsen burner. In both cases there was a very marked recovery in emanating power; the oxide obtained from the sulphate had about 40 per cent, that from the chloride about 55 per cent, whereas the deemanated thoria from which they were both produced had about 13 to 14 per cent of the emanating power of thoria. The oxalates from which the oxides were formed each had about 11 per cent of the power, and in converting them into oxides it was ascertained by a direct trial that too high a temperature had been employed and the thorium oxide had suffered partial de-emanation. At this time, also, it was beginning to be realized that the emanating power was a quantity which varied, not only with the nature of the chemical compound, but also for the same compound very materially with its previous his tory. Thus the oxide from the oxalate does not possess as a rule so great an emanating power as that used for comparison, which would account for the above result. The following two exactly parallel experiments were therefore made, the one with ordinary, and the other with de-emanated thoria possessing 9 to 10 per cent of the emanating power of the first. Each was converted to chloride in the ordinary way, by mixing with sugar solution, carbonizing, and igniting the mixt ure of oxide and carbon so obtained in a current of dry chlorine. Each sampIe was then treated with water, the thorium precipitated as hydroxide with ammonia, and the hydroxides washed and dried at 110°. The hydroxide prepared from the de-emanated thoria possessed 128 per cent, that from the ordinary thoria 108 per cent of the emanating power of ordinary thoria, when tested immediately after drying. Now a sampIe of hydroxide previously obtained had shown no less than three times the emanating power of ordinary thoria. The specimens were therefore tested again after having been kept for four days in loosely corked tubes. They now showed 157 per cent and 139 per cent respectively. The emanating power was thus increasing, so both specimens were exposed side by side in open watch

390

The Collected Papers 0/ Lord Rutherford

glasses under a sheet of glass to keep off the dust. The result is agam conclusive: From de-emanated ThOz From ordinary ThOz

After nine days After three more days

253 per cent 259 per cent

253 per cent 259 per cent

Thus the process of de-emanating thoria by ignition does not irretrievably destroy the emanating power, for after solution and reprecipitation no difference whatever exists in the emanating power between ordinary and de-emanated thoria. The results also bring out another point, the marked effect of time and exposure to air in increasing the emanating power of thorium hydroxide. This will be examined more fully later. A fair conclusion from these experiments is that the cause of the emanating power is not removed by ignition, but only rendered, for the time being, inoperative. Radioactivity 01 De-emanated Thoria. The 'straight line' radiation of thoria, de-emanated as completely as possible by ignition, was compared with that of ordinary thorium oxide by the method described. It was found that within the limits of error no difference whatever could be detected between them. This result serves to bring out the fact that the power of thoria to give an emanation is independent of its power to give a direct radiation. Is the Emanating Power a Specijic Property 0/ Thorium?

Having shown that the de-emanation of thoria by the processes described consists rather in a temporary obliteration of the effect than in a removal of the cause producing it, the next question to be considered is whether it is possible to remove from thorium compounds by chemical methods any constituent to which the property of emanating power can be traced. The thoria used in the investigation is that supplied by Messrs Eimer and Amend of New York, and is obtained from monazite sand by a secret process. It, of course, does not consist of pure thoria, although from superficial investigation it appears to be of excellent quality. There is a small quantity of a substance present which can be precipitated by sodium phosphate after removal of the thorium as hydroxide by ammonia, the nature of which is at present under investigation. The most noticeable impurity is about one per cent of thorium sulphate. Careful washing completely removed this impurity, and the emanating power of the washed sampie was identical with the ordinary. The impurity may therefore be disregarded for present purposes. Emanating power is not confined to thorium from any particular source. Orangeite and thorite from Norway both possess it, as weIl as monazite sand from Brazil. A specimen of thoria prepared from orangeite by the

391

Thr Radioactil'ity of Thorium ('o1l1pmmd'i. I

ordinary processes possessed about the same emanating power as that obtained from monazite sand by the secret process. A quantity of thorium oxide was converted into the anhydrous sulphate and dissolved in iced water. The temperature was allowed to rise and the hydrated sulphate precipitated in four fractions, a fifth being obtained by evaporation of the mother liquor to dryness. These showed no marked difference in emanating power among themselves. The first fraction was dehydrated and again submitted to fractional precipitation as hydrated sulphate. The first fraction of the new series-designated fraction AAwas then compared in the following manner with the mother liquor fraction of the first series-designated as fraction E. Both were dehydrated, dissolved in water, and precipitated as hydroxide by ammonia, washed and dried under the same conditions, and compared together at regular intervals with a comparison sampIe of ordinary thoria. Fraction AA

At first After 1 day After 13 days After 43 days

203 per 240 per 316 per 352 per

cent cent cent cent

Fraction E

200 per 249 per 321 per 372 per

cent cent cent cent

The differences are too small to afford any indication of separation of the emanating material. The straight line radiations of the two fractions tested in the apparatus (Fig. 2) also proved to be identical. It was obviously useless trying any further fractionations by this method. Since there was no appreciable difference in either property in the fractions tried, there was nothing to be gained in a further repetition. These results completely accord with those of Sir William Crookes, toe. eit., with which, however, we were not acquainted until after our own experiments had been performed. Another method for the purification of thoria, employed by Dennis (J. Amer. Chem. Soc., 1896, 18, 947), the precipitation of the hydroxide by potassium azoimide, was next tried. The latter reagent was prepared by Thiele's method (Annalen, 1892, 270, 1) from diazoguanidine nitrate. Hydrazoie acid partially neutralized with potash precipitates thorium hydroxide from the boiling solution of a thorium salt. This hydroxide, compared with a sampie which had been precipitated with ammonia in the ordinary way, showed similar emanating power. These results, whieh fail to give any indication of aseparation of the emanating material by chemical means, taken in conjunetion with those already described in the preceding section on the regeneration of the emanating power in de-emanated thoria, eertainly seemed to point to the conclusion that the power of giving an emanation is really a specific property of thorium. Recent results, wh ich will be given in the last section (p. 396), put the question in a fresh light.

392

The Collected Papers 01 Lord Ruthelford Effect 01 Conditions upon Emanating Power

Before any further work was undertaken, it was necessary to make a elose study of the influence of conditions upon the emanating power of thorium compounds. Effect 01 Temperature. The effect of increase of temperature on the emanating power of thoria has al ready been fully investigated by one of us (Phys. Zeit., 1901, 2, 429). The results stated briefly show that an increase in temperature up to a certain limit, in the neighbourhood of a red heat, correspondingly increases the emanating power. At the maximum, this is between three and four times the value at the ordinary temperature, and is maintained at this increased value for several hours without any sign of diminution with time. When the thoria is allowed to cool, the emanating power then returns to the neighbourhood of the normal value. If, however, the limit of temperature given is exceeded, de-emanation sets in, and even while the high temperature is maintained, the emanating power falls rapidly to a fraction of its former value. On cooling, the substance is found to be more or less de-emanated. It is of interest that no increase of emanating power is observed when de-emanation commences. These experiments were extended to inelude the effects of cooling. The platinum tube which contained the thoria was surrounded with a feIt jacket containing a mixture of solid carbon dioxide and ether. The emanating power immediately fell to 10 per cent of its former value. On removing the cooling agent, it again rose quickly to nearly its normal value. In another experiment, some thoria was surrounded in a platinum crucible with a mixture of solid carbon dioxide and ether, and kept in a vacuum for several hours. On removing it and allowing Hs temperature to rise, it possessed much the same value as an ordinary sampie, and after standing some time in the air it was again tested and no difference could be detected between the two. Thus changes in temperature produce very marked simultaneous changes in emanating power, but between the limits of -1100 and an incipient red heat no permanent alteration in the value occurs. Effect 0/ Moisture (Dorn, Zoc. cit.) had noticed that moisture produced a moderate increase in the power of thoria of giving an emanation, and of exciting radioactivity on surrounding surfaces. We have confirmed and extended his results by the following experiments. Two similar weights of ordinary thoria were exposed in jars sealed with wax, the one containing sulphuric acid and the other water, for aperiod of 4 days. The desiccated sampIe showed 54 per cent and the sampIe exposed to water vapour 134 per cent of the original emanating power. The experiment was repeated and the sampIes left for a week with much the same result: 70 per cent and 141 per cent respectively. It was of interest to see if a more complete desiccation would further reduce the emanating power. Five grams of thoria were sealed up in a tube containing phosphoric oxide, the two

nie

Radioacti~·ity

0/ Thorium

Compoullds. I

393

substances being separated by a plug of glass wooI. Before sealing, the tube was exhausted by a Töpler mercury pump. Mter 26 days, the end of the tube was connected with a closely packed phosphoric oxide tube, the tip broken off inside the connection, and a slow stream of dried air thus allowed to enter. The other end was connected to the testing cylinder, and arrangements were made to send a stream of air through into the cylinder. When all was ready, this end of the tube was broken inside the connection, and the emanating power measured. A similar experiment made with an ordinary sampie of thoria, using the same arrangement, showed that the desiccated sam pie possessed 79 per cent of the emanating power of the ordinary sampie tested under the same conditions. A sampie of thoria sprayed with water gave 125 per cent of its original emanating power. If completely flooded with water, however, the value is much reduced, as would be expected from the reduction of surface. Another trial was made, in which thoria was flooded with concentrated sulphuric acid. Hardly any emanation was observed so long as the mixture remained undisturbed, but when vigorously shaken it gave nearly onehalf of the original emanation. These experiments show that the presence of water, although producing a marked increase, is not apparently essential for the production of the phenomena. It must be mentioned, however, that thoria only ceases to lose weight after prolonged ignition with the blowpipe, that is, under conditions which nearly destroy its emanating power. This, with analogous points, will be taken up, however, in aseparate communication on the more purely chemical side of the question. Thc result of some experiments on the effects of other conditions may be shortly tabulated. In each case the sampie was exposed to the conditions given for four days. The emanating power is that possessed at the end of this period, compared with that of the first sampie, which is regarded as 100 per cent: Kept in sealcd test tube enclosed completely in lead tube (2) Taken from tightly stoppered stock bottle containing the main quantity (3) Sealed up in test tube and exposed to bright, all-day sun (4) Exposed to the air of the laboratory in open watch glass (5) Kept in a continuous stream of ordinary air (I)

100 per cent 100 per cent 100 per cent 105 per cent 88 per cent

The last experiment was made at a different time from the other fOUf, and therefore is not strictly comparable. The most useful result attained is that thoria does not change in emanating power when kept in closed vessels N*

394

Tize Collected Papers of Lord Rutherford

under different conditions, but when exposed to the air the cmanating power varies within comparatively narrow limits. Thorium Hydroxide. The effect of time on the emanating power of the freshly prepared hydroxide already mentioned is one of the most striking observations in this connection. The following additional experiments have been made on this point. A quantity of hydroxide was prepared, and separate portions subjected to different drying temperatures and subsequent conditions, as folIows: (1) Dried at 11 0° and exposed some hours to the air

(2) Dried as before at 110° and kept in desiccator until tested (3) Dried at 200° and kept in desiccator (4) Dried at 250° and kept in desiccator

Emanating power

264 per cent 226 per cent 220 per cent 219 per cent

From this, it appears that the additionalloss of water caused by exposure to increasing temperature is without effect on the emanating power. A similar experiment to that described for thorium oxide was performed with the hydroxide. Two quantities were exposed in c10sed bottles to the action of moist air and of air dried with sulphuric acid respectively, and showed, after four days, emanating powers of 394 per cent and 307 per cent. After having been exposed to the air for 24 hours, these sampies showed 350 per cent and 324 per cent respectively. The next experiment was designed to inc1ude the effect of carbon dioxide, which the hydroxide absorbs from the air to the extent of 2 per cent of its weight. A quantity of hydroxide was tested immediate1y after preparation, and possessed 140 per cent emanating power. A sampIe was sealed up in a test tube, while another similar sampie was tested in the following manner. It was exposed to a current of moist carbon dioxide for an hour, and then possessed an emanating power of 156 per cent. It was then 1eft exposed to the air of the 1aboratory and tested at intervals : Emanating power

After After After After After

2 days 6 days 10 days 11 days 16 days

263 per cent 325 per cent 300 per cent 341 per cent 362 per cent

On the last day, the sealed up specimen was opened and examined, and was found to possess an emanating power of 298 per cent. These experiments show that if the air is fundamental in producing the increase of emanating power with time, a very limited quantity of it is effective. For the present, it is perhaps better to consider it as an effect of time simply, hastened no doubt by the presence of water vapour.

The Radioarfivit)'

0/ Thorium ('ompollnds. I

Oll fhe ehemiral Nature

395

0/ the Emanation

The following work has reference to the emanation itself and not to the material producing it, and was designed to see whether the emanation possesses chemical properties which would identify it with any known kind of matter. It had been noticed at the time of its discovery that it passed unchanged through concentrated sulphuric acid. The same holds true of every reagent that has been investigated. The effect of temperature was first tried. The air containing the emanation, obtained in the usual way by passage over thoria, was led through tbe platinum tube heated electrically to the highest attainable temperature, and also through the tube cooled by solid carbon dioxide and ether. The tube was then filled with platinum black, and the emanation passed through in the cold, and with gradually increasing temperatures, until the limit was reached. The effect of the intense heat was to convert the platinum black completely into platinum sponge. In another experiment, the emanation was passed through a layer of red hot lead chromate in a glass tube. The current of air was replaced by a current of hydrogen and the emanation sent through red hot magnesium powder, and red hot palladium black, and, by using a current of carbon dioxide, through red hot zinc dust. In every case, the emanation passed without sensible change in the amount. If anything, a slight increase occurred, owing to the time taken for the gas current to pass through the tubes when hot being slightly less than when cold, the decay en route being consequently less. It will be noticed that the only known gases capable of passing in unchanged amount through all tbe reagents employed are the recently discovered gases of the argon family. But another interpretation may be put upon the results. If the emanation were tbe manifestation of excited radioactivity on the surrounding atmospbere, then since from the nature of the experiments it was necessary to employ in each case, as the atmospbere, a gas not acted on by the reagent employed, the result obtained might be explained. Red hot magnesium would not retain an emanation consisting of radioactive hydrogen, or red bot zinc dust an emanation consisting of radioactive carbon dioxide. The correctness of this explanation was tested in the following way. Carbon dioxide was passed over tboria, then tbrough aT-tube, where a current of air met and mixed with it, both passing on to the testing cylinder. But between this and the T-tube, a large soda-lime tube was introduced, and the current of gas thus freed from its admixed carbon dioxide before being tested in the cylinder for emanation. The amount of emanation found was quite unchanged, whether carbon dioxide was sent over thoria in the manner described, or whether an equally rapid current of air was substituted for it, keeping the other arrangements as before. The theory that the emanation may consist of the surrounding medium rendered radioactive is thus excluded, and the interpretation of the experiments must be that the

396

The Collected Papers

0/ Lord Ruther/ord

emanation is a chemically inert gas analogous in nature to the members of the argon family. It is perhaps early to discuss these results from a theoretical point of view, although it appears certain that an explanation of the nature of the emanation must precede, as a necessary step, any hypo thesis put forward to account for emanating power. The explanation already advanced and disproved being left out of the question, two other views of the origin and nature of the emanation are still possible. It may be that one of the inert constituents of the atmosphere is rendered radioactive in the presence of thoria and so constitutes the emanation. The actual amount being probably small, and air being a constant impurity in all gases as ordinarily prepared, it is, of course, no argument against this view that emanating power is independent of the gaseous medium surrounding the emanating material. An experiment is in progress, however, to ascertain whether emanating power persists in a current of gas as free from air as present methods of preparation allow. The other alternative is to look upon the emanation as consisting of agas emitted by the thorium compound. It is not necessary that such should contain thorium, it might conceivably be an inert gas continuously emitted in the radioactive state. In the present state of knowledge, it would be premature to attempt to choose between these two alternatives. But in any decision of this point, the work already given on the regeneration of the emanating power of thoria de-emanated by ignition, the continuous loss of emanating power by successive ignition at increasing temperatures, and the increase in the chemical activity of thorium hydroxide with time, must be taken into consideration. Concentration

0/ the Radioactive Material

Since the preceding account was written, developments have been made in the subject which completely alter the aspect of the whole question of emanating power and radioactivity. The first has reference to thorium nitrate, which in the solid state hardly possesses any emanating power. In a careful determination, using 20 grams of the finely powdered commercial salt, this worked out to be only 1·8 per cent of the emanating power of thoria. Dissolved in water, however, and tested for emanation by bubbling a current of air through it, it gives about three times as much emanation as thorium oxide. That is, solution in water increases the emanating power of thorium nitrate nearly 200 times. The emanating power, as in the case of solids, is proportional to the weight of substance present, and within the limits tried is not much affected by dilution, for a solution of 10 grams made up to 25 c.c. in volume possessed a similar value when diluted four times. Solutions of thorium chloride also give a large amount of emanation. In these experiments, the cylinder C (Fig. 1) is replaced by a Drechsel bottle. A drying tube of calcium chloride is inserted between it and the

The Radioacth·ity 0/ Thorium Compounds. I

397

testing cylinder to prevent the moisture destroying the insulation of the latter. In this connection, the method of testing the insulation by varying the voltage is invaluable. The air current under these circumstances cannot of course be kept so constant as when working with solid substances, and the results are not strictly comparable in consequence, but the arrangement works well enough for a first approximation. Simultaneously with this observation of the latent emanating power of thorium nitrate, it was noticed that preparations of thorium carbonate varied enormously in emanating power according to their method of preparation. A sampie prepared from the nitrate by complete precipitation with sodium carbonate showed an emanating power of 370 per cent of that of the ordinary oxide, and this value remained fairly constant with time. In another experiment, the precipitated carbonate was partially redissolved in nitric acid, and the redissolved fraction completely reprecipitated with ammonia as hydroxide. The result was remarkable: the carbonate had an emanating power of only 6 per cent, the hydroxide one of 1225 per cent of that of the ordinary oxide. On repeating the experiments, both fractions proved almost equally inactive, the carbonate showing 14 per cent and the hydroxide 19 per cent of the emanating power of thoria. An even greater difference between these two similar experiments was observed in the effects of time on the different preparations. In the first, the carbonate did not alter in value in seven days, whilst the hydroxide steadily decreased: Hydroxide

Original After 1 day After 4 days After 7 days After 14 days

1225 per cent 1094 per cent 696 per cent 614 per cent 473 per cent

Carbonate

6·2 8·4 4·8 4·7

per cent per cent per cent per cent

In the second experiment, the emanating power of both the carbonate and hydroxide had increased many fold when tested eleven days later, and the former now possessed 109 per cent, the latter 273 per cent (originally ] 4 per cent and 19 per cent respectively). The straight line radioactivity of the carbonate from the first experiment which possessed such a low emanating power is of interest. It proved to be similar to that of a specimen of hydroxide of normal emanating power, which it resembled in density and state of division. After having been bept seven days without showing any sign ofrecovering its emanating power, it was redissolved in nitric acid, and reprecipitated with ammonia as hydroxide. The latter now possessed, when first made, an emanating power of 65 per cent, and after 24 hours, 145 per cent, from which value it did not much alter. These results throw a new light on the question of emanating power. In the first experiment. which we have so far not succeeded in repeating, by an accident in the conditions apparently, two fractions were separated

398

The Collected Papers 01 Lord Rutherlord

from thorium which varied in their emanating power in the ratio of 200 to 1. The active fraction diminished to nearly a third of its original value in fourteen days spontaneously, whilst the activity of the inactive fraction was, to a large extent, regenerated by solution and reprecipitation, in an exactly analogous manner to the behaviour of thoria de-emanated by ignition. Attempts to repeat this result have so far led to the production of two more or less completely de-emanated fractions, which, however, spontaneously increase in activity with time, as in the second experiment, and this seems to be generally the case, whether incomplete precipitation is effected as in the experiment given by re-solution of the carbonate in acid, or by using a deficiency of sodium carbonate in the first instance. The production of preparations of such low emanating power led naturally to an examination being made of the filtrates and washings for radioactivity. It was found that these possess, when concentrated, both emanating power and radioactivity in considerable amounts, although from the nature of their production they should be chemically free from thorium. The behaviour is quite general, a dilute solution of thorium nitrate, after the thorium has been precipitated as hydroxide with ammonia, shows, when concentrated, an emanating power of from one-third to two-thirds that of the original nitrate in solution. It does not matter whether the thorium is precipitated with ammonia directly, or after preliminary partial precipitation as carbonate-either by adding insufficient sodium carbonate in the first place, or by precipitating completely and dissolving part of the precipitate in nitric acid-the thorium-free filtrate invariably possessed emanating power, and when evaporated to dryness exhibited straight line radioactivity also in amounts very much greater than possessed by the same weight of thoria. The result of a careful chemical investigation of the active filtrates produced under the various conditions described was to show that these contained no thorium, or at most only aminute trace, but another substance in very appreciable quantities, which can be precipitated with sodium phosphate, and which, so prepared, is a white substance possessing both emanating power and radioactivity, often many hundred-fold greater than thoria. It has not yet been obtained in suffieiently large quantities for an exhaustive chemical investigation, and it is impossible at present to say what it may prove to be. We may at onee state, however, that we do not ineline to the view that it is ThX, either in the sense of the radioactive or emanating constituent of thorium. The evidence of a long series of experiments in two directions, of whieh the final steps ean only find place here, is quite definite on this point, and in our opinion admits of only one conc1usion. There seems little doubt of the aetual existence of a constituent ThX to whieh the properties of radioactivity and emanating power of thorium must be ascribed, but in all prob ability it is present in altogether minute amount, and must therefore be possessed of these qualities to a correspondingly intense degree. But before the reasons for this view are put forward, it is neeessary to

The Radioaetivit)' of Thorium ('ompoundv. I

399

discuss more nearly the meaning of the experiments already given on the emanating power. Jt has been shown that this is a most uncertain quantity, similar experiments often giving preparations of very varying value. as is clearly shown in the results given, as weil as in many others in the same direction. The most pregnant fact is that although, as has been shown, precipitation with ammonia invariably leaves behind considerable emanating material in the filtrate which is lost, this seems to exert Httle influence on the emanating power of the precipitates. These, prepared under different conditions, often by a different number of precipitations, in which therefore varying amounts of the emanating material are lost, show a surprising uniformity in this property, especially after they have attained their maximum power by keeping. It is only necessary to quote the experiment on the almost completely de-emanated carbonate, which gained in emanating power thirty times by conversion into the hydroxide, although during the process much emanating material must have been lost, to show that the value of the emanating power alone furnishes no criterion of the amount of emanating material present. It may safely be said that three things must be carefully distinguished between in considering the nature of the property possessed by thoria of giving out a radioactive emanation. First, the nature of the emanation itself; secondly, the nature of the emanating power; and thirdly, the nature of the emanating material. The first, the emanation itself, we have shown to possess the negative properties of a chemically inert gas, whose radioactivity is unaffected by any conditions, apparently, except lapse of time. With regard to the second, the emanating power or rate at which the emanation is produced per unit weight of substance, it is certain that this does not depend only, or even mainly, on the quantity of emanating material present. The regeneration of the emanating power of thoria de-emanated by ignition, the enormous variation with time in the emanating power of the hydroxide and carbonate under certain conditions, and the comparatively constant maximum which these substances ultimately attain, although prepared under conditions where different amounts of the emanating material are lost, make this point perfectly dear. These considerations, taken in eonjunetion with the effect of temperature, moisture, ete., on emanating power, and the nature of the emanation itself, make the property appear rather as the result of a dynamical change, possibly in the nature of a chemieal reaetion where the active mass of emanating material is a constant, than as the property of a peculiar kind of matter in the static state, additive with regard to mass. It is, however, neither the emanation itself nor the emanating power with which we are concerned in these experiments, but the third conception, the emanating material, that is, the substance, whether thorium or not wh ich is responsible for the activity. It has been shown that it is difficult to follow, by means of the value of the emanating power, the progress of the removal of the active material. When this was realized, attention was directed to the straight line radioaetivity, whieh is generally unaffected by

400

The Collecled Papers 0/ Lord RUlher/ort!

these changes of conditions and previous history which produce such profound alteration in the former property. The two phenomena are undoubtedly connected. The intensely radioactive preparations obtained from thorium in different ways always show correspondingly great emanating power when the conditions are favourable for the manifestations of the latter. Solution appears to be the most generally favourable condition. The experiments we have been engaged in were therefore repeated in a form which would allow a c10se study of the total radioactivity, in the hope that this value would prove a more suitable indication of the amount of active material present than the emanating power alone. Seventy grams of thorium nitrate were dissolved in four litres of boiling water, and precipitated with ammonia added cautiously in very dilute solution in excess. The filtrates and washings were evaporated to about 60 c.c. and then possessed as much emanating power as 146 grams of thoria. On evaporating the solution to dryness and removing the ammonium salts by ignition the residue weighed 0·0583 gram. The emanating power of this residue in solution was thus about 2500 times that of ordinary thoria. In the solid state, however, the value fell to one-fiftieth. But its total radiation was equivalent to at least 23·6 grams of thoria, that is, was about 400 times as great. It was dissolved in hydrochloric acid, and ammonia added in excess, when a precipitate weighing 0·0015 gram was thrown down. This contained all the thorium present besides iron in appreciable quantity which had been introduced during the evaporation. It equalled in radioactivity 2·73 grams of thoria, the ratio in this case being thus no less than 1800 times. Sodium phosphate precipitated 0·0225 gram of white substance the activity of which was equivalent to 4·4 grams of thoria, that is, 200 times. The sodium salts freed from ammonium still possessed a radioactivity equivalent to 3·6 grams of thorium oxide. In other experiments, however, these had been obtained quite free from activity, and this result is due to the solubility of the phosphate in water, so that some was dissolved during the washing (which the subsequent determination of the weight rendered necessary) and appeared in the filtrate. The radioactive residue obtained in the first place from the filtrate by evaporation and ignition, before it was redissolved, had, however, been tested to determine the penetrative power of the radiations emitted. If the rays from various radioactive substances are made to pass through successive layers of aluminium foil, each additional layer of foil cuts down the radiation to a fraction of its former value, and a curve can be plotted with the thickness of metal penetrated as abscissae, and the intensity of the rays after penetration as ordinates, expressing at a glance the penetrative power of the rays being examined (compare Rutherford, Phi!. Mag., 1899 [v], 47, 122). The curves so obtained are quite different for different radioactive substances. The radiations from uranium, radium, thorium, each give distinct and characteristic curves, whilst that of the last named again is quite different from that given by the excited radioactivity produced by the

Thr RadioaClil'ily of Thorium Compoullds. I

401

thorium emanation. The examination in this way of the penetrative power of the rays from the radioactive residue showed that the radiations emitted were in every respect identical with the ordinary thorium radiation. In another experiment, the nature of the emanation from a similar intensely active thorium-free residue was submitted to examination. The rate of decay was quite indistinguishable from that of the ordinary thorium emanation. That is, substances chemically free from thorium have been prepared possessing thorium radioactivity in an intense degree. The main quantity of thorium hydroxide in the last experiment was redissolved in nitric acid, and the previous round of operations repeated twice, the filtrates from each operation being mixed and then examined exactly as in the former case. The emanating power of the concentrated solution was only equal to that of 8 grams of thoria in this instance, and the radioactivity of the residue to that of 3 grams. From this only a small quantity of the phosphate precipitate was obtained (0' 00 1 gram) the radioactivity of wh ich was equal to that of 0·3 gram of thoria (ratio 200 : 1). The emanating power of the main quantity of the hydroxide when first so prepared was 73 per cent that of thoria, that is, about one-half of its usual value. The hydroxide was converted into oxide by ignition, and its radioactivity compared with that of the oxide from the original nitrate prepared in the same way. It was found to be only about one-third as acti ve, the exact ratio being O· 36 : 1. Only one conc1usion seems possible from this series of experiments. There is no longer any room for doubt that apart of the radioactive constituent ThX has been separated from thorium, and obtained in a very concentrated form, in one instance 1800 times more powerful in its actions. This result, taken into account with the reduction of the radioactivity and emanating power of the main quantity of thorium compound, and the identity of the radiations of the active thorium-free preparations with those of the ordinary thorium radiation, warrant the conc1usion that ThX is a distinct substance, differing from thorium in its chemical properties and so capable of separation therefrom. The manner in wh ich it makes its appearance, associated with each precipitate formed in its concentrated solution, resembles the behaviour of Crookes' UrX, which he found was dragged down by precipitates when no question of insolubility is involved, and suggests the view that it is really present in minute quantity. Even in the case of the most active preparations, these probably are composed of some ThX associated with accidental admixtures probably large in proportion. These results receive confirrnation from observations made in a different method of separating ThX. The experiment was tried of washing thoria with water repeatedly, and seeing if the radioactivity was thereby affected. In this way, it was found that the filtered washings, on concentration, deposited small amounts of material, with an activity often of the order of a thousand times greater than that of the original sampie. In one experi-

402

The Collected Papers 0/ Lord Rutherford

ment, 290 grams of thoria were shaken for a long time with nine quantities, each of 2 litres, of distilled water. The first washing, containing most of the sulphate already referred to, was rejected, the rest concentrated to different stages and filtered at each stage. One of the residues so obtained weighed 6·4 mg. and was equivalent in radioactivity to 11· 3 grams of the original thoria, and was therefore no less than 1800 times more radioactive. It was examined chemically, and gave, after conversion into sulphate, the characteristic reaction of thorium sulphate, being precipitated from its solution in cold water by warming. No other substance than thorium could be detected by chemical analysis, although of course, the quantity was too small for aminute examination. But the absence of the substance precipitable as phosphate, noticed in the other experiments, confirms the opinion that this is an accidental admixture without influence on the qualities of radioactivity and emanating power. The penetrative power of the radiation from this substance again established its identity with the ordinary thorium radiation. In another experiment, a small quantity of thoria was shaken many times with large quantities of water. In this case, the radioactivity of the residue was examined and found to be about 20 per cent less radioactive than the original sampie. There remains only one step to prove beyond doubt that the radioactivity and emanating power of thorium are not specific properties of the thorium molecule-the preparation of thoria free from these properties-and on this problem we are now engaged. To sum up briefly what has already been accomplished, two different methods have effected a concentration of the activity many hundred-fold in one fraction, and a corresponding diminution of activity in the remainder, but in each case the character of the radiation is not thereby affected. In one method, the active fraction appears to consist only of thorium, so far as examination has been possible, while in the other case radioactivity and emanating power appear to be manifested indiscriminately in all the products, without reference to their chemical nature. The simplest explanation of this behaviour, on the present view, is that so far the active constituent of thorium has only been obtained in relatively minute quantity, and therefore does not answer to any definite analytical reactions. Macdonald Physics Building Macdonald Chemistry and Mining Building McGiU University, Montreal

The Existence of Bodies Smaller than Atoms by

Macdonald Professor of Physics, McGilI University, Montreal

E. R UTHERFORD, M.A., D.se.,

From Transactions olthe Royal Society olCanada, 1902, sero ii, section iii, voI. 8, pp. 79-86* (Read May 27, 1902)

DURING the last few years considerable evidence has been obtained of the production, under various conditions, of bodies wbich behave as if their mass was only a small fraction of the mass of the chemical atom of hydrogen. As far as we know at present, these minute particles are always associated with a negative electric charge. For this reason they have been termed 'electrons'. In whatever way they are produced, they always have the same charge and this charge is probably the same as that carried by the hydrogen ion in the electrolysis of water. Abrief historical account will be given of the growth of our knowledge of tbis subject, which seems likely in the near future to profoundly modify our ideas of the constitution of matter. Faraday showed that when a current passed through a conducting solution, the amount of matter deposited or given off at the electrodes depended only on the quantity of electricity which had passed through the solution. For different solutions, the amounts of matter deposited for unit quantity of electricity are chemical1y equivalent to each other. It is now gene rally accepted that the current is carried through the solution by me ans of charged carriers or ions. In an electric field the negative ions travel through the solution to the positive electrode, and the positive ions to the negative. The weight W of hydrogen given off for a passage of Q coulombs of electricity is given by W =-= Z Q. where Z = 10-- 4 is the weight of hydrogen givcn off for a passage of one electromagnetic unit of electricity. Let e = charge on an ion. m = mass of each ion. n = number of ions of hydrogen in a weight W. Then W = mn. Q =ne.

* The abstract of an address before Section III. of the Society, introducing a discussion on the evidence of existence of bodies smaller than atoms. Experiments illustrating points of the theory were kindly shown to the meeting by Dr. J. MacLennan, of Toronto University.

404

The Collected Papers

0/ Lord Ruther/ord

e Q I We therefore have - = - = - = IQ4 m W Z This gives the ratio of the charge of an ion to its mass in the electromagnetic system of units. So far no assumptions have been made as to the actual value of the charge of the mass of an ion. Rough approximations to the values of these quantities can be obtained from considerations based on the kinetic theory of gases, but, as will be seen later, the evidence does not rest on the actual values of the mass but only on the value of the ratio!!..· m Sir William Crookes first drew attention to a remarkable phenomenon which showed itself when an electric discharge was passed through a highly exhausted vacuum tube. Below a certain pressure of the gas in the tube, a peculiar kind of rays are shot off from the cathode. These 'cathode' rays travel in straight lines and produce brilliant phosphorescent effects on the walls of the tube and also on many other substances placed direct1y in their path. Crookes showed the path of the rays could be bent by a magnet. In a strong magnetic field these rays can be made to trace out spirals round the direction of the lines of magnetic force. He showed that they produced strong heating effects by their impact and a considerable mechanical pressure on vanes placed in their path. For a long time two rival theories held the ground as to the explanation of these effects. The German school of physicists took the view that the cathode rays were ether waves of some kind. The English view, as voiced by Crookes, held that they were in reality projected partic1es travelling with high velocity. On the latter view most of the effects observed by Crookes received a simple explanation. The phosphorescent, heating, and mechanical effects were due to the bombardment of material partic1es, driven off from the cathode by a strong electric field. The curvature of the path of the rays by a magnetic field was due to the fact that a moving charge acts like a current. The presence of two riyal theories led to a large amount of investigation of the discharge in vacuum tubes. Hertz tried if the rays were deviated by a strong electric field but failed to get any effect. Lenard, in 1895, showed that the cathode rays were able to pass through thin windows of glass, mica, or metal foil. He was thus able to examine the cathode rays outside the vacuum tube. He showed that the absorption of the rays by matter was independent of its chemical constitution and depended only on its density. This was true whether the matter was in the state of solid, liquid or gas. The fact that these rays could pass through solid matter, together with the absorption results, pointed to the conc1usion that, if the rays were projected material partic1es, they must be so small that they were able to pass through the interstices of matter, or in other words, that to these partic1es matter behaved like a coarse sieve.

The Ex;stence of Bo(lies SmalleI' than A 10ms

405

In 1895 Perrin showed that the rays carried with them a negative charge. This was strong evidence in support of Crookes' hypothesis. The discovery of Röntgen rays gavc a great stimulus to the investigation of the discharge in vacuum tubes. It was found that Röntgen rays werc able to produce charged carriers or ions from the gas, through which they passed, and this made it probable that carriers of a similar kind existed in a vacuum tube. The experiments of J. J. Thomson threw a great deal of additional light on the nature of the cathode rays and laid the foundation of all future work on that subject. He found that the rays were negatively charged particles travelling with enormous velocities and were probably of dimensions small compared with a molecule. These results were deduced from experiment in the following way: Let e m

= =

t' =

n=

charge on the cathode ray particle, mass of the particle, velocity of the particle, number of particles in a single discharge.

The mechanical energy of a single particle is 1- mv 2 and this is mainly transformed into heat in the impact on a metal surface. The energy W in a single discharge was measured by observing the rise of temperature when the rays fell on a specially constructed thermopile. We therefore have W = 1- m n v 2• The amount of electricity Q carried by a single discharge was measured by an electrometer and was given by Q = n e. Dividing the second equation by the first, we obtain e

Qv 2

m=2W Another relation was obtained between e, m, and v by observing the curvature p of the path of the rays when a uniform magnetic field H was applied perpendicular to the direction of the rays. Acharge e, moving with a velocity v, acts as a current of strength e v. The force exerted on the partic1e by the magnetic field is H e v and is mutually perpendicular to the direction of the field and the rays. The rays are made to trace out a curved path under the action of this force. The force which causes the body to move in a path of curvature p is dynamically given by m v 2 and this is equal to the p applied force H e v. Therefore He

t' =

m v2

P

or

e m =

We have already shown that

e

l'

Hp Q v2

in = 2 W

406

The Collected Papers of Lord Rutherford

. b . Prom t hese two equatlons we 0 tam v e

=

2W pH Q 2W

and - = - - - - m p2 H2 Q Substituting the numerieal values of H, e, and Q. it was found that in round numbers e -m = 107 and v = 1010 ems. per sec. We thus have obtained the result that the value of ~ for the particles m is about 1,000 times as great as the same ratio observed for hydrogen on the electrolysis of water. If the charges are the same for both cases, the mass of the carriers in the cathode rays is only about 1/1000 of the mass of the hydrogen atom. The velocity of these particles is very great, approximating to the velocity of light, and enormously greater than any velocity of matter before observed in physics. The theory from which these results are deduced is possibly open to some objections, but the values were confirmed by another independent method. If the rays are charged particles their path should be altered in passing through an electrostatic field. Hertz obtained negative results; but J. J. Thomson, by varying the experimental eonditions, was able to show that the rays are deviated and that the failure of Hertz to observe the effect was due to the masking action of the conducting gas, through whieh the particles moved. This electrostatic deviation supplied hirn with a simple means of determining the velocity and ratio of': of the partieles. The rays m were made to pass between the plates of a eharged condenser and were at the same time acted on by a magnetic field. The strength and direction of the field was so adjusted that there was no deviation of the path of the rays. From the data of the experiment the values of the veloeity and :. were m found to be about the same as those determined by the first method.

J. J. Thomson also found that the ratio:' was independent of the gas m in the vacuum tube, showing that, possibly, particles of the same size were produced from different kinds of matter. It is, however, possible to explain this result by supposing that the discharge is in all cases carried by the traee of water vapour which is always present in the vacuum tube. A eomplete confirmation had thus been given to the projection theory of cathode rays, and the importance of the work was at once recognized by Continental physicists.

FliC! Existente oI Boclies Smallef II/(m Atoll/s

407

Aseries of experiments were performed by Des Coudres, Lenard, Kaufmann and others, which verified and extended Thomson's results. There was always present, however, a doubt that possibly the theory from which the results were deduced might be inapplicable, and that the enormous velocity of the partic1es did not exist in fact. This last doubt was completely removed by Wiechert, who showed, by actually measuring the time taken by the rays to pass from one point to another, that the velocity was of the same order as that obtained by previous observers by the methods already explained. Townsend showed by considerations based on the Kinetic theory that the charge on a gaseous ion was the same as on an ion in electrolysis. By a beautiful method, J. J. Thomson succeeded in determining the actual value of the charge, and this charge was found to be the same from whatever gas the ion is produced.

J. J. Thomson also made determinations of the ratio :. for electrons m produced in two other distinct ways. Since the work of Hertz on electrical waves, it has been known that a clean surface of metal discharges negative electricity when ultra violet light falls upon it. The ultra violet light, in some way, produces negatively charged ions at the surface of the metal plate. At atmospheric pressure these ions are equal in size to the ions produced out of the gas by Röntgen or Becquerel rays, but at low pressures they have been shown to be similar to cathode rays. By observing the deviation of these ions by a magnetic field when they were made to move rapidly in a strong electric field, J. J. Thomson showed that the ratio of.!!. m for the carriers was the same as for the cathode ray produced in a vacuum tube. In a similar way he showed that the same ratio was obtained for negative particles shot off from the glowing carbon filament of an incandescent lamp. We thus see that 'electrons' produced under widely different conditions are all of the same size and of mass about 1/1000 of the hydrogen atom. We have so far considered electrons which have been produced by the agency of light, heat, and the electric discharge, but there is also very strong evidence that these electrons are always present in matter and may manifest their presence under special conditions. Zeeman discovered in 1896 that the bright lines in the spectra of many substances were displaced by a strong magnetic field acting on the source of light. Later experiments showed that under certain conditions one of the D-lines of sodium, for example, was transformed into a triplet by the action of a magnetic field. These lines showed definite peculiarities in regard to the polarization of the light. These results were in direct confirrnation of a theory advanced by Lorentz which considered light to be due to the rotation or oscillations of the electrons in the molecule. The equations representing

408

The Collected Papers of Lord Rutherford

the change of period of the light in a magnetic field involved the ratio !!.-. m From the change in wave length of the light vibration it was possible to

!!. was again found to be m about 107, showing that in all probability that light was due to the rotation or oscillation of electrons in a molecule, and that the mass of the electron was much smaller than the atom itself. Not only are these electrons present in matter, but in some cases they are spontaneously emitted from it. Becquerel has shown that the radioactive substances uranium and radium give out some rays deviable in a magnetic field. The writer has recently found that thorium, the other permanent radioactive substance, also possesses the same property. These rays were found to be analogous in all respects to high velocity cathode rays. They were deviated by a magnetic and by an electric field and carried with them a negative charge. Becquerel also showed that the particles travelIed with deduce the value of this ratio. The value of

a velocity not very different from the velocity of light, while the ratio of !!. was again about 107 • m Many of the electrons shot off from radioactive substances have a much higher velocity than the cathode rays in a vacuum tube. The highest velocity observed for the latter is about one-third of the velocity of light, while Kaufmann recently found that the velocity of some of the radium electrons was about 95 per cent of the velocity of light. Experiments of these very high speed carriers are of great importance at the present time, in order to throw some light on the question as to whether the mass of the electron is apparent or real. On the present electromagnetic theory a rapidly moving charged body increases in apparent mass with increase in velocity. When the carriers travel with the velocity of light the apparent mass would be infinite. n is not yet settled what proportion of the apparent mass is electrical. n may possibly prove that the mass is altogether electrical in origin. If such should prove to be the case* (and it does not seem improbable), it would be very strong evidence in support of the view that all mass is electrical in character. It thus appears that electrons produced by the electric discharge, by a glowing carbon filament, and by ultra violet light, as weIl as those present in incandescent sodium vapour or spontaneously emitted by radioactive substances, alI alike show ab out the same ratio of !!.... Since the charges m • Within the last month, important results bearing on this point have been published by Kaufmann and Abraham. The former has shown that the apparent mass of the electron increases with the speed in the same way as the electromagnetie theory suggests. He has deduced that the apparent diameter of the electron is 10- 13 ems, and that its mass is probably altogether electromagnetie in origin.

Tllc Existence of Bodies SmalleI' Illan Atoms

409

are the same in each case, the masscs must be the same for the electrons produced in such widely different ways. The electron thus appears to be the smallest definite unit of mass with which we are acquainted. The view has been put forward that all matter is composed of electrons. On such a view an atom of hydrogen for example is a very complicated structure consisting possibly of a thousand or more electrons. The various elements differ from one another in the number and arrangement of electrons, which compose the atom. We thus have a kind of modified Prout's hypothesis in which the electron is the ultimate corpuscle of which all matter is composed. The physical existence of electrons is now accepted by many scientific men and there are a large number of prominent physicists who are developing mathematically the logical sequence of the idea. I need only mention a few of the more prominent workers-Drude, Voigt, Riecke in Germany, Lorentz and Zeeman in Holland, Poincare and Becquerel in France, J. J. Thomson, Schuster, Lodge and Lord Kelvin in England, to show that the view has a solid basis of support among the ablest physicists. The view that the atom is a complex aggregate instead of a simple entity, as was first supposed, does not in any way invalidate the basis of chemical theory. All we have to suppose is that the chemical atom is the smallest quantity of matter which takes part in a chemical combination, and that the removal of an electron is a sub-atomic change quite distinct from ordinary chemical action, although a chemical action may in some cases be accompanied by the emission of electrons. The evidence of the complexity of the atoms of the elements is very strong from other points of view than those considered in this paper. The extraordinary complicated spectrum of elements of heavy atomic weights is of itself very strong support of the view that an atom is a very complicated structure.

Penetrating Rays from Radio-active Substances From Nature, 66, 1902, pp. 318-9

permanent radio-active substances uranium, thorium and radium an give out two types of rays, one easily absorbed and non-deviable by a magnetic field and the other more penetrating in character and deviated by a magnetic field. In addition to these rays, Villard, using the photographic method, first drew attention to the existence of some very penetrating rays from radium non-deviable by a magnetic field. This result was confirmed by Becquerel. I have recendy examined all these radio-active substances by the electrical method, and have found that thorium, and also the excited radio-activity produced by thorium and radium, emit some rays as penetrating in character as those from radium. Uranium, in comparison with thorium and radium, emits little, if any, of this radiation. These rays are extraordinarily penetrating in character, and pass readily through great thicknesses of matter. They are certainly as penetrating as the most penetrating rays given out by a hard X-ray tube. The amount ofionisation produced by them is only a very small fraction of that produced by the other two types of radiation. U sing testing vessels of ordinary size, the ionisation due to the penetrating rays is of the order of 1 part in 100 of that due to the deviable rays and 1 part in 10,000 ofthat due to the easily absorbed rays. In the experiments on radium, 0·7 gram of radium chloride, of activity 1000 times that ofuranium, was used. The radiation'from this, after its passage through 1 cm. of lead, caused a rapid movement of the needle in the sensitive electrometer employed. The radium was placed in a thick-walled lead vessel and a piece of aluminium waxed tightly over the top to prevent the escape of the emanation. The following numbers illustrate the diminution of the rate of leak in a testing vessel, placed above the radium, with the thickness of the lead traversed by the radiation:-

THE

Thickness of lead

Current

0·72cm. 0·72 + 0·62 cm. " + 1·24 " " + 1·86 " " + 2·50 "

1 0·60 0·37 0·25 0·16

The current with 0·72 cm. of lead over the radium is taken as unity. It will thus be seen that the current falls off approximately in a geometrical

Pelletratillg Rays /rom Radio-ac/ire Subslallces

411

progression with the thiekness traversed, and that after passing through 1·86 em. of lead the intensity is redueed to about one-quarter. The following table shows the thiekness of different metals traversed before the intensity is redueed to one-half:Metal

Mereury Lead Tin .. Copper Zine .. Iron ..

Thickness in cm.

0·75

0·9

1·8

2·2 2·5 2·5

Assuming this law of absorption to hold, the rays would pass through a thiekness of about 7 em. of lead, 19 em. of iron and about 150 em. of water before the intensity would be redueed by absorption to one per cent of its original value. The amount of the penetrating radiation from thorium is about the same as for radium, taking into account the ratio of their radio-aetivities. As the radium employed was about 1000 times as aetive as thorium, it was neeessary to work with a kilo gram of thorium nitrate to obtain about the same amount of rays as from the 0·7 gr. of radium. Experiments were also made to see if the exeited radio-aetivity, due to thorium and radium, whieh gives out deviable and non-deviable rays, also emits these penetrating rays. In order to get measurable effeets, it was necessary to obtain intense excited activity. For this purpose a zinc plate was exposed as kathode in a closed vessel containing 300 gr. of thoria. A lead wire was also made very active by exposure as kathode for six hours in a vessel eontaining a large amount of radium emanation, obtained by bubbling air through a solution of radium chloride. The exeited radiation from these two sourees was found to include rays about as penetrating in character as those from radium and thorium. The intensity of these rays diminished with the time, rapidly for radium and more slowly for thorium excited radiation. This diminution with time is probably direct1y connected with the rate of deeay of the other known types of radiation from excited bodies. Since the penetrating rays are present in thorium and radium, and also in the excited radiations due to these bodies, and are absent in uranium, it seems probable that the penetrating rays in both radium and thorium are due to the excited radio-activity, produced in the mass of the compound by the emanations whieh are unable to escape into the air. According to this view, the production of penetrating rays is a function of that portion of radioactive matter which eauses exeited radio-activity. Connection between Absorption and Density.-Some experiments were made to see how the absorption of the rays by matter varied with the density. The eoefficient of absorption ,\ was determined by noting the ratio of the

412

The Collected Papers

0/ Lord Rutlzer/ord

intensities of the rays after passing through a known thiekness of matter. The following table illustrates the results:-

Penetrating rays

Deviable rays from uranium

Substance

Water Glass [ron Zine Copper Tin Lead Mereury

A

.:\ density

0·033 0·086 0·28 0·28 0·31 0·38 0·77 0·92

0·033 0·035 0·036 0·039 0·035 0·052 0·068 0·068

.:\ -

14·0 44 60 96 122 -

.:\ density

5·7 5·6

-

7·7 13·2 10·8 -

A eomparison table on the right is added for the deviable rays given out by uranium. It will be seen that the quotient of absorption by density is in neither ease a constant, but the differences are no greater for the non-deviable penetrating rays than for the deviable rays of uranium. It is interesting to observe that the value of ,\ divided by the density is for both types of rays twiee as great for lead as for glass or iron. It will be seen from the above table that the penetrating rays from radium, compared with the deviable rays of uranium, pass through a thiekness of glass about 160 times greater for the same reduction of intensity. Comparison 0/ penetrating Rays with Röntgen and Kathode Rays.-The question at once arises as to whether these very penetrating rays are projeeted partic1es like kathode rays or a type of Röntgen rays. The fact that the penetrating rays are not deviable by a magnetic field seems, at first sight, to show that they cannot be kathode rays. I have repeated the experiments of Villard, and have been unable to obtain any appreciable deviation of the rays, which had passed through 0·6 em. of lead, even in a very strong magnetic field. The photographic method was used, and four days' exposure of the plate was neeessary to get an appreciable impression. In some other respects, however, the rays seem more c10sely allied to kathode than to Röntgen rays. It is well known that Röntgen rays produce mueh greater ionisation in gases like sulphuretted hydrogen and hydrochloric acid gas than in air, although the differences in density are not large. For example, sulphuretted hydrogen gives six times and hydrochloric acid gas nine times the eonductivity of air. On the other hand, with kathode rays the conduetivity observed is only slightly greater than for air.

Pl'l1l'trat illg Rays /1'0111 Radio-acl h'l' Suhslanccs

41 J

The experiment was made of filling the testing vessel with sulphuretted hydrogen. when it was found that the current for the penetrating rays from radium was only slightly greater than for air. Both this experiment and the results for thc variation of absorption of the rays with the density of matter seem to show that the penetrating rays have a c10ser resemblance to kathode than to Röntgen rays. It must, however, be remembered that the observations of the relative conductivity of gases and the relative absorption of metals for Röntgen rays have only been determined for rays far less penetrating in character than these rays from thorium and radium. Benoist has shown that the relative absorption of Röntgen rays by matter depends to a large extent on the kind of rays employed. 'Hard' rays give quite different ratios from 'soft' rays. For penetrating Röntgen rays the absorption of the rays by a given weight of the elements is a continuous and increasing function of their atomic weights. From the curve of absorption, given in his paper, the variations of absorption with density are much greater for Röntgen rays than for the penetrating rays from radio-active substances. A very important question arises in discussing the character of these penetrating rays. According to the electromagnetic theory, developed by J. J. Thomson and Heaviside, the apparent mass of an electron increases with the speed, and when the velocity of the electron is equal to the velocity of light its apparent mass is infinite. An electron moving with the velocity of light would be unaffected by a magnetic field. It does not seem at all improbable that some ofthe electrons from thorium and radium are travelling with a velocity very nearly equal to that of light, for Kaufmann has recently determined the velocity of the most penetrating deviable rays from radium and found it to be about 95 per cent of the velocity of light. The power of these rapidly moving electrons of penetrating through solid matter increases very rapidly with the speed. From general theoretical considerations of the rapid increase of mass with speed, it is to be expected that the penetrating power would increase very rapidly as the speed of light was approached. Now we have already shown that these penetrating rays have very similar properties, as regards absorption and ionisation, to rapidly moving electrons. In addition, they possess the properties of great penetrative power and of non-deviation by a magnetic field which, according to theory, belong to electrons moving with a velocity very nearly equal to that of light. It is thus possible that these rays are made up of electrons projected with a speed of about 186,000 miles per second. An interesting speculation arises from the experimental observation that the excited radiations from bodies inc1ude these very penetrating rays. Elster and Geitel have recently shown that excited radio-activity can be produced from the atmosphere by exposing a negatively charged wire in the open air. This excited activity is very similar in properties to that produced by thorium and radium. Since the earth is negative with regard to the upper atmosphere,

414

The Collected Papers 0/ Lord Rutherford

the surface of the earth is itself made radioactive. From the nature of the phenomenon, it necessarily follows that, not only the surface of the earth, but also the whole interior surface of buildings is covered with an invisible deposit of radioactive matter. From the elose similarity in the nature of this excited activity from the air with that from radio-active bodies, it is not improbable that the excited radiations from the air inelude also some of the penetrating rays. If this is the case, our bodies must be continually subject in a small degree to something very like the Röntgen ray treatment, which is now so popu1ar in medical circles. It would also follow that the 'spontaneous' ionisation of air, observed in elosed vessels by Elster and Geitel and C. T. R. Wilson, may be due, in part at least, to the presence of these rays, which so readily pass through the walls of the containing vessels. E. RUTHERFORD McGill University, Montreal July 6

Comparison of the Radiations from Radioactive Substances by E. RUTHERFORD, M.A., D.se., Macdonald Professor of Physics, and MISS H. T. BROOKS, M.A., McGill University, Montreal

From the Philosophical Magazine for July 1902, sero 6, iv, pp. 1-23

ALL the radioactive substances possess in common the power of acting on a photographic plate and of ionizing the gas in their immediate neighbourhood. The intensity of the radiations may be compared by means of the photographic or electrical action; and in the case of the strongly radioactive substances by the power of lighting up a fiuorescent screen. Such comparisons, however, do not throw any light on the question whether the radiations are of the same or of different kinds. It is well known that such different types of radiation as the short waves of ultra-violet light, Röntgen and cathode rays all possess the property of producing ions throughout the volume of the gas, lighting up a ßuorescent screen and acting on a photographic plate. None of the radiations from the various radioactive substances show any trace of regular refiection, refraction, or polarization. * There are two general methods of differentiating to some extent between the various types of radiations.

(1) By observing whether the rays are appreciably deviated by a magnetic field. (2) By comparing the relative absorption of the rays by solids and gases.

The first method has been utilized by Giesel, Becquerel, t Curie, and others. Of the radioactive substances which have been most c10sely examined, viz. uranium, thorium, polonium, and radium, the latter has been shown by many observers to give out rays defiectable by a magnet. Debiernet states that the radioactive substance which he has termed actinium also gives out some rays defiectable by a magnet. In all cases these deßectable rays are similar in every respect to cathode rays, and are thus probably streams of negatively charged particles moving with very great velocities. Becquerel§ has shown that the

* A very complete and admirable account of radioactive substances by Henri Becquerel and P. and Mme Curie is given in vol. iii of the Reports of the Congres International de Physique held at Paris, 1900. t Paris Report, 1900. t Comples Rendus, cxxix (1899), and cxxx (1900). § Loe. eil.

416

The Collected Papers 0/ Lord Rutlzer/ord

ratio.!!. of the charge to the mass of these negatively charged carriers is about m 104, which is about the same value observed for the cathode rays produced in a vacuum tube. Radium, in addition to the deflectable rays, also emits non-deviable rays. The ionizing and fluorescent action of radium rays in air at atmospheric pressure, at a distance of from 5 or 6 cm. from the surface of the radium, is very largely due to the rays deflected by a magnetic field. For distances less than this, the ionization is partly due to the deflectable rays and partly to rays which are not acted on by a magnet. Close to the surface of the radium the ionization due to the non-deviable rays greatly preponderates over that due to the deviable rays. This is due to the fact that the non-deflectable rays are very largely absorbed in passing through a few centimetres of air at ordinary pressure. Action 01 a Magnetic Field on Uranium Rays

Becquerel has examined the rays of uranium in a magnetic field by the photographic method, and found that some of them are deflectable. We have confirmed these observations by the electrical method, and found that only the penetrating rays of uranium are deviable. One of us* has shown several years ago that the radiation from uranium was complex, and could be divided into two types of radiation, which were called for convenience the IX and ß radiations. The ß radiation is far more penetrating in character than the IX radiation, but is difficult to examine accurately on account of the small conductivity produced by it in the gas, compared with that due to the IX radiation. In order to measure with certainty the very small rate of leak involved, a very sensitive electrometer was employed. The instrument is described by Dolezalekt in arecent paper, and was constructed by Herr BarteIs of Göttingen. It was of the usual quadrant type, but was provided with a very light needle suspended by a fine quartz fibre. When the needle was charged to 200 volts it gave a deflection corresponding to 1,500 mm., with the telescope and scale at a distance of about 150 em., for 1 volt between the quadrants. For the special purpose for which it was employed, it was found necessary to improve the insulation of the quadrants and to alter the quadrant connections. The instrument was easy to work and gave accurate results. It has been employed recently by one of ust to measure the small spontaneous ionization produced in the air, which has been shown by the experiments of Elster and Geitel§ and C. T. R. Wilsonll who used specially designed electroscopes for that purpose. In the experiments on the action of a magnetic field on uranium radiation (Fig. 1) a thick layer of uranium oxide was placed on the bottom of a reetangular lead box 5·7 cm. long, 1·8 cm. wide, and 4·0 cm. deep, which was

*

E. Rutherford, Phi!. Mag., January 1899. t Verh. d. D. Physik. Ges. iii (1901). t Rutherford and Allen, Phys. Zeit., No. 11, 1902. § Phys. Zeit., November 24, 1900.

11

Proc. Roy. Soc., March 1901.

Comparisoll of the Radiations /rom Radioactire Substances

417

placed between thc fiat pole pieces of a large electromagnet. The rays, after passing out ofthe lead box, passed between two parallel insulated plates A and B. One ofthese plates A was charged to a P.D. of 50 volts above the earth by means of a battery. The other plate B was connected to one pair of quadrants of an electrometer in the usua! manner. Electrostatic disturbances were completely eliminated by covering the electromagnet and wires leading to it with tin foil connected to earth. There was always a small current observed between the plates on account of the spontaneous ionization of the air in the testing vesse1 when the uranium oxide was removed to a distance. The layer of uranium oxide was covered with several thin layers of aluminium of sufficient thickness to completely absorb all the 01: radiation. The open

8

oxide. A A

>A/.fuil oxide.

A

oxide.

Ur. oxide. 1 end of the lead vessel was covered with thin aluminium foil. In that case the rate of leak of the electrometer was due to ionization produced between the plates by the ß radiation together with the ions spontaneously produced by the air itself. The latter was accurately determined before the lead vessel containing the oxide was placed between the poles of the electromagnet. As the magnetic field was increased, the rate ofleak observed by the electrometer steadily diminished, until with a strong field the rate ofleak was reduced alm ost to that due to the spontaneous ionization of air. This diminution ofthe rate of leak between A and Bis due to the curvature of the path of the rays by the magnetic field before they reach the testing vessel. Since the rate of leak, due to the action of the ß radiation, with a strong magnetic field is reduced to a small fraction of its value when no magnetic field is acting, we may conclude that the ß radiation is composed almost entirely of rays deviable by a magnetic field. A comparison experiment with radium showed that the ß rays of uranium were deflected to about the same extent as the radium rays for the same strength of field. No action of a magnetic field on the 01: radiation of uranium was observed.

o

418

The Collected Papers

0/ Lord Ruther/ord

Both radium and uranium resemble one another in emitting two types of radiation, one of which is deviated in a magnetic field, and the other not. Absorption

0/ the

ß Radiation by Substances

Since the ß radiation of uranium is acted on by a magnetic field to almost the same extent as radium rays, we may conclude that the deviable rays are due to negatively charged particles emitted with high velocities; for Becquerel has shown that some of the radium rays move with a velocity of at least 1· 6 X 1010 cm. per second. The penetrating power ofthe ß rays is greater than that of the similar radiation for radium in our possession. It readily passes through 2 mm. of glass before complete absorption.

EiJl'th

B

A

Äl roi/

300Y.

tEiJl'th

EiJl'th Fig.2

Lenard, in his well-known experiments on cathode rays, has shown that the absorption of cathode rays in substances depends only on the density of the material through which they pass, and is approximately independent of its chemical constitution. On account of the constancy of the uranium rays, it is possible to determine their absorption in different media with accuracy. A few experiments were consequentIy made to see how c10sely the absorption varied with the density for the high-velocity particles emitted by uranium. The experimental arrangement is shown in Fig.2, where the dotted lines represent insulators. A thick layer of uranium was uniformly spread over a shallow rectangular groove 6 cm. square in lower plate A. The plate A was charged to 300 volts by a battery of small accumulators, the other pole of which was to earth. The current was observed between the plates A and B by means of the sensitive Dolezalek electrometer previously described, with, if necessary, a suitable capacity in parallel. In order to completely absorb the (X radiation an aluminium plate 0·003 cm. in thickness was fastened tightly over the layer of uranium. The P.D. of

Comparison of the Radiatiol1s from Radioactive Substances

419

300 volts between A and B (6 cm. apart) was sufficient to carry over all the ions to the electrodes before appreciable recombination occurred. The rate of movement of the electrometer needle was observed, for different layers of material of uniform thickness successively placed over the uranium. If Ais the coefficient of absorption of the radiation in a material, the intensity I of the radiation after passing through a thickness dis given by 1= Ioe-'J...d, where 10 is the intensity of the radiation at the surface before the plate was applied. The absorption of the radiation in a layer of air is negligible compared with that of an equal thickness of solid matter. The maximum current* between the plates is proportional to the intensity of the radiation. Preliminary experiments showed that the current diminished very approximately in G.P. with the distance of material traversed, so that the value of A determined was independent of the thickness of the plate. This shows that most of the rays emitted have approximately the same penetrating power. The rays of radium, examined in a similar manner, did not fall off regularly, showing that the rays emitted consist of particles having a wide range of velocities, and, consequently, a wide range of penetrating power. This is clearly shown by Becquerel, who examined (by the photographie method) the amount of deflection of the rays in a magnetic field after passing through different thicknesses of various metals. The following table represents the results obtained. Substance

A

Density

A Density

Glass Mica Ebonite Wood Cardboard Iron Aluminium Copper Silver Lead Tin

14·0 14·2 6·5 2·16 3·7 44 14·0 60 75 122 96

2·45 2·78 1·14 0·40 0·70 7·8 2·60 8·6 10·5 11·5 7·3

5·7 5·1 5·7 5·4 5·3 5·6 5'4 7·0 7·1 10·8 13·2

It will be observed that the value of the coefficient of absorption divided by the density is very approximately the same for such different substances as glass, mica, ebonite, wood, iron, and aluminium, The divergences from the law are, however, great for the other metals examined, viz. copper, silver, lead, and tin. In tin the value of Adivided by the density is 2·5 times its value for iron and aluminium. These differences show that the law of the absorption of cathode rays depending only on the density, is not true for all substances.

*

Rutherford, Phi!. Mag., January 1899.

420

The Collected Papers of Lord Rutherford

Experiments are at present in progress to see whether there is any simple numericaI connection between the vaIues of ,.\ divided by density for different metals, and to extend the results so as to inc1ude a variety of substances in the solid and liquid state. Absorption

0/ the

Rays by Solids and Gases

The rays not acted on by a magnetic field can be distinguished from each other by their power of penetrating through thin layers of metaI, and their absorption in gases. If, on examination, the penetrating power of two types of radiation proves to be the same in each case for all substances, it is extremely probable that the two radiations are identicaI. By examining the diminution of intensity of the radiation when sheets of metaI of the same thickness are placed over the radioactive substance, the homogeneity or complexity of the radiations can be tested. Ifthe intensity I ofthe radiation after passing through a distance of metal is given by loe-Nl, where 10 is the original intensity and ,.\ the coefficient of absorption, we can conc1ude that the radiation is homogeneous in character. If this condition is not fulfilled the radiation is complex. One of us* has at different times given results for the absorption of some of the different radiations in solids and gases. In this paper we have extended the results and compared the different types of radiation under, as far as possible, the same conditions. In the case of both uranium* and thoriumt it has been shown that the absorption of the radiation is the same for all the different compounds of each element examined. When the types of radiation are complex, the relative amount of rays of different types may vary for different compounds, but so far there is no evidence that the actual radiations themselves are aItered. It is only necessary, therefore, to examine one compound of each element for the purpose of comparison of the types of rays emitted. The following substances have been employed in the experiments : Uranium oxide and thorium oxide. Two different sampies of each obtained from Schuchart of Germany and Eimer and Amend of New Y ork gave similar results. Polonium. This substance was kindly prepared for me by Dr Walker of McGill University from pitchblende, after the method described in Curie's first paper. Since that time the intensity of the radiation given off has steadily diminished; but the type of radiation has been unaItered. Radium. Two different' specimens were employed. One was of impure radium chloride kindly presented to me by Elster and Geitel two years ago; this did not give off any emanation and only a small proportion of defl.ectable rays. The other, from P. de Haen, Hannover, was not very strong in deflectable rays, but gave out a large amount of emanation when slightly heated. • E. Rutherford, Phil. Mag., January 1899, February and March 1900. t Owens, Phil. Mag., October 1900.

Comparison of the Radiations from Radioactive Substances

421

In the course of the paper we shall examine the following types of radiation.

A. Uranium. (1) The oe or early absorbed radiation. (2) The ß or defLectable rays. B. Thorium radiations. (3) The simple radiation given out by a thin layer. (4) The radiation from the 'emanation.' (5) The excited radiation. C. Polonium. (6) Simple radiation.

n.

Radium. (7) Radiation not affected by a magnetic field. (8) Radiation from the emanation. (9) Excited radiation. (10) Magnetically defLected rays.

Absorption of Radiation by Metals In examining the absorption of the radiation by metal foil and other substances, the apparatus shown in Fig. 2 was employed. The active compound in the form of fine powder was uniformly spread over a shallow depression 6·5 cm. square in a large lead plate. This corresponds to plate A in Fig. 2. The rate of leak was observed by means of an electrometer between plates A and B with, if necessary, a suitable capacity in parallel. The plate A was connected to one pole of a battery of 300 volts, the other pole of which was earthed. Preliminary experiments showed that with this voltage the maximum or saturation current between the plates was obtained for all the radioactive substances examined. In most of the experiments described in this paper an Ayrton electrometer was used. In some of the later experiments, however, the White pattern of Kelvin electrometer was used. The former electrometer could be readily arranged to give 200 mm. divisions for one volt p.n. between the quadrants. As most of the experiments were carried out during the very dry Canadian winter, it was very essential to screen the electrometer and connections with testing apparatus by wire gauze. Unless precautions of this kind were taken, every movement of the observer produced sufficient frictional electrification to disturb the electrometer. For the same reason, and also for convenience, the quadrants were separated by a cord connected to a suitable key and operated at a distance. The method of observing the rate of leak was as follows: A secondspendulum was placed before the observer. At the instant of passing the middle point of its swing the quadrants were separated by a sudden pull of the cord.

422

Tize Collected Papers

0/ Lord Rutherford

After ten or more swings the connection between testing apparatus and the electrometer was broken by means of an insulated key, operated by a second cord. The deflection of the electrometer needle when it came to rest was then observed. The number of scale divisions passed over, divided by the time between the separation of the two keys, was taken as a measure of the rate of leak. This method is more accurate than the usualone of observing the time the electrometer needle takes to pass over, say, 100 divisions of the scale. The final deflection is independent of the amount of damping and of any oscillation or irregularity in the movement of the electrometer needle. In experiments with uranium, thorium, and polonium a very thin layer of the material was employed. This is essential in the case of thorium oxide, in 100

90

'0

70t &O~

50

~ ~

40~

~

3O~

20 1;, ~ 10

o

o

L.AYGRS bF A't,VMII'IIUM FOJt.

o

o

o

o

Fig. 3 order that the rate of leak due to the emanation from it may be negligible compared with the rate of leak produced by the ordinary radiation. In dealing with radium a very small amount of material was dusted by means of agauze as uniformly as possible over a platinum plate. For the specimen of radium employed the rate of leak due to the emanation and rays deviable by a magnet was in this way rendered negligible compared with the rate of leak due to non-deviable radiation. Suitable capacity was, if necessary, placed in parallel with the electrometer to reduce the rate of leak. In Figs. 3 and 4 curves are given for the absorption of the different radiations by thin aluminium foil and Dutch metal respectively. In order to plot the curves on the same scale the rate of leak for the bare radioactive plate is in each case taken as 100. The average thickness of aluminium foll was O' 00036 cm., and of Dutch metal 0·00012 cm. The curves are given for two specimens ofradium, marked(c) and (e), which corresponded to two specimens of radium from P. de Haen, marked 'concentrated' and 'einfach' respectively. The curve for the specimen sent by Herrn Elster and Geitel was not very

Comparison of the Radiations from Radioacti\'e Suhstances

423

different from the two shown. Curves obtained from specimens of the minerals 'thorite' and 'orangite' gave practically the same curves as for thoria. The radiations may b'

t ae-A1x

dx

(I - e-A1d)

= Eta' 2>'1 1f

\ d' I IS arge.

"1

For simplicity, let us suppose a thick layer of radioactive substance spread uniformly over a large plane area. There seems to be no doubt that the radiations are emitted uniformly from each portion of the mass; consequently, radiation which produces the ionizing action in the gas above the radioactive layer is the sum total of all the radiation which reaches the surface of the layer. Let '\1 be the average coefficient of absorption of the rays in the radioactive substance, and a the specific gravity of the substance.

Deviable Rays of Radioactive Substances

469

Let E, be the total energy radiated per second per unit mass of the substance where the absorption of the rays in the substance itself is disregarded. The energy per second radiated to the surface by a thickness dx of a layer of unit area distant x from the surface is given by ~El ue-Axdx .

In a similar way it may be shown that the energy W2 , radiated per second by the ß rays for a very thick layer is given by E2 u W2 =

2.\2

where E2 is the total rate of emission of energy of unit mass disregarding absorption, and 1.2 the coefficient of absorption of rays in the substance. Therefore WI EI 1.2 W2 = E2 Al' EI AI W I or E 2 = 1.2 W2' It is difficult to determine Al and 1.2 directly for the radioactive substance itself; but it is probable that the ratio is not widely different from the ratio of the absorption coefficients for another substance like aluminium, which has been directly determined. This follows from the general result that the absorptions of 0:: and ß radiations in any substance are approximately proportional to the density of the substance.

~: is the ratio of the number of ions produced by the

0::

to those produced

by the ß rays, and can be determined in the way already explained for a thin layer. E

The ratio E~ has been determined for uranium from the following experimental data. A thick layer of uranium oxide was spread over an area 0:: and ß rays determined between two parallel plates distant 6· 1 cm. apart. It was found that _cu_r_r_en_t-:-d_ue_t_o_o::::--ra_y_s = 12. 7 . current due to ß rays Thus W1 total number of ions due to 0:: rays W2 = total number of ions due to ß rays 12·7 X 6·1 154 = Q. 5 approx., 22 sq. cm., and the ratio of the current produced by the

470

The Collected Papers 0/ Lord Ruther/ord

sinee we have previously ealculated number of ions produeed by the ß rays is 154 times the number produeed for 1 em. distanee between the plates.* Now EI AIWI 0·5AI 0·5 X 2740 E2 = A2 W2 = = 1 A = 100 approx.

--x;-

since the value of AI for aluminium = 2740 for the IX rays and A2 for the ß rays = 14. We therefore see that for uranium about ntoo of the total energy radiated is earried off in the form of electrons. The ratio is still smaller for thorium and radium. It thus appears that in the permanent radioactive substances the electrons driven off represent only a small fraction of the energy dissipated. § 11. Discussion

0/ Results

We have seen that the three well-recognized radioactive substances, uranium, thorium, and radium, all emit both deviable and non-deviable rays. In this respeet they differ from polonium, which gives out no deviable rays. As Becquerel has pointed out, there is little doubt that polonium (radioactive bismuth) cannot be considered as a permanent radioactive substance, for its radiation steadily diminishes with the time. We have shown that uranium gives out more deviable rays than radium or thorium, compared with the amount of non-deviable, but the ratio of the amounts of the two types of rays is of the same order. In eonsidering the question of the relation between the IX and ß rays, the results of the chemical separation are of great importance. It seems certain that we cannot regard the IX rays as having the same relation to ß rays as cathode-rays have to Röntgen rays which they produce; for we have shown that the separated active products from uranium and thorium contain all the substance responsible for the ß rays. The radioactive material, which has thus been temporarily freed from ß rays still, however, retains its power of giving out, in the case of uranium, a large proportion, and in the case of thorium about 30 per cent of the original IX rays. This IX radiation persists, in the case of uranium, several days, and, in the case of thorium, several hours, without any appreciable change in intensity. If the IX rays are due directly to the ß rays, it is necessary to assume that the radiation persists for long intervals after its exciting cause is removed. This view also falls to explain, without additional assumptions, why the radiation from UrX does not excite similar IX radiations in itself. Without, at this stage, going into views on the mechanism of radioactivity, it seems probable that most of the deviable rays from uranium and thorium are given out by a secondary product produced by a disintegration of the • Phll. Mag., June 1902.

Deviable Rays of Radioactive Substances

471

uranium or thorium atom or molecule. These secondary products differ in chemical properties from the uranium and thorium, and can be separated from them by chemical means, and thus give rise to UrX and ThX. The non-deviable radiation may be either due to the other secondary product of the reaction, or may be due to an action of the product responsible for the deviable rays in the mass of the radioactive material. McGill University, Montreal May 7,1902

The Cause and Nature of Radioactivity PART

by E.

I

Macdonald Professor of Physics, SODDY, B.A. (OXON), Demonstrator in Chemistry, McGill University, Montreal

RUTHERFORD, M.A., D.se.,

and F.

From the Philosophical Magazine for September 1902, sero 6, iv, pp. 370-396 (Accounts of these researches, during the progress of the investigation, have already been given to the London Chemica1 Society)

CoNTENTS

I. Introduction. 11. Experimental Methods of investigating Radioactivity. 111. Separation of a Radioactive Constituent (ThX) from Thorium Compounds.

IV. V. VI. VII. VIII. IX. X. XI. XII. XIII.

The Rates of Recovery and Decay of Thorium Radioactivity. The Chemical Properties of ThX. The Continuous Production of ThX. The Inftuence of Conditions on the Changes occurring in Thorium. The Cause and Nature of Radioactivity. The Initial Portions of the Curves of Decay and Recovery. Tbe Non-separable Radioactivity of Thorium. Tbe Nature of the Radiations from Thorium and ThX. Summary of Results. General Tbeoretical Considerations.

I. Introduction THE following papers give the results of a detailed investigation of the radioactivity of thorium compounds which has thrown light on the questions connected with the source and maintenance of the energy dissipated by radioactive substances. Radioactivity is shown to be accompanied by chemical changes in which new types of matter are being continuously produced. These reaction products are at first radioactive, the activity diminishing regularly from the moment of formation. Their continuous production maintains the radioactivity of the matter producing them at adefinite equilibrium value. The conclusion is drawn that these chemical changes must be sub-atomic in character. The present researches had as their starting point the facts that had come to light with regard to thorium radioactivity (Rutherford, Phi!. Mag., 1900, vol. xlix, pp. 1 and 161). Besides being radioactive in the same sense as the uranium compounds, the compounds of thorium continuously emit into the surrounding atmosphere agas which possesses the property of temporary

The ('ausr am/ Nature

0/ Radioactivit)'. I

473

radioactivity. This 'emanation', as it has been named, is the source of rays which ionize gases and darken the photographic film. * The most striking property of the thorium emanation is its power of exciting radioactivity on aIl surfaces with which it comes into contact. A substance after being exposed for some time in the presence of the emanation behaves as if it were covered with an invisible layer of an intensely active material. If the thoria is exposed in a strong electric field, the excited radioactivity is entirely confined to the negatively charged surface. In this way it is possible to concentrate the excited radioactivity on a very small area. The excited radioactivity can be removed by rubbing or by the action of acids, as, for example, sulphuric, hydrochloric, and hydrofluoric acids. If the acids be then evaporated, the radioactivity remains on the dish. The emanating power of thorium compounds is independent of the surrounding atmosphere, and the excited activity it produces is independent of the nature of the substance on which it is manifested. These properties made it appear that both phenomena were caused by minute quantities of special kinds of matter in the radioactive state, produced by the thorium compound. The next consideration in regard to these examples of radioactivity, is that the activity in each case diminishes regularly with the lapse of time, the intensity of radiation at each instant being proportional to the amount of energy remaining to be radiated. For the emanation aperiod of one minute, and for the excited activity aperiod of eleven ho urs, causes the activity to fall to half its value. These actions-(I) the production of radioactive material, and (2) the dissipation of its available energy by radiation-wh ich are exhibited by thorium compounds in the secondary effects of emanating power and excited radioactivity, are in reality taking place in all manifestations of radioactivity. The constant radioactivity of the radioactive elements is the result of an equilibrium between these two opposing processes.

11. The Experimental Methods 01 investigating Radioactivity Two methods are used for the measurement of radioactivity, the electrical and the photographic. The photographic method is of a qualitative rather than a quantitative character; its effects are cumulative with time, and, as a rule, long exposures are necessary when the radioactivity of a feeble agent like thoria is to be demonstrated. In addition, Russell has shown that the darkening of a photographic plate is brought about also by agents of a totally different character from those under consideration, and, moreover, under very general conditions. Sir William Crookes (Proc. Roy. Soc. 1900, lxvi, p. 409) has sounded a timely note of warning against putting too much '" If thorium oxide be exposed to a white heat its power of giving an emanation is to a large extent destroyed. Thoria that has been so treated is referred to throughout as 'de-emanated' .

474

The Collecled Papers of Lord Rutherford

confidence in the indications of the photographic method of measuring radioactivity. The uncertainty of an effect produced by cumulative action over long periods of time quite precludes its use for work of anything but a qualitative character. But the most important objection to the photographic method is that certain types of rays from radioactive substances, which ionize gases strongly, produce little if any effect on the sensitive film. In the case of uranium, these photographically inactive rays form by far the greatest part of the total radiation, and much of the previous work on uranium by the photographie method must be interpreted differently (Soddy, Proc. ehem. Soc., 1902, p. 121). On the other hand, it is possible to compare intensities of radiation by the electrical method with greater rapidity and with an error not exceeding 1 or 2 per cent. These methods are based on the property generally possessed by all radiations of the kind in question, of rendering agas capable of discharging both positive and negative electricity. These, as will be shown, are capable of great refinement and certainty. An ordinary quadrant electrometer is capable of detecting and measuring a difference of potential of at least 10-2 volt. With special instruments, this sensitiveness may be increased a hundredfold. An average value for the capacity of the electrometer and connexions is 3 X 10-5 microfarad; and when this is charged up to 10-2 volt, a quantity of electricity corresponding to 3 X 10-13 coulomb is stored up. Now in the electrolysis of water one gram of hydrogen carries acharge of 105 coulombs. Assuming, for the sake of example, that the conduction of electricity in gases is analogous to that in liquids, this amount of electricity corresponds to the transport of a mass of 3 x 10-18 gram ofhydrogen; that is, a quantity of the order of 10-12 times that detected by the balance. For a more delicate instrument, this amount would produce a large effect. The examples of radium in pitchblende and of the thorium-excited radioactivity make it certain that comparatively large ionization effects are produced by quantities of matter beyond the range of the balance or spectroscope. The electrometer also affords the means of recognizing and differentiating between the emanations and radiations of different chemical substances. By the rate of decay the emanation from thorium, for example, can be instantly distinguished from that produced by radium; and although a difference in the rate of decay does not of itself argue a fundamental difference of nature, the identity of the rate of decay furnishes at least strong presumption of identity of nature. Radiations, on the other hand, can be compared by means of their penetration powers (Rutherford, Phil. Mag., 1899, vol. xlvii, p. 122). If the rays from various radioactive substances are made to pass through successive layers of aluminium foil, each additionallayer of foil cuts down the radiation to a fraction of its former value, and a curve can be plotted with the thickness of metal penetrated as abscissae, and the intensity of the rays after

Ihe

('{lUse

475

{md Na/ure of Radioactil'ilY, I

penetration as ordinates, expressing at a glance the penetration power of the rays under examination. The curves so obtained are quite different for different radioactive substances. The radiations from uranium, radium, thorium, each give distinct and characteristic curves, whilst that of the last named again is quite different from that given by the excited radioactivity produced by the thorium emanation. It has been recently found (Rutherford and Grier, Phys. Zeit., 1902, p. 385) that thorium compounds, in addition to a type of easily absorbed Röntgen rays, non-deviable in the magnetic field, cmit also rays of a very penetrating character deviable in the magnetic

ä1rth

Fig.l

Earth

field. The latter are therefore similar to cathode rays, which are known to consist of material particles travelling with a velocity approaching that of light. But thorium, in comparison with uranium and radium, emits a much smaller proportion of deviable radiation. The determination of the proportion between the deviable and non-deviable rays affords a new means of investigating thorium radioactivity. The electrometer thus supplies the study of radioactivity with methods of quantitative and qualitative investigation, and there is therefore no reason why the cause and nature of the phenomenon should not be the subject of chemical investigation. Fig. 1 shows the general arrangement. From O· 5 to O· 1 gram of the cornpound to be tested, reduced to fine powder, is uniformly sifted over a platinum plate 36 sq. cm. in area.

476

The Collected Papers 0/ Lord Rutherford

This plate was placed on a large metal plate eonneeted to one pole of a battery of 300 volts. the other pole of whieh was earthed. An insulated parallel plate was plaeed about 6 em. above it, and the whole apparatus enclosed in a metal box eonneeted to earth, to prevent eleetrostatie disturbance. The shaded portions in the figure represented insulators. A door was made in the apparatus so that the plate eould be rapidly plaeed in position or removed. Both pairs of quadrants are first eonneeted to earth. On eonneeting the one pair with the apparatus, the defiexion of the needle from zero inereases uniformly with time, and the time taken to pass over 100 divisions of the seale is taken by a stop-wateh. The rate of movement is a measure of the ionization eurrent between the plates. The ratio of the eurrents for different substanees is a eomparative measure of their radioaetivity. With this apparatus O' 5 gr. of thorium oxide produces a eurrent of 1 . 1 X 10-11 amperes, whieh, with the eleetrometer used, working at average sensitiveness, eorresponds to 100 divisions of the seale in 36 seeonds. In eertain eases a special modifieation of the Dolezalek eleetrometer was employed whieh is 100 times more sensitive. With this instrument the radioactivity of t milligram of thoria produces a measurable effeet. If the substance gives off an emanation, the eurrent between the plates inereases with time. Under these eonditions, when the thorium eompound is exposed in thin layers with a maximum of radiating surfaee, aU but one or two per cent of the total effeet is due to the straight line radiation. Even when the effeet due to the emanation has attained a maximum, this eonstitutes a very small fraetion of the whole. This effeet, however, may to a large extent be eliminated by taking the eurrent between the eleetrodes immediately after the material is plaeed in the testing apparatus. It may be eompletely eliminated by passing a eurrent of air between the eleetrodes to remove the emanation as fast as it is formed. The eurrent between the plates observed with the eleetrometer at first inereases with the voltage, but a stage is very so on reaehed when there is a very small inerease for a targe additional voltage. A p.n. of 300 volts was sufficient to obtain the maximum eurrent, so that all the ions reaehed the eleetrodes before any appreciable recombination occurred. It must, however, at onee be pointed out that it is difficult to make any absolute measure of radioactivity. The radiation from thorium is half absorbed by a thickness of aluminium of 0·0004 cm.; and since thorium oxide is far denser than aluminium, it is probable that the radiation in this case is confined to a surface layer only 0·0001 cm. deep. It is obvious that different preparations, each containing the same percentage of thorium but with different densities and different states of division, will not give the same intensity of radiation. In comparing two different specimens of the same compound, it is important that the final steps in their preparation should be the same in each case. As a rule absolute measurements of this kind have been avoided. It is possible, however, to trace with great accuracy the change of radioactivity of any preparation with time by leaving it undisturbed

The ('ause and Nature

0/ Radioactil'ity. I

477

on its original plate, and comparing it with a similarly undisturbed constant comparison sampIe. Most of the investigations have been carried out by this method. III. The Separation 0/ a Radioactive Constituent /rom Thorium Compounds During an investigation of the emanating power of thorium compounds, to be described later, evidence was obtained of the separation of an intensely radioactive constituent by chemical methods. It had been noticed that in certain cases thorium hydroxide, precipitated from dilute solutions of thorium nitrate by ammonia, possessed an abnormally low emanating power. This led naturally to an examination being made of the filtrates and washings obtained during the process. It was found that the filtrates invariably possessed emanating power, although from the nature of their production they are chemically free from thorium. If the filtrate is evaporated to dryness, and the ammonium salts removed by ignition, the small residues obtained exhibit radioactivity also, to an extent very much greater than that possessed by the same weight of thorium. As a rule these residues were of the order of one-thousandth part by weight of the thorium salt originally taken, and were many hundred, in some cases over a thousand, times more active than an equal weight of thoria. The separation of an active constituent from thorium by this method is not all dependent on tbe purity of the salt used. By the kindness of Dr Knöfler, of Berlin, who, in the friendliest manner, presented us with a large specimen of his purest thorium nitrate, we were enabled to test this point. This specimen, which had been purified by a great many processes, did not contain any of the impurities found in the commercial salt before used. But its radioactivity and emanating power were at least as great, and the residues from the filtrates after precipitation by ammonia were no less active than those before obtained. These residues are free from thorium, or, at most, contain only the merest traces, and when redissolved in nitric acid do not appear to give any characteristic reaction. An examination of the penetrating power of the rays from the radioactive residue showed that the radiations emitted were in every respect identical with the ordinary thorium radiation. In another experiment the nature of the emanation from a similar intensely active thorium-free residue was submitted to examination. The rate of decay was quite indistinguishable from that of the ordinary thorium emanation; that is, substances chemically free from thorium have been prepared possessing thorium radioactivity in an intense degree. The thorium hydroxide which had been submitted to the above process was found to be less than half as radioactive as the same weight of thorium oxide. It thus appeared that a constituent responsible for the radioactivity of

478

Tlte Collected Papers of Lord RlI1herJord

thorium had been obtained, which possessed distinct chemical properties and an activity of the order of at least a thousand times as great as the material from which it had been separated. Sir William Crookes (Proe. Roy. Soc., 1900, lxvi, p. 409) succeeded in separating a radioactive constituent of great activity and distinct chemical nature from uranium, and gave the name UrX to this substance. For the present, until more is known of its real nature, it will be convenient to name the active constituent of thorium ThX, similarly. Like UrX, however, ThX does not answer to any definite analytical reactions, but makes its appearance with precipitates formed in its solution even when no question of insolubility is involved. This accords with the view that it is present in infinitesimal quantity, and possesses correspondingly great activity. Even in the case of the most active preparations, these probably are composed of some ThX associated with accidental admixtures large in proportion. These results receive confirmation from observations made on a different method of separating ThX. The experiment was tried of washing thoria with water repeatedly, and seeing if the radioactivity was thereby affected. In this way it was found that the filtered washings, on concentration, deposited small amounts of material with an activity often of the order of a thousand times greater than that of the origmal sampie. In one experiment, 290 grams of thoria were shaken for a long time with nine quantities, each of 2 litres of distilled water. The first washing, containing thorium sulphate present as an impurity, was rejected, the rest concentrated to different stages and filtered at each stage. One of the residues so obtained weighed 6· 4 mg., and was equivalent in radioactivity to 11· 3 grams of the original thoria, and was therefore no less than 1800 times more radioactive. It was examined chemically, and gave, after conversion into sulphate, the characteristic reaction of thorium sulphate, being precipitated from its solution in cold water by warming. No other substance than thorium could be detected by chemical analysis, although of course the quantity was too small for a minute examination. The penetrating power of the radiation from this substance again established its identity with the ordinary thorium radiation. In another experiment, a small quantity of thoria was shaken many times with large quantities of water. In this case, the radioactivity of the residue was examined and found to be about 20 per cent less radioactive than the original sampIe. The influence 0/ Time on the activity 0/ Thorium and ThX. The preparations employed in our previous experiments were allowed to stand over during the Christmas vacation. On examining them about three weeks later it was found that the thorium hydroxide, which had originally possessed only about 36 per cent of its normal activity, had almost completely recovered the usual value. The active residues, on the other hand, prepared by both methods, had almost completely lost their original activity. The chemical separation effected was thus not permanent in character. At this time M. Becquerel's paper (Comptes Rendus, cxxxiii, December 9, 1901, p. 977) came to hand,

n/(: ('ausc

llllc!I\'lIIUI'(, 0/

RadioQClirily.

I

479

in which he shows that the same phenomena of recovery and decay are presented by uranium after it has been partially separated from its active constituent by chemical treatment. A long series of observations was at once started to determine: (1) The rate of recovery of the activity of thorium rendered less active by

removal of ThX; (2) The rate of decay of the activity of the separated ThX;

in order to see how the two processes were connected. The results led to the view that may at onee be stated. The radioactivity of thorium at any time is the resultant of two opposing processes : (1) The production of fresh radioactive material at a constant rate by the thorium compound; (2) The decay of the radiating power of the active material with time.

The normal or constant radioactivity possessed by thorium is an equilibrium mlue. where the rate 0/ increase 0/ radioactivity due to the production 0//resh actil'e material is balanced by the rate 0/ decay 0/ radioactivity 0/ that already /ormed. It is the purpose of the present paper to substantiate and develop this hypo thesis. IV. The Rates 0/ Recovery and Decay 0/ Thorium Radioactivity A quantity of the pure thorium nitrate was separated from ThX in the manner described by several precipitations with ammonia. The radioactivity of the hydroxide so obtained was tested at regular intervals to determine the rate of recovery of its activity. For this purpose the original specimen of O' 5 gram was left undisturbed throughout the whole series of measurements on the plate over which it had been sifted, and was compared always with O' 5 gram of ordinary de-emanated thorium oxide spread similarly on a second plate and also left undisturbed. The emanation from the hydroxide was prevented from interfering with the results by a special arrangement for drawing a current of air over it during the measurements. The active filtrate from the preparation was concentrated and made up to 100 c.c. volume. One quarter was evaporated to dryness and the ammonium nitrate expelled by ignition in a platinum dish, and the radioactivity of the residue tested at the same intervals as the hydroxide to determine the rate of decay of its activity. The comparison in tbis case was a standard sampie of uranium oxide kept undisturbed on a metal plate, whicb repeated work had shown to be a perfectly constant source of radiation. Tbe remainder of the filtrate was used for other experiments. The following table gives an example of one of a numerous series of observations made with different preparations at different times. The

480

The Collected Papers ofLord Rutlle/jord

maximum value obtained by the hydroxide and the original value of the ThX are taken as 100: Time in days

Activity of Hydroxide

0 1 2 3 4 5 6 8 9

44 37 48 54 62 68

Activity of ThX

100 117 100

88 72 53

71

78

29·5 25·2 15·2 11'1

83

10 13

15 17 21 28

96·5

99

100

Fig. 2 shows the curves obtained by plotting the radioactivities as ordinates, and the time in days as abscissae. Curve 11 illustrates the rate of recovery of the activity of thorium, curve I the rate of decay of activity 12D

UD 100 9. 10 70

S. S' t

40 I' 10 10

UD

o

t

T/~

4

,

t

IN DArG

'0

'1

Fig.2

'..

.,

"

2.

21

APPARATUS FOR MEASURING THE RATE OF DECAY OF THORIUM EMANATION

See: 'The Cause and Nature of Radioactivity' (P. Soddy), page 472.

APPARATUS FOR STUDYING THE MAGNETIC AND ELECTRIC DEVIATION OF RAYS FROM RADIUM

Plain parallel plates

ELECTROSCOPE (Iower chamber missing) Plates with lips

Insulated parallel plates, connected in alternate pairs.

See: 'The Magnetic and Electric Deviation ofthe Easily Absorbed Rays from Radium,' page 549.

This page intentionally left blank

l'he Cause amI Nature

oI Radioactil'it)'.

/

4~1

of ThX. It will be seen that neither of the curves is regular for the first two days. The activity of the hydroxide at first actually diminished and was at the same value after two days as when first prepared. The activity of the ThX, on the other hand, at first increases and does not begin to fall below the original value till after the lapse of two days (compare section IX). These results cannot be ascribed to errors of measurement, for they have been regularly observed whenever similar preparations have been tested. The activity of the residue obtained from thorium oxide by the second method of washing decayed very similarly to that of ThX, as shown by the above curve. If for present purposes the initial periods of the curve are disregarded and the later portions only considered, it will be seen at once that the time I:

so 60

40

70

T/~

o

....

....

IN DArG

12.

1(,

20

24

28

Fig.3 taken for the hydroxide to recover one half of its lost activity is about equal to the time taken by the ThX to lose half its activity, viz., in each case about four days, and, speaking generally, the percentage proportion of the lost activity regained by the hydroxide over any given interval is approximately equal to the percentage proportion of the activity lost by the ThX during the same interval. If the recovery curve is produced backwards in the normal direction to cut the vertical axis, it will be seen to do so at a minimum of about 25 per cent, and the above result holds even more accurately if the recovery is assumed to start from this constant minimum, as, indeed, it has been shown to do under suitable conditions (section IX, Fig. 4). This is brought out by Fig. 3, which represents the recovery curve of thorium in which the percentage amounts of activity recovered, reckoned from this 25 per cent minimum, are plotted as ordinates. In the same figure the decay curve after the second day is shown on the same scale. Q

482

The Collected Papers of Lord Rutllerford

The activity of ThX decreases very approximately in a geometrical progression with the time, i.e. if 10 represent the initial activity and l e the activity after time t, I t = e-At,

(1)

10

where A is a constant and e the base of natural logarithms. The experimental curve obtained with the hydroxide for the rate of rise of its activity from a minimum to a maximum value will therefore be approximately expressed by the equation It

Io =

(2)

1 - e-At

'

where 10 represents the amount of activity recovered when the maximum is reached, and lt the activity recovered after time t, A being the same constant as before.

Now, this last equation has been theoretically developed in other places (compare Rutherford, Phi!. Mag., 1900, pp. 10 and 181) to express the rise of activity to a constant maximum of a system consisting of radiating partic1es in which (1) The rate of supply of fresh radiating particles is constant. (2) The activity of each partic1e dies down geometrically with the time according to equation (1).

It therefore follows that if the initial irregularities of the curves are dis-

regarded and the residual activity of thorium is assumed to possess a constant value, the experimental curve obtained for the recovery of activity will be explained if two processes are supposed to be taking place: (1) That the active constituent ThX is being produced at a constant rate;

(2) That the activity of the ThX decays geometrically with time.

Without at first going into the difficult questions connected with the initial irregularities and the residual activity, the main result that follows from the curves given can be put to experimental test very simply. The primary conception is that the major part of the radioactivity of thorium is not due to the thorium at a11, but to the presence of a non-thorium substance in minute amount which is being continuously produced. V. Chemical Properties

0/ ThX

The fact that thorium on precipitation from its solutions by ammonia leaves the major part of its activity in the filtrate does not of itself prove that a material constituent responsible for this activity has been chemically separated. It is possible that the matter constituting the non-thorium part of the solution is rendered temporarily radioactive by its association with thorium, and this

The Cause alU1A'ature

0/ Radioactirity. 1

483

property is retained through the processes of precipitation, evaporation, and ignition, and manifests itself finallyon the residue remaining. This view, however, can be shown to be quite untenable, for upon it any precipitate capable of removing thorium completely from its solution should yield active residues similar to those obtained from ammonia. Quite the reverse, however, holds. When thorium nitrate is precipitated by sodium or ammonium carbonate, the residue from the filtrate by evaporation and ignition is free from activity, and the thorium carbonate possesses the normal value for its activity. The same holds true when oxalic acid is used as the precipitant. This reagent even in strongly acid solution precipitates almost all of the thorium. When the filtrate is rendered alkaline by ammonia, filtered, evaporated, and ignited, the residue obtained is inactive. In the case where sodium phosphate is used as the precipitant in ordinary acid solution, the part that comes down is more or less free from ThX. On making the solution alkaline with ammonia, the remainder of the thorium is precipitated as phosphate, and carries with it the whole of the active constituent, so that the residue from the filtrate is again inactive. In fact ammonia is the only reagent of those tried capable of separating ThX from thorium. The result of Sir William Crookes with uranium, which we have confirmed working with the electrical method, may be here mentioned. UrX is completely precipitated by ammonia together with uranium, and the residue obtained by the evaporation of the filtrate is quite inactive. There can thus be no question that both ThX and UrX are distinct types of matter with definite chemical properties. Any hypothesis that attempts to account for the recovery of activity of thorium and uranium with time must of necessity start from this primary conception. VI. The Continuous Production

0/ ThX

If the recovery of the activity of thorium with time is due to the production of ThX, it should be possible to obtain experimental evidence of the process. The first point to be ascertained is how far the removal ofThX by the method given reduces the total radioactivity of thorium. A preliminary trial showed that the most favourable conditions for the separation are by precipitating in hot dilute solutions by dilute ammonia. A quantity of 5 grams of thorium nitrate, as obtained from the maker, was so precipitated by ammonia, the precipitate being redissolved in nitric acid and reprecipitated under the same conditions successively without lapse 0/ time. The removal of ThX was followed by measuring the activity of the residues obtained from the successive filtrates. The activity of the ThX from the first filtrate was equivalent to 4·25 grams of thoria, from the second to 0·33 gram, and from the third to O· 07 gram. It will be seen that by two precipitations practically the whole of the ThX is removed. The radioactivity of the

484

The Collected Papers 0/ Lord Ruther/ord

separated hydroxide was 48 per cent of that of the standard de-emanated sampie of thoria. Rate 0/ production 0/ ThX. A quantity of thorium nitrate solution that had been freed from ThX about a month before, was again subjected to the same process. The activity of the residue from the filtrate in an experiment in which 10 grams of this nitrate had been employed was equivalent to 8·3 grams of thorium oxide. This experiment was performed on the same day as the one recorded above, in which 5 grams of new nitrate had been employed, and it will be seen that there is no difference in the activity of the filtrate in the two cases. In one month the activity of the ThX in a thorium compound again possesses its maximum value. If aperiod of 24 hours is allowed to elapse between the successive precipitations, the activity of the ThX formed during that time corresponds to about one-sixth of the maximum activity of the total thorium employed. In three hours the activity of the amount produced is about one-thirtieth. The rate of production of ThX worked out from those figures weIl agrees with the form of the curve obtained for the recovery of activity of thorium, if the latter is taken to express the continuous production of ThX at a constant rate and the diminution of the activity of the product in geometrical progression with the time. By using the sensitive electrometer, the course of production of ThX can be followed after extremely short intervals. Working with 10 grams of thorium nitrate, the amount produced in the minimum time taken to carry out the successive precipitations is as much as can be conveniently measured. If any interval is a1lowed to lapse the effect is beyond the range of the instrument, unless the sensitiveness is reduced to a fraction of its ordinary value by the introduction of capacities into the system. Capacities of 0·0 land O· 02 microfarad, which reduce the sensitiveness to less than one two-hundredth of the normal, were frequently employed in dealing with these active residues. The process of the production of ThX is continuous, and no alteration was observed in the amount produced in a given time after repeated separations. In an experiment carried out for another purpose (section IX) after 23 successive precipitations extending over 9 days, the amount formed during the last interval was as far as could be judged no less than what occurred at the beginning of the process. The phenomenon of radioactivity, by means of the electrometer as its measuring instrument, thus enables us to detect and measure changes occurring in matter after a few minutes interval, which have never yet been detected by the balance or suspected of taking place. VII. lnfluence

0/ Conditions on the Changes occurring in Thorium

It has been shown that in thorium compounds the decay of radioactivity

with time is balanced by a continuous production of fresh active material. The change which produces this material must be chemical in nature, for

The Cause ami Nature (!f' Radioactivity. I

485

the products of the action are different in chemical properties from the thorium from wh ich they are produced. The first step in the study of the nature of this change is to examine the effects of conditions upon its rate. Effect 0/ conditions on the rate 0/ decay. Since the activity of the products affords the means of measuring the amount of change, the influence of conditions on the rate of decay must be first found. It was observed that, like all other types of temporary radioactivity, the rate of decay is unaltered by any known agency. It is unaffected by ignition and chemical treatment, and the material responsible for it can be dissolved in acids and re-obtained by the evaporation of the solution, without affecting the activity. The following experiment shows that the activity decays at the same rate in solutions as in the solid state. The remainder of the solution that had been used to determine the decay curve of ThX (Fig. 2) was allowed to stand, and at the end of 12 days a second quarter was evaporated to dryness and ignited, and its activity compared with that of the first which had been left since evaporation upon its original platinum dish. The activities of the two specimens so compared with each other were the same, showing that in spite of the very different conditions the two fractions had decayed at equal rates. After 19 days a third quarter was evaporated, and the activity, now very smalI, was indistinguishable from that of the fraction first evaporated. Resolution ofthe residues after the activity had decayed does not at all regenerate it. The activity of ThX thus decays at a rate independent of the chemical and physical condition of the molecule. Thus the rate of recovery of activity under different conditions in thorium compounds affords a direct measure of the rate of production of ThX und er these conditions. The following experiments were performed: One part of thorium hydroxide newly separated from ThX was sealed up in a vacuum obtained by a good Töpler pump, and the other part exposed to air. On comparing the sampIes 12 days later no difference could be detected between them either in their radioactivity or emanating power. In the next experiment a quantity of hydroxide freed from ThX was divided into two equal parts; one was exposed for 20 hours to the heat of a Bunsen burner in a platinum crucible, and then compared with the other. No difference in the activities was observed. In a second experiment one half was ignited for 20 minutes on the blast, and then compared with the other with the same result. The difference of temperature and the conversion of thorium hydroxide into oxide thus exercised no infiuence on the activity. Some experiments that were designed to test in as drastic a manner as possible the effect of the chemical condition of the molecule on the rate of production ofThX broughtto light small differences, but these arealmost certainly to be accounted for in another way. It will be shown later (section IX) that about 21 per cent ofthe normal radioactivity ofthorium oxide under ordinary conditions consists of a secondary activity excited on the mass of the material. This portion is of course a variable, and since it is divided among the total amount of matter present, the conditions of aggregation, etc., will affect the

486

The Collected Papers

0/ Lord Rutherford

value of this part. This effect of excited radioactivity in thorium makes a certain answer to the question difficult, and on this account the conc1usion that the rate of production of ThX is independent of the molecular conditions is not final. The following experiment, however, makes it extremely probable. A quantity of thorium nitrate as obtained from the maker was converted into oxide in a platinum crucible by treatment with sulphuric acid and ignition to a white heat. The de-emanated oxide so obtained was spread on a plate, and any change in radioactivity with time, which under these circumstances could certainly be detected, was looked for during the first week from preparation. None whatever was observed, whereas if the rate of production of ThX in thorium nitrate is different from that in the oxide, the equilibrium point, at which the decay and increase of activity balance each other, will be altered in consequence. There should have therefore occurred a logarithmic rise or fall from the old to the new value. As, however, the radioactivity remained constant, it appears very probable that the changes involved are independent of the molecular condition. It will be seen that the assumption is here made that the proportion of excited radioactivity in the two compounds is the same, and for this reason compounds were chosen which possess but low emanating power. (Compare section IX, last paragraph.) Uranium is a far simpler example of a radioactive element than thorium, as the phenomena of excited radioactivity and emanating power are here absent. The separation of UrX and the recovery of the activity of the uranium with time appear, however, analogous to these processes in thorium, and the rate of recovery and decay of uranium activity are at present under investigation. It is proposed to test the infiuence of conditions on the rate of change more thoroughly in the case of uranium, as here secondary changes do not interfere. VIII. The Cause and Nature

0/ Radioactivity

The foregoing conc1usions enable a great generalization to be made in the subject of radioactivity. Energy considerations require that the intensity of radiation from any source should die down with time unless there is a constant supply of energy to replace that dissipated. This has been found to hold true in the case of all known types of radioactivity with the exception of the 'naturally' radioactive elements-to take the best established cases, thorium, uranium, and radium. It will be shown later that the radioactivity of the emanation produced by thorium compounds decays geometrically with the time under all conditions, and is not affected by the most drastic chemical and physical treatment. The same has been shown by one of us (Phil. Mag., 1900, p. 161) to hold for the excited radioactivity produced by the thorium emanation. This decays at the same rate whether on the wire on which it is originally deposited, or in solution of hydrochloric or nitric acid. The excited radioactivity produced by the radium emanation appears analogous. All

The Cause and Na ture of Radioactivity. I

487

these examples satisfy energy considerations. In the ca se of the three naturally occurring radioactive elements. however. it is obvious that there must be a continuous replacement of the dissipated energy, and no satisfactory explanation has yet been put forward. The nature of the process becomes c1ear in the light of the foregoing results. The material constituent responsible for the radioactivity, when it is separated from the thorium which produces it, then behaves in tbe same way as the other types of radioactivity cited. Its activity decays geometrically with the time, and the rate of decay is independent of the molecular conditions. The normal radioactivity is, however, maintained at a constant value by a chemical change which produces fresh radioactive material at a rate also independent of the conditions. The energy required to maintain the radiations will be accounted for if wc suppose that the energy of the system after the change has occurred is less than it was before. The work of Crookes and Becquerel on the separation of UrX and the recovery of the activity of the uranium with time, makes it appear extremely probable that the same explanation holds true for this element. The work of M. and Mme Curie, the discoverers of radium, goes to show that this body easily suffers a temporary decrease of its activity by chemical treatment, the normal value being regained after the lapse of time, and this can be weH interpreted on the new view. All known types of radioactivity can thus be brought under the same category. IX. The Initial Portions

0/ the Curves 0/ Decay and Recovery

The curves of the recovery and decay of the activities of thorium and ThX with time suggested the explanation that the radioactivity of thorium was being maintained by the production of ThX at a constant rate. Before this can be considered rigidly established, two outstanding points remain to be c1eared up. (1) What is the meaning of the early portion of the curves? The recovery curve drops before it rises, and the decay curve rises before it drops. (2) Why does not the removal of ThX render thorium completely inactive? A large proportion of the original radioactivity is not affected by the removal of ThX. A study of the curves (Fig. 2) shows that in each case a double action is probably at work. It may be supposed that the normal decay and recovery are taking place, but are being masked by a simultaneous rise and decay from other causes. From what is known of thorium radioactivity, it was surmised that an action might be taking place similar to that effected by the emanation of excited radioactivity on surrounding inactive matter. It will be shown later that the ThX, and not thorium, is the cause of the emanating power of thorium compounds. On this view, the residual activity of thorium might consist in whole or in part of a secondary or excited radioactivity produced on the whole mass of the thorium compound by its association with the ThX. The drop in the recovery curve on this view would be due to the

488

The Collected Papers o[ Lord RUfherjord

decay of this excited radioactivity proceeding simultaneously with, and at first reversing, the effect of the regeneration of ThX. The rise of the decay curve would be the increase due to the ThX exciting activity on the matter with which it is associated, the increase from this cause being greater than the decrease due to the decay of the activity of the ThX. It is easy to put this hypothesis to experimental test. If the ThX is removed from the thorium as so on as it is formed over a sufficient period, the former will be prevented from exciting activity on the latter, and that already excited will decay spontaneously. The experiment was therefore performed. A quantity of nitrate was precipitated as hydroxide in the usual way to remove ThX, the precipitate redissolved in nitric acid, and again precipitated after a certain interval. From time to time a portion of the hydroxide was removed and its radioactivity tested. In this way the thorium was precipitated in all 23 times in aperiod of 9 days, and the radioactivity reduced to a constan t minimum. The following table shows the results: Activity of hydroxide per cent

After first precipitation After precipitations at three intervals of 24 hours At three more intervals each of 24 hours, and three more each of 8 hours At three more each of 8 hours At six more each of 4 hours

46 39

22 24 25

The constant. minimum thus attained-about 25 per cent of the original activity-is thus about 21 per cent below that obtained by two successive precipitations without interval, which has been shown to remove all the ThX separable by the process. The rate of recovery of this 23 times precipitated hydroxide was then measured (Fig. 4). It will be seen that it is now quite normal, and the initial drop characteristic of the ordinary curve is quite absent. It is in fact almost identical with the ordinary curve (Fig. 2) that has been produced back to cut the vertical axis, and there is thus no doubt that there is a residual activity of thorium unconnected apparently with ThX, and constituting about one fourth of the whole. The decay curves of several of the fractions of ThX separated in this experiment after varying intervals of time were taken for the first few days. All of them showed the initial rise of about 15 per cent at the end of 18 hours, and then anormal decay to zero. The position is thus proved that the initial irregularities are caused by the secondary radiation excited by ThX upon the surrounding matter. By suitably choosing the conditions the recovery curve can be made to rise normally from a constant minimum, and the decay curve be shown to consist of two curves, the first, the rate of production of excited radioactivity, and the second, the rate of decay of the activity as a whole.

The Cause and Nature

0/ Radioactivity.

I

489

So far nothing has been stated as to whether the excited radioactivity which contributes about 21 per cent of the total activity of thorium is the same or different from the known type produced by the thorium emanation. All that has been assumed is that it should follow the same general law, i.e. the effect will increase with the time of action of the exciting cause, and decrease with time after the cause is removed. If the rate of rise of the excited activity be worked out from the curves given (Fig. 5) it will be found to agree with that of the ordinary excited activity, i.e. it rises to half value in about 12 hours. Curve I is the observed decay curve for ThX; curve 11 I~D

so 60

40

20

TIME INOArs

I

o

4

8

12

16

20

Fig.4 is the theoretical curve, assuming that it decreases geometrically with time and falls to half value in four days. Curve III is obtained by plotting the difference between these two and, therefore constitutes the curve of excited activity. Curve IV is the experimental curve obtained for the rise of the excited radioactivity from the thorium emanation when the exciting cause is constant. But the exciting cause (ThX) in the present case is not constant, but is itself falling to half value in 4 days, and hence the difference curve, at first almost on the other, drops away from it as time goes on, and finally decays to zero. There is thus no reason to doubt that the effect is the same as that produced by the thorium emanation, which is itself a secondary effect of ThX. Curve III (Fig. 2) represents a similar difference curve for the decay of excited activity, plotted from the recovery curve of thorium. Since this effect of excited activity is caused by the emanation, it seemed reasonable to suppose that it will be greater, the less the emanation succeeds in escaping in the radioactive state, and therefore that de-emanated compounds should possess a greater proportion of excited radioactivity than Q*

490

The Collected Papers of Lord Rutherford

those with high emanating power. This conclusion was tested by converting a specimen of thorium carbonate with an emanating power :live times that of ordinary thoria, into oxide and de-emanating by intense ignition. The energy that had escaped in the form of emanation is now, all but a few per cent, prevented from escaping. The radioactivity of the oxide so prepared rose in the first three days about thirty per cent of its original amount, and there thus seem to be grounds for the view that the excited radioactivity will 1'1.0

100

80 60 20

-40

20

o

TIME INOArs 2

2

,

4

5

6

Fig.5 contribute a much greater effect in a non-emanating thorium compound than in one possessing great emanating power. Additional confirmation of this view is to be found in the nature of the radiations emitted by the two classes of compounds (section XI). X. The Non-separable Radioactivity 0/ Thorium It has not yet been found possible by any means to free thorium from its residual activity, and the place of this part in the scheme of radioactivity of thorium remains to be considered. Disregarding the view that it is a separate phenomenon, and not connected with the major part of the activity, two hypotheses can be brought forward capable of experimental test, and in accordance with the views advanced on the nature of radioactivity, to account for the existence of this part. First, if there was a second type of excited activity produced by ThX similar to that known, hut with a very

The Cause ami Nature of Radioactirity. I

491

slow rate of decay, it would account for the existence of the non-separable activity. If this is true it will not be found possible to free thorium from this activity by chemical means, but the continuous removal of ThX over a very long period would, as in the above case, cause its spontaneous decay. Secondly, if the change which gives rise to ThX produced a second type of matter at the same time, i.e. if it is of the type of a decomposition rather than a depolymerization, the second type would also in all probability be radioactive, and would cause the residual activity. On this view the second type of matter should also be amenable to separation by chemical means, although it is certain from the failure of the methods already tried that it resembles thorium much more closely than ThX. But until it is separated from the thorium producing it, its activity will not decay spontaneously. Thus what has already been shown to hold for ThX will be true for the second constituent if methods are found to remove it from the thorium. It has been shown (Soddy, loc. eit.) that uranium also possesses a nonseparable radioactivity extremely analogous to that possessed by thorium, and whatever view is taken of the one will in all probability hold also for the other. This consideration makes the second hypothesis, that the residual activity is caused by a second non-thorium type of matter produced in the original change, the more probable of the two. XI. The Nature of the Radiations from Thorium and ThX From the view of radioactivity put forward it necessarily follows that the total radioactivity of thorium is altered neither in character nor amount by chemical treatment. With regard to the first, the amount of activity, it has been pointed out that the intensity of radiations emitted do not furnish alone a measure of the activity. The absorption in the mass of material must be considered also. The radiations of thorium oxide are derived from a very dense powder; those from ThX, on the other hand, have only to penetrate a very thin film of material. The difficulty can be overcome to some extent by taking for the comparison the radioactivity of a thin film of a soluble thorium salt produced by evaporating a solution to dryness over a large metal plate. Compared in this way, the radioactivity of ThX when first separated almost exactly equals the activity of the nitrate from which it is produced, while the hydroxide retains about two-fifths of this amount. The total activity of the products is therefore greater than that of the original salt; but this is to be expected, for it is certain that more absorption takes place in the filtrate than in the products into which it is separated. Similar difficulties stand in the way of an answer to the second question, whether the nature of the radiations is affected by chemical treatment, for it has been experimentally observed that the penetrating power of these radiations decreases with the thickness of material traversed. The character of the radiations from ThX and thorium have, however, been compared by the method of penetration power. A large number of comparisons justifies

492

The Collected Papers of Lord Rutherford

the view that the character of thorium radioactivity is unaltered by chemical treatment and the separation of ThX, although the different types are unequally distributed among the separated products. Determinations of the proportion of rays deviable by the magnetic field in thorium and ThX throws fresh light on the question. The general result is that ThX gives out both deviable and non-deviable rays, and the same applies to the excited activity produced by ThX. But in the experiment in which the excited radiation was allowed to spontaneously decay, by removing ThX as formed, the thorium compound obtained after 23 precipitations was found to be quite free from deviable radiation. This is one of the most striking resemblances between the non-separable radioactivities of uranium and thorium, and warrants the question whether the primary radiation of ThX is not, like that of UrX, composed entirely of cathode rays. There is, however, no means of deciding this point owing to the excited radiation which always accompanies the primary radiation of ThX, and which itself comprises both types of rays. Finally, it may be mentioned that the proportion of deviable and nondeviable radiation is different for different compounds of thorium. The nitrate and ignited oxide, compounds which hardly possess any emanating power, have a higher proportion of deviable radiation than compounds with great emanating power. This is indirect evidence of the correctness of the view already put forward (section IX), that when the emanation is prevented from escaping it augments the proportion of excited radioactivity of the compound. XII. Summary 0/ Results The foregoing experimental results may be briefiy summarized. The major part of the radioactivity of thorium-ordinarily about 54 per cent-is due to a non-thorium type of matter, ThX, possessing distinct chemical properties, which is temporarily radioactive, its activity falling to half value in about four days. The constant radioactivity of thorium is maintained by the production of this material at a constant rate. Both the rate of production of the new material and the rate of decay of its activity appear to be independent of the physical and chemical condition of the system. The ThX further possesses the property of exciting radioactivity on surrounding inactive matter, and about 21 per cent of the total activity under ordinary circumstances is derived from this source. Its rate of decay and other considerations make it appear probable that it is the same as the excited radioactivity produced by the thorium emanation, which is in turn produced by ThX. There is evidence that, if from any cause the emanation is prevented from escaping in the radioactive state, the energy of its radiation goes to augment the proportion of excited radioactivity in the compound. Thorium can be freed by suitable means from both ThX and the excited radioactivity which the latter produces, and then possesses an activity about 25 per cent of its original value, below which it has not been reduced. This

The Cause and Nature

0/ Radioactivity. I

493

residual radiation consists entirely of rays non-deviable by the magnetic field, whereas the other two components comprise both deviable and nondeviable radiation. Most probably this residual activity is caused by a second non-thorium type of matter produced in the same change as ThX, and it should therefore prove possible to separate it by chemical methods. XIII. General Theoretical Considerations Turning from the experimental results to their theoretical interpretation, it is necessary to first consider the generally accepted view of the nature of radioactivity. It is weIl established that this property is the function of the atom and not of the molecule. Uranium and thorium, to take the most definite cases, possess the property in whatever molecular condition they occur, and the former also in the elementary state. So far as the radioactivity of different compounds of different density and states of division can be compared together, the intensity of the radiation appears to depend only on the quantity of active element present. It is not at all dependent on the source from which the element is derived, or the process of purification to which it has been subjected, provided sufficient time is allowed for the equilibrium point to be reached. It is not possible to explain the phenomena by the existence of impurities associated with the radioactive elements, even if any advantage could be derived from the assumption. For these impurities must necessarily be present always to the same extent in different specimens derived from the most widely different sources, and, moreover, they must persist in unaltered amount after the most refined processes of purification. This is contrary to the accepted meaning of the term impurity. All the most prominent workers in this subject are agreed in considering radioactivity an atomic phenomenon. M. and Mme Curie, the pioneers in the chemistry of the subject, have recently put forward their views (Comptes Rendus, cxxxiv, 1902, p. 85). They state that this idea underlies their whole work from the beginning and created their methods of research. M. Becquerel, the original discoverer of the property for uranium, in his announcement of the recovery of the activity of the same element after the active constituent had been removed by chemical treatment, points out the significance of the fact that uranium is giving out cathode rays. These, according to the hypotheses of Sir William Crookes and Professor J. J. Thomson, are material partic1es of mass one thousandth of the hydrogen atom. Since, therefore, radioactivity is at once an atomic phenomenon and accompanied by chemical changes in which new types of matter are produced, these changes must be occurring within the atom, and the radioactive elements must be undergoing spontaneous transformation. The results that have so far been obtained, which indicate that the velocity of this reaction is unaffected by the conditions, makes it c1ear that the changes in question are different in character from any that have been before dealt with in

494

The Collected Papers of Lord Rutherford

chemistry. It is apparent that we are dealing with phenomena outside the sphere of known atomic forces. Radioactivity may therefore be considered as a manifestation of subatomic chemical change. The changes brought to knowledge by radioactivity, although undeniably material and chemical in nature, are of a different order of magnitude from any that have before been dealt with in chemistry. The course of the production of new matter which can be recognized by the electrometer, by means of the property of radioactivity, after the lapse of a few hours or even minutes, might conceivably require geological epochs to attain to quantities recognized by the balance. However the well-defined chemical properties of both ThX and UrX are not in accordance with the view that the actual amounts involved are of this extreme order of minuteness. On the other hand, the existence of radioactive elements at all in the earth's erust is an apriori argument against the magnitude of the change being anything but smal1. Radioactivity as a new property of matter eapable of exact quantitative determination thus possesses an interest apart from the peculiar properties and powers which the radiations themselves exhibit. Mme Curie, who isolated from pitchblende a new substance, radium, which possessed distinct ehemical properties and spectroseopic lines, used the property as a means of chemieal analysis. An exact parallel is to be found in Bunsen's discovery and separation of caesium and rubidium by means of the speetroscope. The present results show that radioactivity can also be used to follow chemical changes occurring in matter. The properties of matter that fulfil the necessary conditions for the study of chemical change without disturbance to the reacting system are few in number. It seems not unreasonable to hope, in the light of the foregoing results, that radioactivity, being such a property, affords the means of obtaining information of the processes occurring within the ehemical atom, in the same way as the rotation of the plane of polarization and other physical properties have been used in chemistry for the investigation of the course of molecular change. Macdonald Physics Building Macdonald Chemistry and Mining Building McGill University, Montreal

The Cause and Nature of Radioactivity PART

by E.

11

Macdonald Professor of Physics, and B.A. (OXON), Demonstrator in Chemistry, McGill University, Montreal

R UTHERFORD, M.A., D.se., F. SODDY,

From the Philosophical Magazine for November 1902, sero 6, iv, pp. 569-585

CONTENTS

I. 11. 111. IV. V. VI. VII. VIII. IX.

Introduction. Method of measuring emanating power. The de-emanation of thoria and the regeneration of the emanating power. The effect of conditions upon emanating power. The cause of the emanating power of thorium. The chemical nature of the emanation. The nature of emanating power. The excited radioactivity from thorium. Further theoretical considerations.

I. Introduction

THE investigation of the radioactivity of thorium, detailed in the first part of this communication, * arose out of an examination of the power possessed hy thorium compounds of giving out a radioactive emanation. The nature of this property and its relation to the radioactivity of thorium remain to be considered. A short resume of what was known at the commencement of the work may be of interest. Thorium radioactivity was discovered by Schmidt and Curie independently in 1898, and Owens in the following year investigated its nature in detail (Phi!. Mag., 1899, p. 360). He observed the inconstancy of the radiation and the effect of air currents in reducing its value. The discovery of the thorium emanation which explained these results, and its power of exciting activity on surrounding matter, following shortly after (Rutherford, Phi!. Mag., 1900, pp. 1 and 161). It was shown that the radiation from the emanation decays rapidly, hut at a perfectly defined rate, falling to ahout one-half the original value at the end of one minute. The emanation passes unchanged through cotton-wool, weak and strong sulphuric acid, and aluminium and other metals in the form of foil, but not through an extremely thin sheet of mica. The emanating

* For Part I

see Phi!. Mag., September 1902 (p. 472 of this volume).

496

The Collected Papers of Lord Rutherford

power of thoria is independent of the surrounding atmosphere, but is destroyed to a large extent by intense ignition, and does not return when the substance is kept. There is a very elose connection between the excited radioactivity produced by thorium compounds and the emanation. It was shown that the amount of the former produced under various conditions was proportional to the amount of the latter, and if the emanating power of thoria be destroyed by ignition, its power to excite radioactivity correspondingly disappears. Simultaneously with the appearance of the papers referred to, Curie showed that radium also possessed the power of exciting activity on surrounding objects. Later, Dom (Abk. der Natur/orsch. Ges. für Halle-a.-S., 1900) repeated the work quoted for thoria, and extended it to inc1ude two preparations of radioactive barium compounds (radium) prepared by P. de Haen, and apreparation of radioactive bismuth (polonium). He found that radium, but not polonium, gave an emanation, especially on warming, and this possessed the power of exciting activity on surrounding objects. Radium and thorium are in this respect completely analogous and different from other radioactive substances, but the phenomena in the two cases are quite different. The emanation from radium retains its activity for many weeks, while the excited radioactivity it produces, on the other hand, decays much more rapidly than that from thorium. One of the most interesting advances in this connection was made during the progress of the work by Elster and Geitel (Phys. Zeit., 1901, Ü, p. 590), who found that it is possible to produce excited radioactivity from the atmosphere, without further agency, by simply exposing a wire highly charged to a negative potential in the air for many hours, and that this also possesses the property of being dissolved off by acids, and of being left behind unchanged on the evaporation of the latter. But here again the rate of decay is different from that of the excited radioactivity produced by thorium. At the commencement of the work the presumption seemed to be in favour of considering emanating power as aseparate phenomenon not directly connected with the ordinary radioactivity of thorium. The former could be destroyed in thorium oxide by ignition without reducing the latter. Later many external conditions were found to affect the value of emanating power without influencing the radioactivity. The nature of the phenomenon had been fully examined from this point of view with very puzzling results, but the conelusion was arrived at that emanating power is probably the manifestation of a change of the nature of a chemical reaction. The discovery of ThX and its continuous production, however, revealed tbe true interpretation of the results, and enables a fairly complete explanation of the phenomenon to be given.

11. Method 0/ Measuring Emanating Power The emanation from thorium (and from radium) behaves in all respects like a temporarily radioactive gas, and diffuses rapidly through porous substances

The Cause and Nature

0/ Radioactirity.

497

II

as, for example, thick cardboard, wbicb are completely opaque to the straight line radiation. Each particle of the emanation behaves as if it were a radiating centre, producing charged carriers throughout the gas in its neighbourhood. The emanation passes through plugs of cotton-wool and can be bubbled through liquids without appreciable loss ofradioactivity, whereas the charged carriers, produced by tbe emanation in common with the straight line radiation from radioactive substances, on the contrary, completely disappear on passing through a plug of cotton- or glass-wool, or by bubbling through liquids. The means of eliminating the effects of the straight line radiation and of measuring the amount of the emanation alone thus suggest themselves. Air passed over uranium or other non-emanating radioactive substance will no longer conduct a current after passage through cottonwool. The conductivity in the case of thorium, however, will persist, and afford a measure of the amount of emanation present. FROH 'GAS~METER

EARTH

E

C

F

H

0

A

EARTH

Fig.l Fig. 1 shows the experimental arrangement for comparing the emanating power of substances. These are placed in the form of fine powder in a shallow lead vessel inside the glass cylinder, C, 17 cm. in length and 3·25 cm. in diameter, provided with indiarubber corks. A current of air from a large gas-bag, after passing through a tube containing cotton-wool to remove dust particles, bubbled through sulphuric acid in the vessel A. It then passed through a bulb containing tightly packed cotton-wool to prevent any spray being carried over. The emanation mixed with air was carried from the vessel C through a plug of cotton-wool, D, which completely removed all the charged carriers carried with the emanation. The latter then passed into a long brass cylinder 75 cm. in length and 6 cm. in diameter. The cylinder insulated on paraffin blocks was connected to one pole of a battery of small lead accumulators, the other pole of which was connected to earth. Three electrodes E, F, H, of equal length were placed along the axis of the cylinder. The current through the gas was measured by means of a Kelvin electrometer of the White pattern. The electrometer and connections were suitably screened by means of wire gauze connected to earth. An insulating key was arranged so that either of the electrodes E,

498

The Collected Papers o[ Lord Ruther[ord

F, H, or aU of them together, could be rapidly connected to one pair of

quadrants of the e1ectrometer, the other two being always connected to earth. The measurements were carried out in the usual way by observing the rate of movement of the electrometer-needle after the one pair of quadrants were connected with the electrodes. On placing the emanating substance in C and continuing the air current for several minutes at a constant rate, the current through the gas due to the emanation attains a steady state. The number of divisions of the scale passed over per second may be taken as a measure of the current. With this apparatus the emanation from 10 gr. of ordinary thorium oxide produces a current of 3·3 X 10-11 ampere between the three electrodes connected together and the cylinder. With the electrometer working at an average sensitiveness, this corresponds to a deflection of 100 divisions of the scale in 12 seconds, so that one-hundredth part of this current could be readily measured-that is, the emanation produced by one-tenth of a gram of thorium oxide. The electrometer one hundred times more sensitive than this failed to detect the presence of an emanation or radioactivity in the oxides of tin, zirconium, and titanium, the other elements of the same group in the periodic table. Rate o[ Decay 0/ the Radiation [rom the Emanation. The three electrodes E, F, H, were used to compare the 'rates of decay' of the radiations from the emanations of different substances. In the previous papers quoted, it has been shown that the radiating power of the thoria emanation falls to half its value in about a minute. In consequence of this, the current observed for the electrode Eis greater than for the electrode H. Knowing the velocity of air along the cylinder and the respective currents to the electrodes E, F, H, the rate of decay of the radiation can be readily deduced. If, however, we merely require to compare the rate of decay of one emanation with another, it is only necessary to compare the ratio of the currents to the electrodes E, F, H in each case, keeping the current of air constant. If the ratio of the currents is the same we may conc1ude that the radiating power of each diminishes at the same rate. The comparison of the emanation is thus rendered qualitative as well as quantitative. In most of the experiments the current to the electrode E was about twice that to the electrode H; the velocity of the current of air along the cylinder was thus about 0·8 cm. a second. Comparison 0/ Emanating Power. The experiments in all cases on the amount of emanation from different substances are comparative. The standard of comparison was usua1ly a sampie of 10 gr. of thoria as obtained from the maker, which gave out a conveniently measurable quantity of emanation. Preliminary experiments were made to find the connection between the weight of thoria and the amount of emanation as tested in the cylinder. The following numbers show that the amount of emanation is within the limits of accuracy desired directly proportional to the weight of substance:

The Cause and Nature of Radioactil'ity. II Weight of thoria

2 gr. 4 gr. 10 gr. 20 gr.

499

Divisions of scale per second

1·41 2·43 6·33 13·2

Correction Jor Natural Leakage. Even with no emanating material in C the electrometer generally indicates a slight movement on separating the quadrants. This is caused by a small current, chiefly made up of leakage due to conduction over the ebonite, as well as the current produced by the excited radioactivity which has collected on the negative electrode during the course of the day's experiments. It varies from day to day, and is as a rule negligible; but in case of bodies possessing very low emanating power it is necessary to correct for it. The number of divisions of the scale per second indicated by the electrometer needle when no emanating material is present is subtracted from the number obtained with the specimen being tested. The corrected number indicates the current due to the emanation alone. Alternative Method oJ Comparing Emanating Power. The apparatus (Fig. 1) described in the first paper (Phi!. Mag., September 1902) for the comparison of radiations, can also be quite weIl employed for a comparison of emanating power. In this case, a thick layer of thoria (several grams) is spread over the plate and covered with two thicknesses of ordinary paper. This has been found almost completely to stop the straight line radiation, whilst allowing the emanation to pass through unimpeded. The current is now measured when a steady state has been reached, due to the accumulation of the emanation. This takes some time, and draughts of air must be guarded against. For this reason, it is less convenient than that first described, but the results obtained by the two methods are almost exactly the same. Thus a sampIe of 'de-emanated' thoria, which gave 12 per cent of the emanating oower of the comparison sampIe by the first method, gave 13 per cent by the second method, whilst a sampIe of oxide prepared from thorium oxalate gave 37 per cent and 39 per cent by the two methods respectively. This elose agreement in the values by methods so completely different in character is a proof that the indications of the methods are worthy of a great degree of confidence. 111. The De-emanation oJ Thoria and the Regeneration oJ the Emanating Power

The emanating power of thoria, as has been stated, is destroyed to a large extent by intense ignition. A eloser study of this is the first step in the investigation of the phenomenon. Previous experiments had not succeeded in completely de-emanating thoria, although a reduction to about 15 per

500

The Collected Papers 0/ Lord Ruther/ord

cent of its original value had been accomplished. A sampIe of this preparation which had been kept for two years had not altered from tbis value. An experiment was performed in which thoria was heated for one hour by means of a powerful gasoline furnace to the highest temperature wbich could be safely employed with platinum vessels. The temperature was such that the fireclay walls fused, and the pipec1ay of a triangle showed signs of having been softened. It was found that the sampIe retained about 8 per cent of its original emanating power. In another experiment, a small platinum crucible fil1ed with thoria was heated for half an hour in a small furnace by a large blowpipe and powerful pair of bellows. Some asbestos-wool had completely fused on the outside of the crucible, and the temperature was probably but little lower than in the previous experiment. This sampIe also retained about 8 per cent of its emanating power. No further attempt has yet been made to completely destroy the emanating power. A small quantity of thoria heated in a platinum crucible in the open over an ordinary-sized blowpipe and bellows for five minutes retained about 45 per cent of its emanating power.The effect of time as well as of temperature was studied by heating about equal quantities in a platinum crucible over an ordinary Bunsen burner for different periods. Heated 10 min. Heated 1 hr. Heated 24 hr.

Emanating power = 61 per cent Emanating power = 59 per cent Emanating power = 42 per cent

It thus appears that there is a large and practically sudden decrease of emanating power for each temperature above a red heat, followed by a very gradual decrease with time when the temperature is maintained; thus, 5 min. on the blowpipe, whilst much more effective than the same time at the temperature of the Bunsen burner, produced rather less effect than 24 hr. at the latter temperature. Effect 0/ Moisture. The next point to be examined was whether the loss of emanating power could be attributed to the loss of water and desiccation of the thoria by ignition. A sampIe of de-emanated thoria (retaining about 14 per cent) was placed in the middle of a Jena glass tube, one end of which was c10sed and contained water, tbe other end being drawn out to a jet. This was supported in a powerful tube-furnace in a sloping position, and the part containing the thoria heated to the highest possible temperature, while a sIow current of steam from the water at the end was passed over it, escaping by the jet. When all the water was evaporated, the jet was drawn off and the tube allowed to cool in an atmosphere of steam free from air. The thoria, on testing, was found to have been lowered in emanating power to about 7 per cent. The further heating had thus reduced the emanating power without the steam having at all regenerated it. In the next experiment, the reverse was tried. Two exact1y parallel processes were carried out for ordinary thoria possessing the normal amount of

The Cause and Nature

0/ Radioactivity. II

501

emanating power. In the first, it was heated in a porcelain tube in the tubefurnace for three hours, while about 500 c.c. of water were distilled over it from a retort. In the second, another quantity of thoria was heated in exaetly the same way for the same time, only a eurrent of well-dried air was substituted for the steam. The result was conc1usive: each sampie had had its emanating power reduced to exactly the same amount, that is, about 50 per cent of the original. These experiments prove that water-vapour exerts no influence either in de-emanating thoria or in effecting a recovery of its lost emanating power. The Regeneration 0/ the Emanating Power by Chemical Processes. The task of subjecting de-emanated thoria to aseries of chemical changes to see if it would recover its lost emanating power was then undertaken. It may first be mentioned that thoria which has been subjected to ignition has changed very materially in chemical and physical properties. The pure white colour changes at temperatures corresponding to the first stages of de-emanation to a light brown, and after subjection to the very highest temperature to a pure pink. At the same time the solubility of the substance in sulphurie acid is greatly diminished. Apart always obstinately refuses to dissolve, even after long and repeated boiling with the concentrated acid, although the part is diminished on each successive treatment, and appears to be in no way different from the rest of the substance. No differenee, however, oecurs in the readiness with which chlorine attacks it when intimately mixed with carbon. The formation of the chloride by this method is the easiest way of dissolving ignited thoria. Preliminary experiments went to show that emanating power is a quantity which varies, not only with the nature of the chemical compound but also for the same compound very materially with its previous history . Thus the oxide from the oxalate does not possess as a rule so great an emanating power as that used for comparison. The following two exactly parallel experiments were therefore made, the one with the ordinary, and the other with de-emanated thoria possessing 9 to 10 per cent of the emanating power of the first. Each was converted to chloride in the ordinary way, by mixing with sugar solution, carbonizing, and igniting the mixture of oxide and carbon so obtained in a current of dry chlorine. Each sampie was then treated with water, the thorium precipitated and dried at 110°. The result was conclusive, for each sampIe showed the same emanating power. For the first few days after preparation this value increased rapidly, but after having been kept a fortnight both specimens showed about 260 per cent of the emanating power of the thoria used as a comparison sampie. Thus the process of de-emanating thoria by ignition does not irretrievably destroy the emanating power, for after solution and reprecipitation no difference whatever exists in the emanating power between ordinary and de-emanated thoria. A fair conc1usion from these experiments is that the cause of the emanating power is not removed by ignition, but only rendered for the time being inoperative.

502

The Colleeted Papers

0/ Lord Ruther/ord

IV. Effect 0/ Conditions upon Emanating Power

The experiments just described brought out two new points. Thorium hydroxide possessed an emanating power which increased with time since preparation, and when it attained its maximum it was much greater than that of the oxide. Before any further work was undertaken, it was necessary to make a elose study of the influence of conditions upon the emanating power of thorium compounds. Effeet 0/ Temperature. The effect of increase of temperature on the emanating power of thoria has already been fully investigated by one of us (Phys. Zeit., ii, p. 429, 1901). The results, stated briefly, show that an increase in temperature up to a certain limit, in the neighbourhood of a red heat, correspondingly increases the emanating power. At the maximum this is between three and four times that at the ordinary temperature, and is maintained at this increased value for several hours without any sign of diminution with time. When the thoria is allowed to cool, the emanating power then returns to the neighbourhood of the normal value. If, however, the limit of temperature given is exceeded, de-emanation sets in, and even while the high temperature is maintained, the emanating power falls rapidly to a fraction of its former value. On cooling, the substance is found to be more or less de-emanated. It is of interest that no increase of emanating power is observed when de-emanation commences. These experiments were extended to inelude the effects of cooling. The platinum tube which contained the thoria was surrounded with a feit jacket containing a mixture of solid carbon dioxide and ether. The emanating power immediately fell to 10 per cent of its former value. On removing the cooling agent it again rose quickly to nearly the normal. In another experiment some thoria was surrounded in a platinum crucible with a mixture of solid carbon dioxide and ether, and kept in a vacuum for several hours. On removing it, and allowing its temperature to rise, it possessed much the same value as an ordinary sampie, and after standing some time in the air it was again tested, and no difference could be detected between the two. Thus changes in temperature produce very marked simultaneous changes in emanating power, but between the limits of -1100 and an incipient red heat no permanent alteration in the value occurs. Effeet 0/ Moisture. Dorn, loe. eit., had noticed that moisture produced a moderate increase in the power of thoria of giving an emanation, and of exciting radioactivity on surrounding surfaces. We have confirmed and extended his resuIts by the following experiments. Two similar sampies of thoria left sealed up for a week, the one in a desiccated atmosphere, the other in air saturated with water-vapour, showed an increase and decrease in emanating power respectively. The moist sampie possessed nearly twice as much emanating power as the dry. More complete desiccation, by sealing up the specimens in vacuo with phosphorus pentoxide

The Cause and Nature of Radioactirity. II

503

for a month, did not further reduce the emanating power. Some thoria mixed with concentrated sulphuric acid gave about one half of the usual emanation when vigorously shaken. These experiments show that the presence of water, although producing a marked increase, is not essential for the production of the phenomenon. Other experiments were made on the effect of light and air on emanaging power. The most useful result obtained is that thoria does not change in emanating power when kept in closed vessels under different conditions, but when exposed to the air the emanating power varies within comparatively narrow limits. Thorium Hydroxide. This compound, like the oxide, has its emanating power increased by water-vapour. A similar experiment to that described for the oxide gave as the result an emanating power of 400 per cent of that of thoria for the moist sampie and 300 per cent for the dry. Exposure to the air for a short time again equalized the two values. Carbon dioxide, which thorium hydroxide absorbs from the air to the extent of 2 per cent of its weight, is without influence on the emanating power. Effects 0/ M olecular Condition and State 0/ Aggregation 0/ Thorium on the Emanating Power. Unlike the radioactivity, the emanating power of thorium

compounds is by no means mainly controlled by the proportion of thorium present. The effect of temperature in de-emanating thoria and the high value of the emanating power of thorium hydroxide illustrate this. Thorium sulphate, oxalate, and nitrate possess but low emanating power, while thorium carbonate has been obtained with a value five times as great as that of thoria. In general, a dense crystalline compound in not very fine powder possesses a much higher emanating power than a light floury compound in a much finer state of division. Solution, however, has been found generally to greatly increase the emanating power of soluble thorium salts. In a eareful determination, using 20 gr. of finely-powdered thorium nitrate, this worked out to be only 1· 8 per eent of the emanating power of thoria. Dissolved in water, however, and tested for emanation by bubbling a eurrent of air through the solution it gives about three times as much emanation as thorium oxide. That is, solution in water increases the emanating power of thorium nitrate nearly 200 times. The emanating power, as in the case of solids, is proportional to the weight of substance present, and within the limits tried is not mueh affeeted by dilution, for a solution of 10 gr. made up to 25 c.e. in volume possessed a similar value when diluted four times. V. The Cause

0/ the Emanating Power o/Thorium

The separation from thorium of ThX, detailed in the first part of this communication, showed that not only the radioactivity but also the emanating power of thorium is conneeted with the presence of a non-thorium type of matter, ThX. The solutions from which thorium hydroxide had been

The Collected Papers 0/ Lord Rutherford

504

precipitated by ammonia possessed, when concentrated, about as much emanating power as the solutions from which they were prepared, while the precipitated hydroxide was more or less completely de-emanated. On aUowing these preparations to stand, the emanating power of the filtrates gradually disappeared, while that of the hydroxide in most cases rose steadily with time, ti11 at the end of a fortnight they had attained a maximum between three and four times that of ordinary thoria. This recovery of the emanating power in the case of the hydroxide was noticed long before the similar change of its radioactivity was observed, but the two phenomena admit of a similar explanation. If, in the precipitation by ammonia, care is taken to remove the ThX complete1y, the thorium hydroxide is at first almost devoid of

40

40

40

40

40

o

TIME INDAYS

2

4

6

8

'0

/2

Fig.2 emanating power. The small fraction that remains-only a few per cent of the maximum-can be accounted for by the reproduction of ThX during the time taken to dry the precipitate. The Rate of Recovery and Decay 0/ Emanating Power. The rate of decay of the emanating power of ThX, and the recovery of this property by the thorium from which it had been separated, were then investigated in parallel with the similar experiments on radioactivity already described. One quarter of the concentrated filtrate used for the latter purpose was taken, and the decrease of its emanating power with time measured. The increase of emanating power of the thorium hydroxide from which it had been prepared was also measured. Fig. 2 expresses the results. The decay-curve is merely approximate, for it is not easy to accurately take the emanating

The Cause and Nature of Radioactivity. II

505

power of a liquid without special arrangements to assure the constancy of the air current and the shaking of the solution. The experiments, although only of a preliminary character, bear out the conc1usion that emanating power decays and recovers according to the same law and at the same rate as the radioactivity of ThX, and that it is therefore one of the properties of the latter and not of thorium. The decay curve given, so far as it can be relied upon, shows that the emanating power of ThX at any instant is proportional to its radioactivity. VI. The Chemical Nature

0/ the Emanation

The following work has reference to the emanation itself, and not to the material producing it, and was designed to see whether the emanation possesses chemical properties which would identify it with any known kind of matter. It had been noticed at the time of its discovery that it passed unchanged through concentrated sulphuric acid. The same holds true of every reagent that has been investigated. The effect of temperature was first tried. The air containing the emanation, obtained in the usual way by passage over thoria, was led through the platinum tube heated electrically to the highest attainable temperature, and also through the tube cooled by solid carbon dioxide and ether. The tube was then filled with platinum black, and the emanation passed through it in the cold, and with gradually inereasing temperatures, until the limit was reached. The effeet of the intense heat was to eonvert the platinum black completely into platinum sponge. In another experiment the emanation was passed through a layer of red-hot lead chromate in a glass tube. The eurrent of air was replaeed by a eurrent of hydrogen, and the emanation sent through redhot magnesium powder and red-hot palladium blaek, and, by using a eurrent of carbon dioxide, through red-hot zinc dust. In every ease the emanation passed without sensible change in the amount. If anything, a slight inerease occurred, owing to the time taken for the gas current to pass through the tubes when hot being slightly less than when cold, the decay en route being consequently less. It will be noticed that the only known gases eapable of passing in unchanged amount through all the reagents employed are the recently discovered members of the argon family. But another interpretation may be put upon the results. If the emanation were the manifestation of excited radioactivity on the surrounding atmosphere, then since from the nature of the experiments it was necessary to employ in each case, as the atmosphere, a gas not aeted on by the reagent employed, the result obtained might be explained. Red-hot magnesium would not retain an emanation consisting of radioactive hydrogen, or redhot zinc dust an emanation eonsisting of radioaetive earbon dioxide. The correctness of this explanation was tested in the following way. Carbon dioxide was passed over thoria, then through aT-tube, where a eurrent of air met and mixed with it, both passing on to the testing eylinder. But between

506

The Collected Papers of Lord Rutherford

this and the T-tube a large soda-lime tube was introduced, and the current of gas thus freed from its admixed carbon dioxide before being tested in the cylinder for emanation. The amount of emanation found was quite unchanged, whether carbon dioxide was sent over thoria in the manner described, or whether an equally rapid current of air was substituted for it, keeping the other arrangement as before. The theory that the emanation is an effect of the excited activity on the surrounding medium is thus excluded. It is apriori improbable on account of the very different rates of decay of the activity in the two cases. The interpretation of the above experiments must therefore be that the emanation is a chemically inert gas analogous in nature to the members of the argon family. In light of these results, and the view that has already been put forward of the nature of radioactivity, the speculation naturally arises whether the presence of helium in minerals and its invariable association with uranium and thorium may not be connected with their radioactivity. VII. The Nature 0/ Emanating Power The foregoing resu1ts therefore find their simplest explanation on the view that, just as a chemical change is proceeding in thorium whereby a nonthorium material is produced, so the latter undergoes a further reaction, giving rise to a gaseous product which in the radioactive state constitutes the emanation. It will be seen at once that this secondary change is of a different kind from the primary, for it is affected apparently by the conditions in a very marked manner. It was shown that moisture, the state of aggregation, and temperature infiuenced the value of the emanating power. From -80 to a red heat the latter regularly increases in the ratio of 1 : 40 in the case of thorium oxide, while the ratios between the values for thorium nitrate in the solid state and in solution is as 1 : 200. The secondary reaction appears therefore at first sight much more nearly allied to ordinary chemical reaction than the primary. It must not be forgotten, however, that the laws controlling the manifestation of the two phenomena-radioactivity and emanating power-are of necessity very different. In the former we deal with the intensity of radiations emitted by asolid; in the latter with the rate of escape of agas into the surrounding air from either asolid or a liquid. Since this gas is detected by its radioactivity, and this decays extremely rapidly with time, a very slight delay in the rate of its escape will enormously affect the experimental value obtained for emanating power. On the other hand, it is now well established by experiment that sometimes thorium compounds de-emanated chemically by removal of ThX do not recover their normal emanating power with time, but remain constant at a lower value. On one occasion a carbonate was prepared which possessed hardly any emanating power until it was again dissolved and precipitated. In another experiment two sampIes of hydroxide prepared from different nitrates were tested together for rise of emanating power. That of the one 0

The Cause and Nature

0/ RadioactMty. 1I

507

rose normally to its maximum (as in Fig. 2), which was twenty times the minimum. The other started from the same minimum, but rose to a maximum only one fourth as great. When the experiment was repeated under the same conditions, using the same sampie of nitrate, the compound behaved normally. It thus appears that the emanation can be almost entirely prevented from escaping in the radioactive state in some cases, and partially prevented in others, where no visible peculiarity of physical condition exists, and where other preparations similarly prepared behave normally. These are outstanding points in the theory which remain to be explained. It is not possible at present to decide whether these variations of emanating power are caused by an alteration in the velocity of the reaction which produces the emanation, or by an alteration in the time taken for the latter to escape. The experiments detailed in the first paper on the augmentation of the proportion of excited activity in compounds de-emanated by ignition appear to favour the view that the change still proceeds, but the emanation does not succeed in escaping. The experiment on the regular variation of emanating power with temperature might be explained quite well by either hypothesis. VIII. The Excited Radioactivity [rom Thorium Since the emanation gives rise to the phenomenon of excited radioactivity, and the latter appears to be caused by an intensely active invisible deposit of matter, it must be supposed that a tertiary change is taking place. The emanation, a gaseous product of the secondary reaction, is again changing and giving rise to a third reaction product causing the excited activity. The fact that it is manifested entirely on the negative electrode in an electric field, points to the positive ion being the means by which it is transported. Without, in the present paper, going further into the consideration of excited radioactivity, it may be mentioned that the successive changes occurring in the thorium atom are not yet ended at this stage. The fact that the excited radiation consists in part of cathode rays may be recalled here. Further, the intensity of excited activity at first increases from the time of its formation, exact1y as in the case of ThX newly separated from thorium, the increase reversing the effect of the normal decay. The radium excited activity behaves in a somewhat analogous manner. The matter in this case causing excited activity does not appear to be homogeneous, but behaves in its action towards acids, etc., as if consisting oftwo different kinds (compare Rutherford, Phys. Zeit., p. 254, 1902).

IX. Further Theoretical Considerations Enough has been brought forward to make it c1ear that in the radioactivity of thorium, and, by analogy, of radium, we are witnessing the effect of a most complex series of changes, each of which is accompanied by the continuous production of a special kind of active matter. The complexity of the phenomenon gives rise to an important quest ion concerning the fundamental

508

The Collected Papers 0/ Lord Rutherford

relation between the changes which occur and radioactivity. So far it has been assumed, as the simplest explanation, that the radioactivity is preceded by chemical change, the products of the latter possessing a certain amount of available energy dissipated in the course of time. A slightly different view is at least open to consideration, and is in some ways preferable. Radioactivity may be an accompaniment of the change, the amount of the former at any instant being proportional to the amount of the latter. On this view the non-separable radioactivities of thorium and uranium would be caused by the primary change in which ThX and UrX are produced. The activity of ThX would be caused by the secondary change producing the emanation, the activity of the emanation by a tertiary change in which the matter causing the excited activity is produced, the activity of the latter being derived from still further changes. The law of the decay of the activity with time (equation I, first part) in all cases but the primary then appears as the expression of the simple law of chemical change, in which one substance only alters at a rate proportional to the amount remaining. In the primary change the amount remaining is infinitely great compared with the amount that alters in short time, and therefore the velocity of reaction is constant. This view certainly affords an explanation of why the emanating power of ThX is proportional to the radioactivity. So long as the latter is considered a consequence of what has occurred there is no reason why this should be so. But if it is considered the accompaniment of the change in which the emanation is formed the result follows naturally. Further and more exact determinations of the rate of rise and decay of emanating power are therefore called for. In the case of uranium the changes so far as they can be followed by the radioactivity appear to be at an end with that which causes the activity of UrX. It is of interest that this substance gives only cathode rays, and that it continues to do so for many weeks after its separation from uranium. This gives rise to the question whether any connexion can be established between the nature of the radiation and the kind of change producing it. The only consideration which is opposed to this view is the existence of polonium. The radiations of this body resemble c10sely the non-separable radioactivity of uranium, both in penetrating power and the absence of deviable rays. But all attempts (Soddy, loc. cit.) have so far failed to separate polonium from uranium, and until this is done its existence does not of itself affect the present question. It seems as if a more satisfactory explanation of the residual activities common to both uranium and thorium, and of the connexion between the emanating power and radioactivity of ThX, is obtained on the modified view. But further work, both on this latter point and on the nature of polonium, must be awaited before the connexion between radioactivity and chemica1 change can be considered exactly determined. Macdonald Physics Building, Macdonald Chemistry and Mining BuiIding, McGill University. Montreal

Excited Radioactivity and Ionization of the Atmosphere by

Maedonald Professor of Physies, Demonstrator in Physies, MeGill University, Montreal

E. R UTHERFORD, M.A., D.se.,

and s.

J. ALLEN, M.SC.,

From the Philosophical Magazine for December 1902, sero 6, iv, pp. 704-723 Communicated to the American Physical Society, December 27, 1901; Abstract published in the Phys. Zeit., 11 1902

THE experiments of Elster and Geitel· and C. T. R. Wilsont have shown that a weIl insulated charged conductor placed inside a closed vessel gradually loses its charge, and that this loss of charge is due to a small spontaneous ionization of the volume of air inside the closed vessel. Wilson calculated from his results that about 19 ions per c.c. are produced in the air per second. In a later paper~ Wilson has shown that the ionization in different gases varies approximately as the density and pressure of the gas. These results point to the possibility that the ionization observed in gases may be due, in part at least, to the emission of an ionizing radiation from the walls of the containing vessel. Recently, Elster and Geitel§ made the interesting discovery that a negatively charged conductor, placed in the open air, becomes temporarily radioactive. This radioactivity decays in the course of a few hours. The phenomena appear to be closely analogous to the 'excited' radioactivity produced by the radioactive emanations of thorium and radium. The excited activity from the air can be concentrated on the negative electrode in exactly the same way as one of the authorsll has shown for thorium-excited activity. In addition Elster and Geitel have shown that the substance responsible for the radioactivity ean be removed by solution in acid. On evaporating the solution to dryness an active residue, whieh decays with time, is left behind in the vessel. This also is in striking agreement with what one of us, loe. eil., had previously shown for thorium-excited aetivity. In the experiments of Elster and GeiteI, and Wilson, the amount of ionization of air has been determined by observing the rate at which the leaves of a charged electroscope of special construction fall together. This

* Phys. Zeit., November 24,

1900.

Ibid., December 1901. § Phys. Zeit., iii, p. 76 (1901); xl, p. 590 (1901).

t Proc.

~

11

Roy. Soc., March 1901.

Phil. Mag., February 1900.

510

The Collected Papers

0/ Lord Ruther/ord

method, while very simple and advantageous for some experiments, is, in general, slow, and in many cases does not allow of sufficient variation of experimental conditions. In the present investigation the authors have utilized a sensitive quadrant electrometer for examination both of the ionization and excited radioactivity produced in air. The electrometer employed is a modification of the Oolezalek electrometer which is described in Instrumenten Kunde, Oecember 1901. It is of the ordinary quadrant type, with a very light needle of silver paper suspended by a fine quartz fibre. The apparatus, as constructed by Herr Barteis, of Göttingen, was for determination of small P .Os. for electrochemical work. For our purpose it was necessary to completely alter the insulation and method of connection of the quadrants. In the present experiments the needle was charged at intervals of two days by lightly touching the needle by a fine wire connected to a battery of 200 volts. It was found that the needle did not lose more than 10 per cent of its charge in 24 hours. The damping of the needle, on account of its lightness, was fairly rapid, and no extra damping vane was required. The de:tlection was observed by a telescope and scale at a distance of 2 m. The zero point was found to be very steady. For the first suspension employed the electrometer gave a de:tlection of about 1,800 mm. of scale corresponding to one volt P.o. between the quadrants, when the needle was charged to 200 volts. This suspension was accidentally broken in the course of the experiments and was replaced by a quartz fibre which gave only about ! of this de:tlection for the same voltage. When dealing with the very small rate of discharge which is produced by the spontaneous ionization of air, it is very essential that every precaution should be taken to guard against external electrostatic disturbances. The electrometer and all the connecting wires were enclosed in gauze cylinders connected to earth. The :tloor and woodwork in the immediate neighbourhood of the testing apparatus were covered with metal connected to earth. The separation of the quadrants was done by means of a special mercury key operated from a distance by a cord. The insulating substances necessary in experimental arrangements were completely diselectrified by means of flames. Production

0/ Excited Radioactivity

The simplest method of obtaining a large amount of excited radioactivity from the air is to expose a long insulated wire charged to a high negative potential in the open air. After exposure for several hours the wire is removed and wound on a frame, or in the form of a :tlat helix. The ionization produced by the radioactive wire in the testing vessel is then observed, by means of the electrometer, in the usual way. Elster and Geitel in their experiments have used an electroscope to measure the ionization produced. In order to produce a considerable quantity of activity on the conductor

L.\"cited Radioactil'ity amllvni=atiol1 of fhe Afmosphere

511

it is neccssary to charge the wirc to a high negative potential. Potentials varying [rom -5,000 to -100,000 volts have been used in the experiments. A positively charged wire remains quite inactive however long it may be exposed.

Decay

0/ Excited Radioactivity

The excited radiation [rom the air decays with the time in a manner similar to the excited radiation from thorium and radium, but at a different rate. The excited radiation from thorium falls to half value in about 11 hours, while the excited activity from air falls to half value in about 45 minutes for the range of voltages examined. It has been shown* that the excited radiation from radium decays in an irregular manner, the rate of decay depending on the time of exposure. The rate of decay is rapid at first, then nearly stationary for some time, and then a regular decay to zero, falling to half value in about 30 minutes. It is thus seen that the rate of decay of excited activity, due to the atmosphere, is more nearly allied to that from radium than to that from thorium. In the experiments detailed below the excited activity was produced on a long insulated wire, 15 m. long, suspended outside the laboratory window about 15 ft. from the ground. The wire was kept charged by means of a Wimshurst machine driven by a motor. The potential of the wire was measured by means of the sparking distance between two brass knobs. In order to regulate the potential of the exposed wire to any desired value a needle point connected to earth was placed near a small plate connected to the charged wire. The distance between the point and the plate was adjusted until the sparkjust refused to pass between the knobs. This method was found to be more satisfactory than varying the speed of the machine. After the wire had been exposed a definite time, it was rapidly removed and wound on a rectangular metal frame 120 cm. 10ng and 10 cm. wide. The method of winding is shown in Fig. 1, where the frame A is seen in position inside the testing cylinder B. The testing cylinder was of metal, about 150 cm. long and 30 cm. diameter. The outside cylinder was connected to a battery of 100 volts. In order to ensure that there was no conduction current between Band A over the supports the insulator CD was cut into two parts and separated by a metal ring connected to earth. The arrangement can be c1early seen from the figure. The radiation from the wire ionized the gas inside the testing cylinder and the current between the electrodes was observed with the sensitive electrometer in the usual way. On account of the weak ionization of the air by the radiation a P.D. of 100 volts was sufficient to remove all the ions to the electrodes before appreciable recombination, and to give the maximum current through the gas.

* Rutherford, Phys. Zeit., xii, p. 254 (1902); and Rutherford and Miss Brooks, Phil. Mag. July 1902.

512

The Collected Papers of Lord Rutherford

Fig. 2 (I) shows the decay curve for a copper wire exposed 210 min. inside the laboratory at a P.D. of -26,000 volts; Fig. 2 (11) the curve for the same wire exposed 270 min. in the open air at -24,000 volts. The

c

o

C'

Earth

A

B

Eartn

Fig.l ordinates represent divisions per second of the electrometer and the abscissae time in minutes. In most of the experiments (especially when the wire was exposed for several hours) it was found necessary to use a condenser in parallel with the electrometer to decrease the rate of movement of the needle.

I. \ ci/cd Radio(/('/ il'i/)' will !o/li;{aio/l of Ihe A Il1Iv.\phere

513

Therc was always a current (about 2· 5 divisions per second) in the testing vessel when the wire was inactive, due to the spontaneous ionization of the air in the cylinder. Allowing for this it will be seen that the current (which is proportional to the intensity of the radiation) falls off in a geometrical progression with the time, falling to half value for both cases in about 45min. Fig. 3 is a decay-curve for a lead wire exposed in an attic 190 min. at -25,000 volts. In this case the lead wire was wound in the form of a "10. 36

32 Z8

DECAY CORVIS Ff)1( J!«C"GD RAblDACTlV'"-

24

20

~. ~.

16 12

~

't

;S

B 4

T1MEINMl VI.

o

I

2D

ofO

60

80

100

'20

140

'SO

,eo

200

Fig.2 Hat spiral and placed inside a testing vessel consisting of two parallel plates, one of which was connected to the electrometer and the other to the battery. This again falls to half value in about 45 min. A large number of curves of decay have been determined under very varying atmospheric conditions, but no certain differences in the rate of decay have been observed, although the amount of excited activity in a given time varies greatly with the weather and amount of wind. The rate of decay was the same for a copper as for a lead wire, and was independent of the diameter of the wire. The rate of decay for a brass rod charged at -100,000 volts was about the same as for a lead or copper wire exposed at -5,000 volts. The rate of decay for low voltages has not been investigated. We may thus conc1ude that over the range examined the rate of decay is regular and independent of conditions. In this respect also it resembles the excited radiations produced by thorium and radium. R

514

The Collected Papers o[ Lord Ruther[ord

It will be seen, from these results on the rate of deeay, that if the intensity of the excited radiation is initially 1o• the intensity I after a time t is given by 1= loe-At •

Sinee 1 =

~ 10 when t = 45 min. ,\ = O· 00026 .

If the exeited aetivity produced on the wire is due to a uniform rate of deposit of radioaetive material the radiation from whieh deeays with the

so 70 50

DlCAY OF ilC/TED

so

RADIOACTIVITY

40

30 20

llS

11

~

'0 71ME 111 MINS.

o

20

40

60

10

'00

l2D

MO

Fig.3 time according to the above equation, it necessarily follows· that the intensity 1 after a time of exposure t is given by I

= 10 (1

- e-At ),

where 10 is the maximum value of the intensity reaehed after a very long exposure. If this result is eorreet the amount of excited aetivity in a given wire for a fixed voltage should reach half its final value in 4S min. Some experiments have been made on this point with wires exposed in the open air for different times. The amount of exeited radioactivity in the air was found, however, to be too variable to test the truth of the equation. The results, however, showed that the amount of aetivity inereased at first • E. Rutherford, Phil. Mag., February 1900, p. 180.

F:xcited Radioarlivity and lonizalion o/Ihe Atmosl'here

5] 5

roughly in proportion to the time, but after three or four hours' exposure reached a practical maximum. More ace urate experiments on this point are at present in progress, using a closed room instead of the open air, when the amount of excited activity is much more constant. The amount of excited radioactivity from the air increased with the voltage of the exposed wire. On account of the variation of the amount of excited radioactivity in the air from hour to hour and day to day, no definite results on the variation of the amount of excited radioactivity with the voltage were obtained.

Effect 0/ Weather Conditions A large number of experiments were made on the effect of atmospheric conditions on the amount of excited radioactivity from the air. The wire was usually exposed for 30 min. at a voltage of -25,000 volts, outside a laboratory window, at a height of about 15 ft. from the ground. The results showed that the amount of excited activity produced from the air varied greatly with the atmospheric conditions. Other conditions being the same, a bright clear day gave more excited activity than a dull cloudy day. The effect of temperature was not very marked. If anything, slightly more activity was obtained on a bright day during the Canadian winter, with a temperature of about -20°C, than on a bright warm day in the spring. The most powerful factor in determining the amount of activity given to the wire is the presence or absence of wind. A windy day always gave much greater effects than a quiet day, when other conditions were the same. This is true whether the air was cold or warm, or the day bright or dul1. Most of the experiments were made during the Canadian winter, when there was about two feet of snow over the ground. The prevailing wind was from the north, and had been carried over snow-covered lands. The fact that the amount of activity was uninfluenced by the presence of snow shows that the excited activity is not likely to be due to any effect arising from vegetation. The amount of water-vapour in the air appears to have Httle influence on the result, for at -20°C the air is extremely dry.

Penetrating Power 0/ Excited Radiation It has been shown in a previous paper· that the penetrating power of the

excited radiations of thorium and radium was the same. As the penetrating power is one of the methods of distinguishing between the various radiations, a special experiment was made to compare the penetrating power of the excited radiation from the air with that of other known radiations from radioactive substances. Lead wire was employed in these experiments as it could readily be retained in the form of a flat helix. The wire was excited by exposure of 2 to 3

* E.

Rutherford and Miss H. T. Brooks, Phil. Mag., July 1902.

711(.' Collectcd Papers

516

0/ Lord Rutllc/lord

hours at -30,000 volts. It was then wound to form a flat helix and placed between a parallel plate apparatus. The ionization current between the plates was observed for different numbers of sheets of thin aluminium foil placed over the helix. The average thickness of the aluminium foH was 0·00034 cm. The results are shown in curve I, Fig. 4 where the penetrating powers of other known types of radiation are added for comparison. The excited radiation due to air has greater penetrating power than any 100

80 'A8rQ) 'PrION cu}, VES

80

...~

~ ~

40 ~ ~

...

~

~

20 -LArEN! OF Al UM. FOIL-

o

2

2

3

4

5

6

Fig.4 of the types of radiations, not deviated by a magnetic field, from the radioactive substances uranium, thorium, and radium, and is also more penetrating than the excited radiation produced by radium and thorium. No special experiment has been made to determine the absorption of the excited radiation in its passage through the air, but its approximate amount can be readily deduced from known data. In all the different types of radiations examined it has been generally found that if one radiation is more easily absorbed than another, in aluminium for example, it is also more easily absorbed in air. Since the excited radiation from the air is slightly less absorbed in aluminium than that due to thorium, we can thus conclude that it is slightly less absorbed in air. Now it is known that the intensity of the excited radiation from thorium falls to half value after passing through 1· 6 cm. of air. It thus follows that the intensity of the excited radiation from air falls to half value after passing through about 2 cm. of air, and is almost completely absorbed in a distance of 10 or 12 cm.

L.rcitecl Radioaetil'iry ami IOlli=atiofl o! tlle Atmospllere

517

From thc differences observed for the penetrating power and ratio of decay we can conc1ude that the excited radiation from air cannot be ascribed to the presence of any known radioactive substance in the atmosphere.

Transmission

0/ Excited Activity

We have seen that the excited radiation from the air is similar in all respects to the known types of excited activity by thorium and radium. In both cases the activity is confined to the cathode in an electric field, and can be partly removed by rubbing with a cloth or by solution in acid. The differences observed in the rate of decay and penetrating power of the radiations show that the effects obtained cannot be ascribed to the presence of aminute quantity of thorium or radium emanations in the atmosphere. The close resemblance in the phenomena, however, renders it probable that the excited activity from the air is due to a process similar in character to that which produced excited activity from the emanations of thorium and radium. One of the authors· has recently shown that in the case of radioactive substances the excited activity is due to a transmission of positively charged radioactive carriers to the cathode. These carriers travel in an electric field with about the velocity of the positive ions produced in air by Röntgen or Becquerel rays. There seems to be little doubt that the excited activity is due to a deposit of aminute quantity of intensely active radioactive matter. Such an hypothesis is essential to explain the facts of solutions, and that the radioactivity can be transferred from the radioactive body to the cloth by rubbing. The production of excited activity from the air cannot be ascribed to any surface action on the conductor due to the electric field. A wire does not give any appreciable activity if it is confined in a cylinder where the volume of air is small, although the wire is subjected to the same voltage as in the open air. All the evidence obtained up to the present points strongly to the conclusion that the excited activity is derived from the volume of the air surrounding the charged wire. Since the activity is confined to the cathode, the carriers to which the activity is due must possess a positive charge. These carriers may obtain a positive charge either by the condensation of temporary radioactive matter of some kind round the positive ion already existing in the air, or by the expulsion of a negati ve electron from the carrier. The latter explanation seems the more probable, for we now knowt that all the radioactive substances, thorium, radium, and uranium, as weIl as the excited activity due to thorium and radium, possess the property of spontaneously expelling electrons. There is as yet no definite evidence of the origin or mode of production of these radioactive carriers in the air, but assuming their presence, many of the experimental facts observed receive a simple explanation. • Phys. Zeit., x, p. 210 (1902).

t Rutherford and Grier, Phys. Zeit., xvii, p. 385 (1902).

518

The Collected Papers of Lord Rutherford

The higher the potential of the wire the greater the distance from which the carriers are conveyed to the cathode. The amount of excited activity on a wire exposed in free space, on this view, should increase rapidly with increase of voltage. There is strong evidence that a wire charged to a high potential attracts the carriers over a large volume of air. It was experimentally found that the amount of excited activity obtained from a wire charged to -20,000 volts in a cylindrica1 vessel of volume 141,000 C.C., when outside air was drawn through it at a rate of 500 cm. per second, was only a few per cent of the amount obtained from the same wire in the open air. The increase of excited activity observed on days on which a strong wind is blowing is, on this view, due to the continued supply of fresh carriers which are brought in the sphere of action of the electric field. Since the exposed wire merely acts as a collector of radioactive carriers under the infiuence of the electric field the amount and nature of the excited radiation should be independent of the nature of the conductor, and tbis is found to be the case. It thus appears probable that radioactive carriers are continually produced from some constituent of the atmosphere, but at a rate depending on atmospheric conditions. Bright clear weather appears to be the most favourable condition. Since the earth is nearly always charged negatively with regard to the upper atmosphere, it follows that these radioactive carriers are being continually deposited over the surface of the earth. We must thus regard the earth as covered with an invisible layer of intense radioactive material which ionizes the air strongly within a few centimetres of the surface. The presence of these carriers in the volume of the air will also cause the production of fresh ions throughout the atmosphere, for each carrier acts as a radiating centre. A hill or mountain peak, or any high mass of rock or land, concentrates the earth's electric field upon itself and consequently it will receive more radioactivity per unit area than the level plain. Elster and Geitel have pointed out that the greater ionization observed in the neighbourhood of projecting peaks, receives a satisfactory explanation on this view. Spontaneous Ionization

0/ the Air

The experimental arrangement shown in Fig. 5 was employed for determining the number of ions produced per cubic centimetre per second in air and the variation of the ionization current with the strength of the electric field. The current was observed by means of the electrometer between two concentric zinc cylinders A and B, 154 cm. in length, 25· 5 and 7·5 cm. in internal diameter. The cylinders were placed vertical and the base of both cylinders closed. The large cylinder was closed at the top by a zinc plate, in the centre of which was a circular opening slightly larger than the internal cylinder. A metal fiange, fixed round the top of the inner cylinder, rested

Fxdted Radioartil'ify amI

!oni=atioll of ,!Je A tmospl!e1'r

519

on an ebonite ring C. Between the ebonite and the zinc plate D was placed a ring of thin metal E, connected to earth, which res ted on a similar ring of a partial insulator like cardboard. The thin metal ring, connected to earth, served as a guard ring, and ensured that even with a large p.n. between the

B

B

B

EJrth

B

j..-A

&rtiI

Fig.5 cylinders, no current could leak across the insulator to the inner cylinder, which was connected to the electrometer in the usual way. The outer cylinder was connected to one pole of a battery of small storage cells, the other pole of which was earthed. The electrometer needle showed quite a rapid movement due to the

520

Tlte Collected Papers

0/ Lord Ruthelford

ionization current between the electrodes with a P.D. of a few volts between the cylinders. The cylinder was made fairly airtight and allowed to stand undisturbed. Observations of the current between the cylinders were made at intervals for over a month. In order to avoid correction for the slight variations in sensitiveness of the electrometer from day to day, the ionization current between two paralleled insulated plates due to a standard sampie of uranium oxide was observed at the same time. Previous experiments have shown that the uranium oxide is a very constant source of radiation. The following tables show the variation of the current, due to the spontaneous ionization of the air, with the P.D. between the cylinders. Table I is for air which has stood undisturbed for a month inside the vessel; Table 11 for the ordinary air of the room several hours after it had been introduced into the apparatus. TABLE

P.D. in volts

0·4 0·8 2·1 4·2 6·5 13

26 52

I

TABLE

Current in divisions per second of electrometer

0·34 0·50 0·59 0·65 0·67 0·71 0·72 0·73

P.D. in volts

0·2

1·05 2·1 6·5

13

26

39

52

II

Current in divisions per second of electrometer

0·04

0·22 0·32

0·52

0·61 0·65 0·67 0·68

The results are expressed graphically in Fig. 6, curves land 11 respectively. The curves are very similar to those observed when the air is ionized by Becquerel or Röntgen rays. The current first increases approximately direct1y as the voltage, but soon reaches a stage in which large variations of the voltage only cause a slight increase in the current. On account of the very small amount of ionization of the air and consequent slow rate of recombination of the ions, the maximum current is reached for a very small voltage. The current for 50 volts is not very different for the two sampies of air, but in curve I the current reaches an approximate maximum much earlier than in curve 11. This difference is probably due to the presence of dust partides in the air in the latter case. Some of the ions in their slow passage between the cylinders give up their charges to the dust nudeL This action causes an increase in the rate of combination of the ions and consequently a larger electric field is required to produce the maximum current. The capacity of the electrometer, cylinder, and connections, was 150 E.S. units when 1 mrn. division of electrometer corresponded to 0·00182 volt.

Lxdled

Radiuaetivity ami JOllizafion

0/ fhe

Atl1losphere

521

The average valuc of the movement of the electrometer needle was 100 divisions in 132 seconds for 50 volts between the cylinders. The current between the cylinders was thus 6·9

X

10-4 E.S. units

or 2·3 X 10-13 amperes. The volume of air between the cylinders was 71,200 c.c. Taking the value of 6· 5 x 10-10 E.S. units, found by J. J. Thomson* as the charge on an ion, the number of ions produced per cubic centimetre per second is 15.

70

60 -SATURATION CURVFS-

so

IIF

CYl.INDEIt 40 30

20

~

~

JO

! o

JO

~D.in /!Q11:s j etwo. 'rJ EA. 'rJ EA. _~

20

~O

40

so

Fig.6 This is not very different from the value of 19 found by Wilson for air inside a silvered glass vessel, using the electroscope methode No certain difference was observed in the current for aperiod of time extending over one month. The production of excited radioactivity from the air suggested the possibility that a radioactive emanation was present in the air and that this might cause the ionization observed. If this is so, the radiating power decays at an extreruely slow rate, or the emanation is being continuously reproduced in the enc10sed space. R*

* Phi!.

Mag., 1898.

522

The Collected Papers ofLord Rutherford Application o[ the Ionization Theory

In the spontaneous ionization of air we are dealing with an extremely slow rate of production of ions, and it is of interest to see how far the experimental results are in agreement with the ionization theory of gases, which has been previously tested in cases where the ionization is many thousands of times more intense than the present one. We have already noted that the variation of current with the voltage is in general agreement with the theory. Ir q is the constant rate of production of ions per second and no electric field is acting, the number n of ions per cubic centimetre increases until the rate of production is equal to the rate of recombination of the ions, or q = tX1'l2, where IX is the constant of recombination. Now we have shown that q = 15, and McClung* has found from the recombination of ions of Röntgenized air that IX

= 3,400 e about,

where e is the charge on an ion. Substituting these values, we find n = 2,600,

Le. when a steady state is reached, the number of ions per cubic centimetre is 173 times the number produced per second. The time t taken for this number of ions to diminish to half, supposing the rate of production stopped, is given by 1 t=-

IXN

= 174 sec. We can obtain a rough approximation of the agreement with theory of the current voltage curve shown in curve I (Fig. 6) from the following considerations. The electric field X, at any point distant r from the centre of two concentric cylinders of radii band a, is given by X=

V

b=

r10~-

0·82 V r

a

substituting values of b and a of cylinders in Fig. 5. Now if N is the maximum number of ions per cubic centimetre, the current i per unit length of cylinder over any cross-section, when a small p.n. V is applied, is given by i = 2Trr.N.e.u.X,

where u = sum of velocities of positive and negative ions in unit field. • Phil. Mag., March 1902.

Excifl'd Radioactil'i/.V aml Ioni=ation l?f fhe Atmosphere Sllb~tituting

523

the value of X i = 1·641TNeuV.

If I is the maximum eurrent when all the ions produeed reaeh the eleetrodes I = qe1T(b 2

and

i j

a2),

-

1·64N.u. V =

q(b 2

-

a2)

•

Now for a P.D. of O' 36 volt the eurrent i is 0·4 of its maximum value (see Table I, p. 520). Now it will be shown later in the paper that the veloeity of the ions produeed in air is about the same as that of the ions produced by Röntgen rays. The value of u (the sum of the veloeities of the positive and negative ions) for a gradient of I volt per em. is thus a bout 3·2 em. per see. Substituting these values, we obtain N q

-=

32.

Taking into eonsideration that 0·4 of the ions are removed by the eurrent before reeombination, it follows that when no voltage is aeting N 32 = q 0·6

-

=

53 roughly.

Now we have shown that if IX has the same value as that obtained for intense ionizations N - should equal 174, q a value over three times as great. There are, however, several eauses at work which tend to make the observed values less than the theoretieal. In the first place, no correetion has been made for the disappearanee of ions by diffusion to the sides of the vessel. This ean be shown to be quite an important faetor in eausing a N

low value of -. q Curve II (Fig. 6) shows what an important influenee the presenee of dust has on the shape of the eurrent voltage eurves. In addition, it has been assumed, for simplicity of ealeulation, that the potential gradient is not disturbed by the movement of the ions. Experiment and theory have, however, shown that there is a sudden drop of potential near both eleetrodes and that the eleetrie field some distanee from them is less than if no ions were present. All of these three eauses aet in the same direetion and tend to give too low

524

The Collected Papers of Lord Rutherford

a value of N. The agreement between theory and experiment is thus as c10se q as could be expected under the experimental conditions. The results show c1early that, when air is kept in a closed vessel and no electric field is applied, the number of ions per unit volume, when equilibrium occurs between the rates of production and dissipation, is more tban 50 times the number produced per second per unit volume. Velocity of the Ions

Some experiments were made to obtain an approximate estimate of the velocity of the ions which are spontaneously produced in air and at tbe same time to determine the number of ions per unit volume present in tbe outside air.

E8rfJ1

e~rth

E8rfJ1

:A :A :A

Et1l'th

Fig.7 For this purpose, tbe apparatus sbown in Fig. 7 was employed. Air from the outside of the building was drawn through a zinc cylinder, length 200 cm., diameter 30 cm., by means of a fan driven by a motor. Tbe air in its passage through the tube passed througb three circular parallel wire gauzes, A, B, C, 2 cm. apart and insulated from eacb other. Tbe first gauze A was connected to earth, the second B to the electrometer, and the third C to one terminal of a battery of storage ceIls, tbe other terminal of which was to earth. A guard ring, connected to earth, was arranged between Band C to ensure tbere was no conduction leakage across the insulators between Band C.

t;seite" Radioactivity and Ionization

0/ the Atmosphere

525

Suppose gauze C is charged positive. The positive ions, carried with the current of air between the gauzes, start to travel up against the current of air, while the negative ions travel to the positive electrode with the current. Ir the velocity of the positive ions in the electric field is greater than the current of air, they will all reach the gauze Band for a given current of air the current observed with the electrometer will be unaltered when the strength of the electric field is increased. It was experimentally observed that even with a small electric field, there was some current to the gauze B. This amount increased with the voltage to a practical maximum. The experimental results are shown graphically in Fig. 8 for velocities of the current of air of 100, 205, and 250 cm. ISICMJ.!'IR SEC.

5

+ ~

~ 3

.... ~

IDO CMS. I'EIt SIC

~

~~("f./'/~/e.

or;:

lIi::

2

2

-P./J.6ttWtM Eleef,ror/,s i" Volts 2

200

100

300

+00

500

600

Fig.8 per sec. respectively, when the gauze C was charged positively. It will be seen from the curves that, for a velocity of 250 cm. per sec., the maximum current is reached with a p.n. of about 350 volts. Since the gauzes were 2 cm. apart, the velocity K of the positive ions for a potential gradient of I volt per centimetre is given by K=

2u

V

=

2 x 250 'lA::f\

= 1·4cm. per sec.,

where u is velocity of current of air and V the smallest p.n. for which the maximum current is reached. The other two curves also give a value of the

526

The Collected Papers 0/ Lord Rutherford

positive ion of about 1· 4 cm. per sec. It was not found possible to obtain more than an approximate result for the velocity of the ions, on account of the variation in the conductivity of the air drawn through in the course of aseries of observations. The curves shown in Fig. 8 were obtained on special days when the variation of the conductivity was small. Observations made in a similar way, to determine the velocity of the negative ion, were not very definite on account of variations during an experiment. The results showed that the velocity of the two ions was about the same, but it was not possible to decide whether the negative ions move slightly faster than the positive, as is the case for ions produced by Röntgen and Becquerel rays in air. The results obtained for the velocity of the ions are only approximate in character, but they point to the conc1usion that the ions produced spontaneously in the atmosphere travel at about the same rate in the electric field as the ions produced in air by Röntgen and Becquerel rays. In arecent determination Zeleny* has shown that the sum of the velocities of the positive and negative ions, produced by Röntgen rays in dry air, is about 3·2 cm. per sec.

Variation 0/ the Number 0/ Ions in the Air By noting the maximum current between the gauzes, an estimate can be

made of the number of ions per unit volume present in the air drawn through. If A is the area of the cross-section of the cylinder, u the mean velocity of the current of air, N the number of ions per unit volume, the maximum current i observed by the electrometer is given by

i = A.u.N.e, where e is the charge on an ion. Substituting the observed values of i, A, and u in this equation, the value of N can be deduced. The value of N was found to be variable both from hour to hour and day to day. The following numbers illustrate a few of the results obtained. Number of ions per unit volume

Date

November 20, November 21, November 23, November 27, November 30,

1901 1901 1901 1901 1901

40 30

14 16

13

The temperature of the air in most of these cases was about -12°C. A bright c1ear day was found to give a greater value of N than a dul1 day. • Phil. Trans. Roy. Soc., 1900.

Excited

Radioacti~'ity

ami Ion i=at ion

0/ the Atmosphere

527

A very similar apparatus has becn employed by H. Ebert* to determine the number of ions present in the air, only in his experiments the air was drawn between concentric cylinders, and an electroscope employed instead of an electrometer. We see from the above results that the number of ions per unit volume in the air varies considerably, but on three days was almost the same as the number produced per second in a c10sed vessel. This is a surprisingly small number if we consider the outside air to be ionized at the same rate as the air inside the closed vessel; for we have shown earlier in the paper, that in a c10sed space the number of ions per c.c. increases to 50 times the number produced per second before the rate of recombination is equal to the rate of production. After making due allowance for the causes tending to remove the ions, viz. the presence of dust and other particles in the outside air, and the electric field between the upper atmosphere and the earth, the number per unit volume is far lower than would be expected. It is possible that the spontaneous ionization of the air observed in c10sed vessels may be due (in part at least) to a radiation continuously emitted from the walls of the vessel. The spontaneous ionization of the outside air may, on this view, be much smaller than that observed in closed vessels, and the number of ions present per unit volume correspondingly less. McGill University. Montreal June 9th, 1902

• Phys. Zeit., No. 46, 1901.

Note on the condensation points of the thorium and radium emanations by

E. R UTHERFORD

and F. SODDY

From the Proceedings 01 the Chemical Society, 1902, pp. 219-20

IT was shown (Trans., 1902, 81, 342) that the radioactive emanation from

thorium compounds passed in unchanged amount through a tube cooled to -78° with solid carbon dioxide and ether. The acquisition of a liquid air plant has enabled the authors to repeat the experiment at a lower temperature, with the result that they have now succeeded in condensing both the thorium and the radium emanations. A current of hydrogen (or of air) was passed through the thorium or radium compound, and thence through a copper spiral cooled in liquid air. No trace of emanation escaped in the issuing gas either in the case of radium or of thorium. At the temperature of liquid air the emanations are therefore either condensed or lose their activity. The radium (or thorium) compound was then removed, and the gas current sent directly through the spiral tube, which was quickly taken out of the liquid air and placed in cotton wooI. Nothing happened for two or three minutes as the temperature of the spiral slowly rose, until suddenly the presence of the emanation was observed in large amount in the escaping gas. The result therefore cannot be explained by supposing that the emanations lose their ionising power at the temperature of liquid air. It is c1ear that they are condensed and again volatilised. From the sharpness and suddenness of the phenomena there appeared to exist for eacb emanation adefinite temperature of volatilisation. This temperature is very nearly the same in the two cases and considerably above the boiling point of liquid air. By measuring the temperature of the copper spiral at the instant of volatilisation, it was found that successive determinations agreed amongst themselves to within a degree. The exact temperatures cannot yet be given, but as a first approximation it may be said that for the radium emanation the volatilisation point lies in the neighbourhood of -130° and for the thorium emanation about five degrees lower. This experiment furnishes an additional proof-if such were needed-that radioactivity is accompanied by the continuous production of special kinds of active matter, which possess distinct and sharply defined chemical and physical properties.

Excited Radioactivity and the Method of its Transmission hy E. R UTHERFORD, M.A., D.se., Macdonald Professor of Physics, McGiIl University, Montreal

From the Philosophical Magazine for January 1903, ser. 6, v, pp. 95-117 (Preliminary accounts of these results were communicated to the American Physical Society, New York, December 27, 1901, and to Phys. Zeit., No. 10,1902)

CONTENTS

* 2. Connection between the excited activity and the emanations.

§ § § § § § § § § § §

I. Introduction.

3. 4. 5. 6. 7. 8. 9. 10. 1I. 12.

Method of transmission of excited activity. Velocity of carriers of thorium-excited activity. Increase of excited radiation with time. Radium-excited activity. Distribution of excited activity on the anode. Velocity of the carriers. Origin of the carriers. Nature of the radiations. Evidence of chemical changes. Summary of results.

§ 1. Introduction

of the most interesting properties of the radioactive substances thorium and radium, is their power of communicating or exciting* temporary radioactivity to all bodies in their neighbourhood. If a wire charged to a high negative potential is placed in a c10sed metal vessel containing a thorium or radium compound the excited radioactivity is confined almost entirely to the negative electrode. If the wire is charged positively, it remains inactive, but the excited radioactivity is produced on the walls of the vessel. When no electric field is acting, excited radioactivity is produced on the surfaces of aB bodies in the c10sed vessel, independently of their being good conductors or insulators. In previous papers the author has shown that there is a direct connection between the presence of the radioactive emanation from thorium and radium ONE

* (Note). The term' excited' has been used throughout these investigations rather than 'induced', which has found favour with many physicists. I have avoided using the latter term as, to my mind, it conveys the idea that the effect is in some way due to an action across the medium; while the experiments in this paper show conclusively that excited radioactivity is transmitted by means of a convection of positively charged carriers.

530

The Collected Papers of Lord Rutherford

and the production of excited radioactivity. It will be shown in this paper that the production of excited radioactivity is one of the properties of the emanation from thorium and radium. This excited radiation is caused by the deposit on the surface of bodies of radioactive matter, which is transmitted by positively charged carriers travelling through air in an electric field with about the same velocity as the positive ion, produced in air by Röntgen rays. § 2. Connection between Excited Radioactivity and Emanation

In a previous paper (Phi!. Mag., January 1900) I have shown that thorium compounds continuously give off a radioactive emanation. This emanation loses its radiating power rapidly, falling to half value in the course of one minute. Dorn showed later that radium also gave off an emanation, especially when heated. This emanation decayed much more slowly than that from thorium. In some experiments where the radium emanation, mixed with air, was kept in a c10sed metal vessel, I have found that the activity of the emanation fell to half value after standing several days, but was quite appreciable after a month's interval. These emanations from thorium and radium behave in all respects like radioactive gases or vapours. They diffuse rapidly through gases and through porous substances like paper, but unlike the gaseous ions which they produce in their path, pass through plugs of cotton-wooI, and bubble through solutions with no appreciable absorption. In a more detailed investigation (Rutherford and Soddy, Phi!. Mag., September 1902) it has been shown that the thorium emanation behaves like an inactive gas, and that its activity is not appreciably influenced by temperature or by the most drastic treatment. From the rate of diffusion of these emanations, it appears that they must possess a considerable molecular weight. An investigation of the rate of diffusion of the radium emanation into air, a preliminary account of wh ich appeared in Nature, 1901, p. 157, and Proc. Roy. Soc., Canada, 1901, showed that its molecular weight probably lay between 40 and 100. On account of the rapid loss of activity of the thorium emanation it has not so far been found possible to determine with certainty its rate of diffusion into air or other gases. The emanations from thorium and radium possess very similar properties. They both readily diffuse through glass and porous substances; they both possess the power of ionizing the gas in their neighbourhood and producing excited radioactivity on bodies. The differences between them can be readily accounted for by supposing them to be radioactive gases or vapours of different molecular weights. According to the results which have been given in previous papers, radioactivity is an accompaniment of chemical change. Taking this view, the difference in the rates of decay of the radioactivity of the emanations from

1:\citC'cI RadioaCfivit)' {md fhc Mefhod of ifs li'al1smissiol1

531

thorium and radium merely indicates a difference in their rate of chemical change, and does not imply any fundamental difference in nature. Unlike the radiations from the emanations, the excited radiation due to thorium decays much more slowly than that due to radium (see Rutherford and Miss Brooks, Phi!. Mag., July 1902, p. 18). In this case, the chemical change proceeds more rapidly in the material responsible for the excited radioactivity from radium than in that from thorium. Excited radioactivity is always produced on bodies when the radioactivity emanations from thorium and radium are present. In order to show the very elose connection existing between the presence of these emanations and excited radioactivity, the following experimental facts may be mentioned: (I) Only the radioactive substances which emit emanations, viz. thorium and radium, have the power of exciting radioactivity. Uranium and polonium, not giving off any emanation, do not possess the power of exciting radioactivity. (2) The amount of excited radiation obtained from thorium and radium compounds is directly proportional to the amount of emanation present. For example, thoria gives out far more emanation and produces far more excited activity than thorium nitrate in the solid state. Thoria and radium chloride, partly de-emanated by strong heating, lose their power of exciting activity in like ratio. (3) Excited radioactivity can be produced on bodies if the emanation and not the radioactive substance itself is present. It can be produced at long distances from the radioactive compound by blowing the emanation mixed with air along tubes. In the case of radium, the emanation, which has been introduced into avessei by blowing a current of air over the active substance, produces excited radioactivity after a month's interval, although the radioactive substance itself has not been placed in the neighbourhood. On the other hand, the power that a thorium or radium compound has of producing excited activity on a body near it is almost completely lost by blowing over the compound a current of air which removes the emanation as rapidly as it is formed.

The amount of emanation or excited activity has no direct connection with the radioactivity of the compound in its neighbourhood, and cannot be ascribed to any action of the 'straight line' radiation in the gas through which it passes. For example, de-emanated thoria produces only a small fraction of the amount of excited activity of an equal weight of ordinary thoria although the amount of the straight line radiation is not much affected by the process of de-emanation. § 3. Method 0/ Transmission

0/ Excited Activity

The characteristic property of excited radioactivity is that it can be confined to the cathode in a strong electric field. It is probable, therefore, that

532

The Collected Papers

0/ Lord Rutherford

the radioactivity is due to the transport, in the electric field, of positivelycharged carriers of some kind. Experiments were undertaken to test tbis and to find the rate at wbich these carriers moved in an electric field, in order to obtain a rough estimate of their dimensions compared with a gaseous ion. The method employed to determine tbis velocity is a modification of one already used in adetermination of the velocity of the negative ion, produced at the surface of a metal by ultra-violet light. * It depended on the use of an alternating electric field. By means of a revolving commutator, a direct P.D. was commuted into an alternating P.D. of known frequency. If such an alternating field is applied to two parallel plates, between which a radioactive emanation is kept uniformly distributed, equal amounts of excited activity are produced on each electrode. Ifin series with an alternating P.D. ± E o a battery is placed of E.M.F. EI less than E o, the positive carrier moves in a stronger electric field in one-half alternation than in the other. A carrier consequently moves over unequal distances during the two half alternations, since the velocity of the carrier is proportional to the strength of the electric field in which it moves. It follows from this that the excited radioactivity will be unequally distributed over the two electrodes. If the frequency of alternation is sufficiently great, only the positive carriers within a certain small distance of one plate can be conveyed to it, and the rest, in the course of several succeeding alternations, are carried to the other plate. BB B EMANATION A

B

Fig. I Suppose A and B (Fig. 1) are two parallel plates to be made radioactive. The emanation is supposed to be uniformly distributed between them. When B is negatively charged suppose the P.D. between the plates is Eo-E h when B is positive the P.D. is E o + EI. Let d = distance between the plates. T = time of a half alternation. p = ratio of the excited radioactivity on the plate B to the sum of radioactivity on the plates A and B. K = velocity of the positive carriers for unit-potential-gradient.

* Rutherford, Proc.

Camb. Phil. Soc., 1897.

Excitecl Radioaclivil)' al/d Ihe Met/IOd o.f its Transmission

533

On the assumption that the electric field between the plates is uniform, and that the velocity of the carrier is proportional to thc electric field, the velocity of the positive carrier towards B is Eo - E I • K d '

and in the course of the next half alternation Eo +EI • K d

towards the plate A. The greatest distances· Xio X2 passed over by the positive carrier during two succeeding half alternations is thus given by XI

=

Eo - EI d K.T, and

X2

=

Eo + EI d K.T.

Suppose the positive carriers are produced at a uniform rate of q per second for unit distance between the plates. The number of positive carriers which reach B during a half alternation may be divided into two parts: (1) One half of those carriers which are produced within the distance X of the plate B. This number is equal to !xlqT.

(2) All the carriers which are left within the distance Xl from B at the end of the previous half alternation. The number of these can be readily shown to be 1

2XI'

XI

-qT. X2

Now all the rest of the carriers produced between A and B during a complete alternation will reach the other plate A in the course of the succeeding altemations, provided no appreciable recombination takes place. This must obviously be the case, since the positive carriers travel further in a half alternation towards A than they return towards B during the next half alternation. The carriers thus move backwards and forwards in the changing electric field, but on the whole move towards the plate A. The total number of positive carriers produced between the plates during a complete alternation is 2dqT. The ratio p of the number which reach B to tbe total number produced is thus given by I

2X lqT

p=

XI + 2x l-qT X2 I

2dqT

1 XI

=

4d'

XI

+ X2 X2

• In the equations that follow it is assumed that XI is less than the distance between the plates. If x 1 >d the equations have to be modified.

534

0/ Lord Rutlle/ford

The Collected Papers

Substituting the value of Xl and K=

X2

we obtain

+

2(Eo EI) d 2 . - . p. Eo(Eo - EI) T

In the experiments the values of E o, EI' d, and T were varied, and the results obtained were in general agreement with the above equation. § 4. Velocity

0/ Carriers 0/ Thorium-Excited Activity

For experiments on thorium emanation, a thick layer of thoria was placed in a shallow copper vessel inside an ebonite box 11 cm. square and 3 cm. deep, which was tightly waxed down to a metal base. The thoria was completely covered with two layers of filter paper, which cut off most of the direct radiation, but readily allowed the emanation to pass through. The apparatus was rendered air tight by a metal lid, dipping into a mercury trough round the top of the ebonite box. At the beginning of an experiment a square sheet of aluminium foil was placed over the paper covering the thoria, a zinc plate on top of the ebonite box, and the lid placed in position. This was done as quickly as possible, and the alternating electric field was then applied. The emanation rapidly diffused through the paper and thin aluminium foil, and distributed itself between the plates in the electric field. After an interval, varying in the experiments from 20 to 90 min., the aluminium and zinc plates were removed and their radioactivity tested in the usual way with the Dolezalek electrometer. The ratio of the excited radioactivity on the two exposed plates was thus determined. This ratio was found to be independent of the time the plates were left before testing, as the radioactivity on each plate decays at the same rate. The amount of thoria used in these experiments varied from 25 to 100 grammes. The amount of excited activity in a given time varied with the amount of thoria, but the ratio of the activity in the two plates was unaltered. For a given voltage and time of alternation the value of p was slightly greater when the lower plate was negative. This is due to the une qual distribution of the emanation between the plates; for on account of the time taken in diffusion, the emanation is more concentrated near the surface of the thoria. The mean of the values of p with top plate negative and lower plate negative was taken as the true value. In the early experiments a two-part commutator driven by a motor was used. In the later experiments, for a more rapid rate of alternation, a fourpart commutator was employed. With this arrangement of apparatus a large number of experiments were made in order to test the truth of the general theory. Comparisons of the velocity of the carrier have been made over a wide range of period of alternation and of voltage, and for different distances between the plates. The results obtained were in general agreement with the theory put forward.

/:'.\('/lrc/ Radiol/Cli"ity and fhe Met/IOd of its Transmissiun

535

When the voltage was kept constant, the value of p was found to decrease with increase in the number of alternations per second. With a constant speed of alternation the value of p increased with the voltage. When the value of p is small, the velocities of the carriers, deduced from the equation, were found to be all too high, and also inconsistent among themselves. There are several disturbing factors which have a great influence on the value of p when p is small. These factors are: (1) Recombination and diffusion of the carriers. U nless the electric field is strong, the carriers recombine and diffuse to the electrodes. With a weak electric field the excited radioactivity is disturbed on both the positive and negative electrodes. (2) Inequality of the electric field. In this simple theory we have assumed that the potential gradient between the plates is uniform. This is far from being the case. The experiments of Child and Zeleny have shown that there is always a sudden drop of potential near the electrodes. The electric field near the electrodes is consequently stronger than the average. For this reason, when the carriers which reach the plate Bare only abstracted from within a short distance of the plate, the value of p leads to too high a value of the velocity. (3) Initial velocity of the carrier. From some considerations which will be developed later (see § 9), it seems probable that the positive charge on the carrier is due to the expulsion of a negatively charged partic1e of some kind from the neutral molecule. The positive carrier may thus have enough initial velocity imparted to it to carry it some distance against the electric field. This will result in a distribution of some excited activity on the anode, even with a strong field. It is difficult to obtain direct experimental evidence on this point, but there seems Httle doubt that such an effect is present.

In order to obtain consistent results, it was found necessary to have a considerable difference between the strength of the electric fields during the succeeding half alternations. If the difference is small, the carriers take so long to reach the plate A that recombination and diffusion of the ions become important factors in determining the distribution of excited activity. For the reasons we have explained above, it was necessary to use fairly high voltages and correspondingly rapid speed of alternation in the experiments. The following tables are examples of some of the results obtained for different voltages and distances between the plates. Temperature 18°C. Air fairly dry. PLATES

Eo+El

75

152 225 300

Eo-El

50

101

150 200

1· 30 CM.

APART

Alternations

per sec.

p

57 57 57 57

0·17 0·27 0·38 0·44

K

1·7 1·25 1·17 1·24

536

The Collected Papers

0/ Lord Rutherford

The value of K is given in centimetres per second for a potential gradient of 1 volt per centimetre. For the last example, since the carrier travelled over a distance greater than 1·30 cm. during each half alternation, a modified form of the equation was necessary to calculate the velocity. The value of 1·6 cm. per sec. for 50 volts is too high for the reasons explained above. PLATES

Eo+E1

273 300

2 CM.

APART

Altemations per sec.

Eo-E1

207 200

44

53

p

K

0·37 0·286

1·47 1·45

An average value of the velocity from a large number of alternations for different distances, voltages, and speed of alternations was about 1 . 3 cm. per sec. for atmospheric pressure and temperature. This velocity is not very different from the velocity of the positive ion produced by Röntgen or Becquerel rays. The most accurate determination of this velocity by Zeleny* gave a value of 1· 37 cm. per sec. for dry air. § 5. Increase

0/ Excited Radiations with Time

In the course of these experiments a remarkable effect was observed. It was found that a plate which has been exposed a short time in the presence of thoria emanation, after being removed, gradually increased in radioactive power for several hours. The amount of increase varied with the time of exposure to the emanation, but in short exposures it increased to three or four times its initial value. For exposures of several hours the effect is not so marked, and is difficult to detect after a day's exposure. The following tables illustrate the results obtained: (1) PLATINUM WlRE, CHARGED -110 VOLTS, EXPOSED 15 MIN. IN A CYLINDER CQNTAINING THORIA. FIRST OBSERVATION 5 MIN. AFTER REMOVAL OF WIRE FROM EMANATION CYLINDER.

Time in min.

Movement of electrometer in scale-divs. per sec.

7·5 24 43 58 78

2·8 4·0 4·6 5·2 5·9 6·5

o

99

* PhU.

1·9

Trans. Roy. Soc., 1900.

1:"xC'iI(,c! Rat/inaeril'i,}' am/Ihr MN/IOd of i/S Transmission

537

In this case the activity had increased over three times in 99 min. and had not reachcd its maximum value. (2) ALUMINIUM FOIL AS CATHODE IN PARALLEL PLATE APPARATUS OF FIG. 1. TIME OF EXPOSURE 41 MIN. FIRST OBSERVATION 6 MIN. AFTER REMOVAL.

Time

o

21 min. 31 min. 57 min. 70 min. 91 min. 120 min. 160 min. 180 min. 22 hr. 49 hr.

Radioactivity

1 1·6 l·g 2·0 2·2

2·5 2·9 2·9 2·9

1·0 0·21

In this case, for the purpose of comparison, the initial value of electro~ meter current is taken as unity. The activity increases to nearly three times its initial value after an interval of 2 hr., and then slowly decreases at the normal rate, Le. it falls to half value in about 11 hours. Similar results were obtained if the plate was made active without the action of the electric field. The increase of activity with time is independent of the nature of the electrode, or of the concentration of the radioactive matter upon it. It was not found possible to infiuence the rate of increase of activity with time or the final maximum by heating the wire to about a red heat. With increase of time of exposure of the electrode in the thorium emanation, the ratio of increase of activity after removal decreases. For a long interval of exposure the activity begins to decrease at once after removal. This result is to be expected, for the activity of each portion of the radioactive matter deposited increases with time for two or three hours, and then diminishes. Consequently, after the electrode has been exposed for about ten hours or more, the increase of activity of the matter deposited in the last few hours does not compensate for the decrease of activity of the radioactive matter as a wh oIe. This increase of activity with time explains an irregularity in the curve of increase of excited activity from thorium with time of exposure. It was pointed out in a previous paper (Phi/. Mag., February 1900, p. 178) that, on the hypothesis of a uniform rate of deposit of radioactive matter, the activity of which decreased in a G.P. with the time, the curve of rise of excited activity with time of exposure is the same as the curve of rise of an electric current in a circuit of constant self-inductance. It was experimentally observed, however, that the rate of increase for the first few

538

Tlte enTleeret! Papers oJ Lord Ru/herJorel

ho urs was much smaller than would be expected on this hypothesis. In the light of the present results, the explanation of this effect is simple. The matter deposited during the first few hours does not reach its maximum activity for several hours, and the initial effect is consequently much sm aller than would be expected on the simple theory.

§ 6. Radium-Excited Radioactivity Experiments were made to determine the velocity of the carriers responsible for the excited radioactivity of radium in the same way as for thorium. The radium in my possession gave out too little emanation at ordinary temperatures in the solid state to enable me to use it in the apparatus in place of thoria. The amount of emanation from radium can, however, be increased several thousand times by heating the radium compound below a red heat. A more convenient method of obtaining a large amount of emanation is to dissolve a small quantity of radium chloride in water. Radium in solution gives off several hundred times more emanation than in the solid state. If the solution is kept in a c10sed vessel the emanation continually collects in the upper part of the vessel. It can be transferred at any time to another vessel by bubbling a slow current of air through the solution. The procedure adopted in introducing the emanation into the apparatus was as follows: A large amount of emanation was introduced into a metal cylinder of about 3 litres capacity. The plates to be tested were then placed in position in the apparatus of Fig. 1, and the alternating E.M.F. applied. By means of side tubes, the apparatus was put into connection with the emanation vessel and a small fraction of the emanation introduced into it by sending a slow current of air into the cylinder. The tubes leading into the apparatus were then c1osed, and the alternating E.M.F. continued for an interval of 15 to 30 min. Before stopping the commutator, the emanation was blown out of the apparatus by a slow current of air. The plates were then removed, and the amount of activity on them was compared by means of the electrometer. On account of the initial rapid decay of the radiations, a difficulty arose in comparing the amount of radiation on the two plates. As shown in a previous paper, * the excited radiation from radium, for short exposures, decreases rapidly for the first 5 min. after removal, but about 15 min. after removal reaches a value which is maintained fairly constant for an interval of about 10 min. It then decays to zero, falling to half value in about 30 min. The comparison of the activity on the two plates was made during this constant interval. When experiments were made under the same conditions as those for thorium, somewhat higher values of the velocity of the carriers were obtained, and the numbers, for different frequencies and voltages, differed considerably among themselves. These discrepancies were found to be due

* E. Rutherford and Miss Brooks, Phil. Mag., July 1902.

L\ciled RadioQclil'il)' ami Ilw Melhod of ils li"allsmissicm

539

to the fact that even in a strong electric field from 5 to 10 per cent of the total excited activity was distributed on the anode. In this respect the activity excited by radium differs from that of thorium. Consequently, the value of p would be greater for the radium than for the thorium experiment, under the same conditions, even if the carrier of excited radioactivity travelled at the same rate in both cases. § 7. Distribution of Excited Activity on the Anode

In order to throw more light on the cause of this distribution of excited activity on the anode, some experiments were made with the apparatus shown in Fig. 2a. A

B

III'1Jr

c

D

E

e.rth

EMANATION CYLINDER RAD. CL

Fig.2a The emanation vessel A consisted of a brass cylinder 25· 5 cm. long and 8·30 cm. diamater. A long central brass rod B C D E, diameter 0·518 cm., passed through an ebonite cork at one end of the tube. The outside cylinder was connected to one pole of a large battery, the other pole of which was earthed. The central rod was connected to earth. The emanation was introduced into the vessel by sending a slow current of air through a radium chloride solution contained in the Drechsel bottle F. The air passed through a tube containing cotton wool, and through a drying-tube T of calcium chloride. The central rod was made of three removable parts, BC, CD, DE, screwed together. After exposure for a known time in the presence of the emanation the rod was removed, and the activity on the portion CD, length 15 cm., determined by the electrometer in the cylindrical testing vessel L, shown in Fig. 2b. By this means the excited activity was determined on that portion of the rod where the electric field was sensibly uniform. TESTINS CYL.lNDER

A

e.rth

TESTINS CYL.lNDER

Fig.2b

540

The Collected Papers 0/ Lord Ruther/ord

In most of the experiments the emanation was introduced into the vessel A a day or two before observations were taken. This ensured a uniform distribution of the emanation throughout the cylinder by the process of diffusion. If observations were required soon after the introduction of the emanation, the emanation was uniformly mixed with the air by means of astirrer not shown in the figure. Some experiments were made with this apparatus on the amount of excited activity on the central rod when positively charged for different voltages. For the purpose of comparison, the results are expressed in terms of the percentage amount on the same electrode exposed for the same time when negatively charged with 300 volts between the electrodes. The rod was exposed in the presence ofthe emanation for 15 min., then removed and a fresh rod introduced. In the course of tbree or four hours' work, the amount of excited activity, obtained for a given time of exposure, diminished about 20 per cent. This was partly due to the decay of the radiating power of the emanation during the interval, and partly to a slight escape of the emanation in removing and replacing the central rod. By determining the ratio of excited activity at the beginning and end of the experiments, a correction was readily made for this diminution. TABLE OF DISTRIBUTION ON ANODE, DIAMETER

8·3

Voltage

Percentage

-300 +300 +150 + 50 + 20

100 6 6

o

MM.

9 10

14

It will be seen from the table that the amount on electrode with +300 volts P.D. is 6 per cent of amount on electrode with -300 volts P.D. The percentage increases with diminution of voltage, rising to 14 per cent for zero voltage, when the distribution is due to diffusion alone of the carriers to the central electrode. In order to see how much of this amount on the electrode was due to transmission by the electric field and how much to diffusion, the experiments were repeated with the central rod of 0·8 mm. diameter instead of 8· 3 mm. ;

-300 +300 + 50

100 4 7

Now, with a central rod of only about 1/10 of the surface area, it is obvious that the effect due to diffusion must be very much reduced. We may thus conclude from these experiments that a proportion of the excited

Freitet! Radioa('lil'il.\' am/lh(' M('lhod 0/ its Transmission

541

radioactivity from radium (about 5 per cent) travels to the positive electrode in an electric field, and the carrier must, in consequence, have a negative charge. A special experiment was made to determine the amount of radiumexcited activity on the anode with an apparatus consisting of parallel plates. For this purpose the emanation vessel of Fig. 1 was used, with the plates 1 . 3 cm. apart. With 300 volts between the plates about 10 per cent of the total activity was confined to the anode. From the previous experiments we have seen that about 5 per cent reaches the anode in consequence of transmission by the electric field. The remaining 5 per cent must thus reach the electrode by other agencies. With such a strong el ectric field, the effect due to pure diffusion must be very smalI. It thus seems likely that some of the radioactive carriers have sufficient initial velocity to carry them to the electrode against the electric field.

§ 8. Velo city

0/ the Carrier

Some experiments were made on the velocity of the carrier of excited radioactivity, using the concentric cylinders shown in Fig. 2a. If a and bare the radii of the internal rod and the cylinder, the electric field X at a distance r from the centre, for a p.n. of V volts between the cylinders, is given by

X=

V

b' rIo&: a

Using the same notations and assumptions as in the case of parallel plates, it can be shown that the velocity K of the carrier of excited radioactivity is given by

K

E o + E, EoCEo - EI)

(b' - a') log.,

~

T a . p.

The following table shows some of the results obtained for different voltages and periods of alternation: E o - EI

205 205 205 205 385 385

Eo

+ EI

308 308 308 308 580 580

Altemations per sec.

5·7 11·2 17·2

34

18·3 47

p

0·35 0·24 0·19 0·16 0·30 0·16

Corrected value of p

K

0·32 0·20 0·13

1·0 1·2 1·2

0·27

1·5

? ?

The values of the velocity of the carrier determined for the uncorrected values of p in the above table vary much among themselves. It will be seen,

542

The Collected Papers 0/ Lord Rutherford

however, that the value of p increases with the time of alternation and the voltage, as we should expect from the elementary theory. It was not found possible to reduce the observed value of p below about 0·16, whether the voltage was diminished or the frequency of alternation increased. This is to be expected, for we have previously shown that the amount of excited activity in the central rod, when diffusion alone is acting, is 0·14 of the total; for in cases where the carrier is only able to travel over a small fraction of the distance between the electrodes during a half alternation, only a small amount of the excited radioactivity on the central rod is due to the deposit of positively charged carriers by the electric field. The greater proportion is due to the diffusion of carriers to the electrode and to the carriers which are deposited when the central rod is the anode. The amount of this 1atter is, as we have shown, about 5 per cent of the total. It is difficult to make more than a rough estimate of the amount of excited activity on the elec-trode in the various cases. For these reasons, the corrected values of p in the table for the frequencies of 34 and 47 per second are probably not more than about one-third of the observed value 0·16. In the above table a rough correction is made for some of the values of p and the resulting velocity calculated. It will be seen that the values of the velocity of the carrier He between 1 ·0 and 1· 5 cm. per sec. for a potential gradient of one volt per centimetre. This is about the same range of values as that obtained for the carrier of thorium-excited radioactivity. From the nature of the results it is not possible to definitely decide whether the carriers of thorium- and radium-excited activity travel at exactly the same speed. The results, however, indicate that the carriers in the two cases are not very different in speed, and that consequently they do not differ much in size. We may conclude from these experiments that the greater part of the excited radioactivity from both thorium and radium compounds is due to the deposit of positively charged carriers, produced from the emanations on the cathode, and that these carriers travel at about the same rate as the positive ion produced in the air by Röntgen rays. When no electric field is acting excited radioactivity is transferred by the diffusion of these carriers to the surface of all bodies immersed in the emanation. § 9. Origin

0/ the Carriers

Before discussing the question of the method of production of these positive carriers which cause excited activity, abrief resume is necessary of the physical properties of the emanations from thorium and radium. In the first place the emanations behave in all respects like radioactive gases of high molec-ular weight. They do not carry with them any charge of electricity, and are consequently unaffected by the presence of an electric field. In my first paper on the thorium emanation, loc. cit., it was pointed out

l:;xcited Radioaetivity and the Method 0/ its Transmission

543

that the particles constituting the emanation certainly did not move with a velocity greater than 0·00001 cm. per sec. for a gradient of one volt per centimetre. The conclusion was drawn that the emanation itself was initially uncharged. A similar result is true for the emanation from radium; for the emanation still persists in a closed vessel after several weeks' exposure in a strong electric field. For these reasons the suggestion made by Becquerel, * that the emanations are composed of positive ions directly emitted from radioactive bodies, is untenable; for if such were the case, the ions would at once be swept to the electrodes by the electric field, and would very rapidly disappear from the gas. These emanations possess the property of ionizing the gas and of producing from themselves positively charged carriers which cause excited activity in bodies on which they are deposited. This property lasts only a few minutes in the case of the thorium emanation, and for several weeks for the radium emanation. Two hypotheses may be put forward to account for the origin of these charged carriers: (1) The radioactive matter constituting the emanation condenses on the positive ions, produced in the gas by the radiation, and is thus transferred to the cathode. (2) The particles of the emanation possess the property of expelling from themselves a negatively charged body of some kind. The particle would thus be left with a positive charge, and would be carried to the cathode by the electric field. It is not easy to decide definitely between these two hypotheses, but the evidence as a whole is strongly in favour of the second. In regard to (1) it might be supposed that the emanation condensed more readily on the positive than on the negative ion; on the principle that waterand a1cohol-vapour condense more readily on the negative than on the positive ion. If this were the case, it would be expected that the emanation would be removed more rapidly if the number of ions were increased in the gas through which the emanation was distributed. There is no evidence that such an effect exists. I have tried the experiment of passing the emanation through aspace strongly ionized by radium rays; but the amount of excited activity in a given time on the cathode, placed in this space, was not appreciably altered. I have also tried experiments to see if the radiation from the emanation was affected by exposure in a strong electric field, but with negative results. In order to test this, the thorium was placed in the bottom of a small lead box, and covered with two layers of paper to cut off the direct radiation. The top of the box was tightly covered with a very thin layer of mica. This prevented the escape of the emanation, but allowed the radiation from the emanation to pass through and ionize the gas above the vessel. The amount of this ionization outside the vessel was unchanged

* Comptes Rendus, December 9,

1901.

544

The Collected Papers

0/ Lord Ruthelford

if the emanation was exposed to a strong electric field by charging insulated conductors placed inside the vessel. If the emanation is removed to the cathode in an electric field by condensation on the ion, it is to be expected that it would continue to radiate at the same rate on the electrode as in the gas from which it is removed. On this view the radiation from the cathode should rapidly decrease for the first few minutes after removal from the emanation. Some special experiments were tried to settle this point, but no decrease was observed, although even aminute effect could have been readily detected. There is thus considerable indirect evidence against the condensation hypothesis; and it has consequently been discarded in favour of (2), which offers a satisfactory explanation not only of the production of positive carriers, but also of the origin of the radiation given out by the emanation itself. On this view the emanation consists of matter in an unstable state, which undergoes further chemical change. The change consists in the expulsion of a negative particle from the neutral molecule. The residual portion of the molecule retains a positive charge, and is carried at once to the cathode in an electric field. This matter again undergoes chemical change, giving rise to the phenomena of excited radioactivity. The experimental data in favour of this view are best considered in the next section (§ 10) on the nature of the radiations. It has been shown that the carriers of excited activity for both thorium and radium travel at about the same rate as the ions produced in the air by Röntgen rays. From data of the Kinetic Theory of Gases it has been shown that the ion in air is probably large compared with the molecule of oxygen or hydrogen. This has been explained by supposing that the ion, immediately after its production, becomes the centre of a cluster of molecules which move with it. On this view that part of the emanation molecule which retains a positive charge immediately becomes the nucleus of an aggregation of molecules of the surrounding gas. The size of the cluster is probably about the same for the positive ion, since the size is mainly determined by the electric charge which is the same for both. The velocities in an electric field are thus the same for the carriers of excited activity and for the gaseous ion. Since the size of the cluster is large compared with the original nucleus, the velocities of the carriers of thorium- and radium-excited activity would be about the same, even if the original nuclei were of different masses. § 10. Nature 01 the Radiations

In considering the question of the size of the body expelled from the molecule of the emanation, and of the nature of the radiation from the emanation, it is necessary to take into account the nature of the emanations emitted from all the known radioactive bodies; for there is no reason to suppose that the processes which are taking place in the moleeule of the emanation are essentially different in character from those occurring in the other

h'xcilet! Raclioaclirily

al/(l lire ,'-'[eI/lOt! 0/ ils Transmission

545

radioactivc bodies. lt is known that uranium, thorium, and radium emit two types of radiation. One type is not appreciably deviable by a magnetic or an electric field, and is very easilyabsorbed in matter. These will be called the oe rays. The others are deviable and more penetrating in character, and will be called the ß rays. In addition, I have shown that thorium and radium emit some rays non-deviable in character, but of very great penetrating power. All of the radioactive substances inc1uding polonium as well as 'excited' bodies and the emanations give out these oe rays. Their power of ionizing the gas is very much greater than for the other types of rays emitted ; and it is probable that the greater proportion of the energy radiated into the gas is in the form of oe rays. The oe rays from different radioactive substances, inc1uding the emanations of 'excited' bodies, do not vary very much in penetrating power. The 'excited' radiations for thorium and radium are the most penetrating in character and that of uranium the least. It has been difficult to offer a satisfactory explanation of the nature of these rays. I have previously shown as untenable the view that they are secondary rays due to the emission of ß rays. I have been recently led, by a mass of indirect evidence, to the view that the oe rays are in reality charged bodies projected with great velocity. The ionizing effect of the rays is due to the collision of the projected body with the molecules of the gas, in the same way that the cathode rays ionize the gas in their path. Such a projected partic1e probably produces many thousand ions in its path before its velocity is reduced to the point below which it can no longer ionize the gas. Strutt has put forward the view that the oe rays were positively charged bodies since the ß rays emitted from the same body carried a negative charge. This view has also been advanced in arecent paper by Sir W. Crookes. The evidence in favour of the projection nature of the oe rays is so far all indirect in character, and is briefly summarized below: (1) The absorption of the oe rays in matter (like the ß rays which we know are projected partic1es) is approximately proportional to the density of the material. It has been shown that the absorption of uranium, thorium, and radium rays is roughly proportional to the density for air and for aluminium. (2) The absorption of the oe rays by a given thickness of matter increases rapidly with the thickness traversed. I have found that this is a general property of the oe radiations not only for the radioactive elements proper, but for the radiations from the emanation and excited bodies. This is to be expected if the rays consist of projected particles, but is difficult to explain if the radiations are ether waves similar to Röntgen rays. (3) In the case of the emanations we have direct evidence that a negatively charged particle* is projected. The radiation from the emanation is due to

* I was at first inclined to suppose that the particle expelled from the emanation was a negative electron, since it is known that both thorium and radium compounds and bodies excited by them emit some deviable rays. I have, however, made a close examination S

546

The Collected Papers

0/ Lord Ruther/ord

these projected partic1es which ionize the gas in their path. This satisfactorily explains the experimental observation that the amount of excited activity is directly proportional to the amount of radiation from the emanation. It also serves to explain the fact, especially noticeable in the experiments on the radium emanation, that some of the carriers of excited activity have sufficient initial velocity to move against the electric field. This velocity is due to the recoil consequent upon the projection of the charged body. If these rays are due to projected charged partic1es, they should possess the properties of the IX rays of deflectability by a magnetic and e1ectric field. No deviation of the IX rays has so far been detected in a strong magnetic field, but the experiments have not yet been pushed to the necessary limit. The results, however, indicate that if the rays are deflectable, the deviation is minute compared with the ß rays. This is to be expected if the mass of the expelIed partic1e is large compared with the electron. If, for example, the projected body had a mass 10 times that of the hydrogen atom, it would require a magnetic field about 10,000 times as strong to produce the same deviation as for the electron moving with the same velocity. There is evidence that large carriers moving with a high velo city are produced in vacuum tubes. W. Wien* has shown that the 'Canal Strahlen' of Goldstein are positively charged particles moving with high velocity. These rays are deviated by a magnetic and electric field. When the vacuum tube is filled with hydrogen the ratio of the charge to the mass, ~, of these carriers is m about 104, showing that the carriers have the same mass as the hydrogen atom. In an atmosphere of oxygen the size of these carriers is considerably greater than the hydrogen atom. It is possible that the electric charge on the expelled partic1e may be different for different radioactive bodies under different conditions. For the emanations of thorium and radium the expelled particles are for the most part negative. It has been shown that some of the radium carriers of excited activity have a negative charge, showing that the expelled body is positive. In addition, Dorn has shown that in a radium solution the excited radioactivity is produced on the anode, and not on the cathode. This shows that the carriers of excited activity in solution have a negative charge, so that the expelled body is positive in sign. § 11. Evidence 0/ Chemical Changes

In previous papers by Mr Soddy and myself, the view has been put forward that radioactivity is an accompaniment of continuous chemical change. of the radiation from the emanation by the electrical method, but was unable to detect the presence of any penetrating deviable rays. If such deviable rays are present, they certainly exist in far less proportion compared with the CI: rays than in the other radioactive substances. * Drude's Annal., No. 6, p. 244, 1902.

Fxrifed Radioartirify and the Met/IOd of its Transmission

547

Taking, for example, thorium, which has been worked out more thoroughly than the other radioactive bodies, it has been shown that a chemical substance ThX is produced at a constant rate by the thorium compound. This ThX undergoes further chemical change, one of the products of which is the emanation. This emanation itself is not stable, but expels from itself a negatively charged body. The positively charged portion of the emanation is carried to the electrodes, and this again undergoes further chemical change, giving rise to the phenomenon of excited radioactivity. There is thus evidence of four distinct changes, in each of which the matter produced has distinct chemical properties. For example, ThX is soluble in ammonia, while thorium and products of the later changes are not. The emanation is not soluble in hydrochloric or sulphuric acid, unlike the matter responsible for excited radioactivity. There is strong evidence also that the chemical changes in the matter responsible for excited radioactivity are complex in character. It has been shown that the excited radiation in a body increases after removal when the body is exposed for a short time in the presence of the emanation. This effect is analogous to the increase of radioactivity of ThX for the first day after separation, which has been shown to be due to the excited activity produced in the matter constituting the ThX. In order to account for the increase in radiating power after removal, one must suppose that the matter which is deposited from the thorium emanation gradually undergoes a chemical change. The transformed matter undergoes a secondary change, the time-rate of which is slower than the primary, but which gives rise to greater radioactivity. From data of § 5 half the matter has undergone change about 1 hour after deposit, while in the secondary change the corresponding time is about 11 hours. Summary of results (1) Excited radioactivity produced by thorium and radium compounds is due to the deposit ofradioactive matter, which is derived from the emanation given out by these bodies. (2) Excited radioactivity is transmitted by positively charged carriers, produced from the emanation, which travel in an electric field with about the same velocity as the positive ions produced in air by Röntgen rays. This velocity (about 1· 3 cm. per sec. for 1 volt per cm.) is ab out the same for the carriers of thorium- and radium-excited activity. (3) These positively charged carriers are due to the expulsion of a negatively charged body from the molecule of the emanation. (4) Evidence is adduced for the view that the easily absorbed and apparently non-deviable rays of radioactive substance are due to the expulsion of charged bodies at a high velocity. The rays are thus analogous to the 'Canal Strahlen' of Goldstein, which Wien has shown to be positively charged bodies projected at a great speed.

548

The Collected Papers of Lord Rlitherford

In the case of the emanations the expelled particles are for the most part negative in sign. (5) In the case of radium about 5 per cent of the carriers of excited activity are distributed on the anode in a strong electric field. (6) The excited radiations from thorium due to a short exposure in the presence of the emanation increase in the course of several hours after removal, to three or four times their initial value. (7) The emanations and the matter which gives rise to excited activity are the result of a succession of chemical changes occurring in radioactive matter. In thorium there is evidence of at least four distinct chemical changes. Macdonald Physics Building McGill University, Montreal July 29, 1902

The Magnetic and Electric Deviation of the Easily Absorbed Rays from Radium by E.

R UTHERFORD, M.A., D.se.

Macdonald Professor 0/ Physics, McGill Unil'ersüy, Molllreal From the Philosophieal Magazine for February 1903, sero 6, v, pp. 177-187

RADIUM

gives out three distinct types of radiation:

(1) The oe rays, which are very easily absorbed by thin layers of matter, and which give rise to the greater portion of the ionization of the gas observed under the usual experimental conditions. (2) The ß rays, which consist of negatively charged particles projected with high velocity, and which are similar in all respects to cathode rays produced in a vaeuum tube. (3) The y rays, whieh are non-deviable by a magnetic field, and which are of a very penetrating eharaeter. These rays differ very widely in their power of penetrating matter. The following approximate numbers, which show the thickness of aluminium traversed before the intensity is reduced to one half, illustrate this differenee: Radiation oe rays

ß rays

y rays

Thickness of aluminium

0·0005 em. 0·05 em. 8 em.

In this paper an aecount will be given of some experiments which show that the oe rays are deviable by a strong magnetie and eleetrie field. The deviation is in the opposite sense to that of the cathode rays, so that the radiations must consist of positively charged bodies projected with great velocity. In a previous paper* I have given an account of the indirect experimental evidence in support of the view that the oe rays eonsist of projected eharged particles. Preliminary experiments undertaken to settle this question during the past two years gave negative results. The magnetie deviation, ... Phi!. Mag., January 1903, p. 113. It was long ago suggested by Strutt (Phi!. Trans. Roy. Soc., 1900) that the oe rays consist of positively charged partic1es projected from the active substance. The same idea has lately been advanced by Sir Wm Crookes (Proc. Roy. Soc., 1900).

550

The Collected Papers

0/ Lord RU/helford

even in a strong magnetic field, is so small that very special methods are necessary to detect and measure it. The smallness of the magnetic deviation of the IX rays, compared with that of the cathode rays in a vacuum tube, may be judged from the fact that the IX rays, projected at right angles to a magnetic field of strength 10,000 C.G.S. units, describe the arc of a circle of radius about 39 cm., while under the same conditions the cathode rays would describe a circle of radius about 0·01 cm. In the early experiments radium of activity 1000 was used, but this did not give out strong enough rays to push the experiment to the necessary limit. The general method employed was to pass the rays through narrow slits and to observe whether the rate of discharge, due to the issuing rays, was altered by the application of a magnetic field. When, however, the rays were sent through sufficiently narrow slits to detect a small deviation of the rays, the rate of discharge of the issuing rays became too small to measure, even with a sensitive electrometer. I have recently obtained a sampie of radium* of activity 19,000, and using an electroscope instead of an electrometer, I have been able to extend the experiments, and to show that the IX rays are all deviated by a strong magnetic field. Magnetic Deviation

0/ the Rays

Fig. la shows the general arrangement of the experiment. The rays from a thin layer of radium passed upwards through a number of narrow slits, G, in parallel, and then through a thin layer of aluminium foll 0·00034 cm. thick into the testing vessel V. The ionization produced by the rays in the testing vessel was measured by the rate of movement of the leaves of a gold-Ieaf electroscope B. This was arranged after the manner of C. T. R. Wilson in his experiments on the spontaneous ionization of air. The gold-Ieaf system was insulated inside the vessel by a sulphur bead C, and could be charged by means of a movable wire D, which was afterwards earthed. The rate of movement of the gold-Ieaf was observed 'by means of a microscope through small mica windows in the testing vessel. In order to increase the ionization in the testing vessel, the rays passed through 20 to 25 slits of equal width, placed side by side. This was arranged by cutting grooves at regular intervals in side plates into which brass plates were slipped. A cross section of the system of metal plates and air spaces is shown in Fig. Ib. The width of the slit varied in different experiments between 0·042 and O·lcm. The magnetic field was applied perpendicular to the plane of the paper and parallel to the plane of the slits. The testing vessel and system of plates were waxed to a lead plate P so that the rays entered the vessel V only through the aluminium foi!.

* The sampIe of radium of greater activity than that usually sold was obtained from the Societe Centrale de Produits Chimiques, through the kindness of M. P. Curie.

Ma~l1ctic a1ll1 .

F.!('ctric DCl'iation or Fasilr. Ahsorhed Rars ..{rom Radium 551

It is nece.fO I-..

~

i::: 20

\J ~

o

2

i

6

8 10 '2 TIME IN DAYS

1+

/6

Fig.l emanation drawn from the table in § 2. The two curves are quite analogous to those obtained for the decay and recovery of UrX and ThX respectively. The proportion of the activity regained after an interval t is given by , It - = l-e-I\t 10

574

The Collected Papers of Lord Rutherford

where Ais the coefficient of decay of activity of the radium emanation. The curve of recovery of the activity of radium can thus be deduced if the rate of decay of the emanation is known. In these experiments the non-separated activity was 25 per cent of the final activity. The activity of the emanation, together with the excited activity it produced, made up the other 75 per cent. The above equation is only approximate owing to the existence of aperiod of retardation of a few hours, due to the time taken for the excited activity to reach a steady value for any given quantity of emanation present. It is, however, too small to affect the results appreciably and may be neglected. Somewhat similar experiments were performed for thorium. A sampie of thorium hydroxide of high emanating power was ignited over a blastlamp and converted into the de-emanated oxide. Its radioactivity was found to rise, in consequence, about 20 per cent in three days, and remained constant at the higher value. In the converse experiment the thorium hydroxide was kept for three days in liquid air, Le. under conditions where its emanation is condensed and produces the excited activity in the compound itself. It was spread on a plate and its activity found to decrease about 12 per cent after a few days. These results are to be expected if the rate of production of emanation is constant and independent of chemical or physical conditions. When the emanation is prevented from escaping, its activity, and also the excited activity produced by it, cause an increase in the intensity of the radiations emitted. In the second case, the activity decreased, since some of the emanation escaped from the compound at ordinary temperatures, and, in consequence, some of the excited activity deposited in the compound gradually decayed. § 5. The Radiations of Radium

Radium, like thorium and uranium, emits two types of radiation, the ce, or easily absorbed rays (deflectable in very intense magnetic fields), and the ß, or penetrating rays, readily deviated in a magnetic field. It also emits some very penetrating rays, which, however, have not yet been fully investigated. The non-separable activity of radium, which remains after the emanation and excited activity have been removed, consists only of ce rays, the ß radiation being less than 1/200 of the amount normally present. In this respect the three radio-elements are analogous. The radiation from the radium emanation was tested by introducing it in a cylinder made of copper sheet 0·005 cm. thick, which absorbed all the ce rays and allowed the ß rays to pass through with but little loss. The external radiation from this cylinder was determined at intervals commencing about 2 minutes after the introduction of the emanation. The amount at first observed was extremely smalI, but increased rapidly and reached a practical maximum in 3 or 4 hours. Thus the radium emanation also only gives ce rays, the ßrays appearing after the latter has changed into the excited activity. On sweeping out the emanation from the cylinder by a current of air there

.4 ("o/ll/,orati\,(' :"::i/udy of fhe Radioaetirity of Radium ami 71/Orium

575

was no appreciable decrease of the radiation immediately, but the radiation commenced to decay rapidly with the time, falling to half value in about 30 minutes. A similar result has been obtained by P. Curie. Attention has been called (Rutherford, Phi!. Mag., January 1903) to the irregular character of the curves of decay of both thorium- and radiumexcited activity, as measured by the radiation, and the view was put forward that this stage probably represents a double change in the case of thorium, and a treble change in the case of radium. In the latter there is (for a short exposure to the emanation) a very rapid decrease for the first 10 minutes to about 20 per cent of the original value, then aperiod of very slow change, and then a more regular decay in which the remaining activity falls to half value in about 30 minutes. Now the decay curve of the ß radiation of the radium-excited activity shows a fairly regular decrease to half value in 30 minutes. Hence there is strong evidence that the ß rays are not given out in the first change of the excited activity, but only in the second or third change. Radium therefore fully supports the view al ready advanced that the rays are in all cases the first to be produced, the ß rays only resulting in the last stages of the process that can be experimentally traced. (X

(X

§ 6. The Chemical Nature

01 the Radium Emanation

The experiments already described on the chemical nature of the thorium emanation were repeated for that of radium. As in the former case, all the reagents tried were without effect. The emanation passed unchanged through phosphorus pentoxide, sulphuric, nitric, and hydrochloric acids, and over red-hot lead chromate and metallic magnesium. Water does not dissolve the emanation appreciably, and the activity of the water is solely due to the presence of the excited activity. The emanation in both dry and moist atmospheres is unaffected by passage through a platinum tube electrically heated to the point of incipient fusion. An interesting effect, however, was observed as the temperature approached a white heat in this experiment. The ionization current due to the emanation decreased with rise of temperature, but returned to its original value when an increased voltage was applied sufficient to give a saturation current through the gas. This effect is due to fine platinum dust given off from the white-hot platinum, and is quite analogous to that of tobacco smoke observed by Owens (Phi!. Mag., 1899, xlviii, p. 377) in this laboratory. The condensation of the radioactive emanations of thorium and radium at the temperature of liquid air (Proe. Chem. Soc., 1902, p. 219) will be dealt with in detail in aseparate communication. McGill University. Montreal February 20, 1903

Some Remarks on Radioactivity From the Philosophical Magazine for April 1903, sero 6, v, pp. 481-5

To the Editors

0/ the Philosophical Magazine

GENTLEMEN,

IN arecent number of the Comptes Rendus, Jan. 26, 1903, there appeared a paper by M. Henri Becquerel, entitled 'Sur la deviabilite magnetique et la nature des certains rayons emis par le radium et le polonium', and also one by M. P. Curie, 'Sur la radioactivite induite et sur l'emanation du radium', in the former of which certain criticisms of my experimental methods, and in the latter of my theoretical views were made. I am very pleased that M. Becquerel, with the very active material at his disposal, has confirmed in such a direct manner by the photographie method the results which I had previously obtained by the electric method, * showing that the cx or easily absorbed rays of radium are deviated by an intense magnetic field. In the course of his paper M. Becquerel states, 'M. E. Rutherford, avec une grande habilete, et par une methode electrique relativement grossiere, a reconnu un phenomene d'une extreme delicatesse. Cependant la methode employee laisserait prise a diverses objections et a un doute sur l'existence du phenomene en question si l'on n'en apportait pas d'autre preuve. L'une des objections resulte de la disposition experimentale qui fait traverser des espaces laminaires par un rayonnement dont la partie cathodique est rejetee :,ur les parois, et les rayons secondaires qui en resultent peuvent donner lieu a des effets dans le sens observe par M. Rutherford. Je me suis alors propose de mettre en evidence le phenomene par une experience plus simple et plus sure. J'ai eu recours a l'une des dispositions photographiques que j'emploie depuis longtemps et qui permettent certaines distinctions qui n'apparaissent pas toujours quand on emploie exc1usivement la methode electrique' . I fully recognize the great simplicity and utility of the photographie method, so ably developed by M. Becquerel himself, for determining the magnetic deviation of rays and determining accurately the curvature of their path, provided that the rays emitted are sufficiently intense to affect a photographic plate without too long an exposure. The difficulty of obtaining satisfactory photographs in cases of less active material is c1early shown by M. Becquerel's own experiments on the rays of polonium described in the same paper. I have unfortunately not had at my • Phys. Zeit. No. 8, p. 235, Jan. 15 (1903); Phil. Mag., Feb. 1903.

.\'0111('

Rel11urks

Ull

Rudiuac:th·ity

577

disposal very active preparations of radium, and in consequence have had to adapt my methods to obtain effects from such active matter as I possessed. 1t was for this reason that I employed the electric method, which is capable of extreme refinement, and can be used to compare rapidly the intensities of radiations which would require very long exposures to act appreciably on the photographic plate. As an example of the comparative sensitiveness of the two methods I may recaIl some of M. Becquerel's own experiments (Comp. Rend. p. 209, 1902) in which he was unable to detect photographicaIly any action of the oe rays from uranium, or any evidence of the existence of very penetrating rays from 1t, although the times of exposure in two experiments were as long as 20 and 42 days. By the electric method the oe rays from I milligramme of uranium can be quickly measured; and I have recently found that, using 100 grs. ofuranium oxide, the existence of very penetrating rays can be detected in a few minutes, through 1 cm. of lead, by the increased rate at which the leaves of an electroscope fall together. The objection raised by M. Becquerel that the effects observed may have been due to secondary rays set up by the ß or 'cathode' rays striking the metal boundaries is in direct opposition to the data given in my paper. I have there shown that 89 per cent. of the discharge in the electroscope was due to oe rays, since the discharge was diminished by that amount by placing a thin layer of mica 0·01 cm. thick over the active material. It is weIl known that such a thickness of mica completely cuts off the oe rays, but allows the passage of the ß rays through it with but little absorption. The effect in the electroscope, due to the ß rays and the secondary rays set up by them, was thus only slightly altered by the addition of the mica plate, and therefore could not have been initially much more than 11 per cent. of the total. As a matter of fact, I showed that the 11 per cent. was hardly affected by an intense magnetic fie1d, and was due chiefly to the very penetrating rays from radium. With the uncovered active specimen the rate of discharge in very intense fields was reduced by 89 per cent. of the original, showing that a11 the oe rays were deviated in passing through the narrow slits. It has been shown in previous papers that, with a thin layer of active material, the ionization due to the ßrays (including that due to the secondary rays produced by them) is very smaIl compared with that due to the oe rays under ordinary experimental conditions. On the other hand, the ß rays are photographicaIly very active compared with the oe rays; and M. Becquerel, in several of his papers, has drawn attention to the marked photographic action of the secondary radiation set up by them. This may have led him to believe that their electrical effect would be equally marked, but such is not the case. M. P. Curie in his paper, after giving some experimental results showing that the decay of activity of the radium emanation is unaltered by variations in temperature between 4500 C. and -1800 C., proceeds as folIows:T

578

The Collected Papers

0/ Lord Rutherford

'Pour expliquer les phenomenes de lä. radioactivite induite et la transmission de l'activite par les courants des gaz, M. Rutherford a admis que le thorium et le radium emettent une emanation radioactive qui provoque la radioactivite des corps sur lesquels elle vient se fixer. C'est cette emanation qui entretient l'activire induite dans une enceinte fermee activee. M. Rutherford semble croire ä. la nature materielle de l'emanation et, dans run de ses Memoires les plus recents," il considere comme vraisemblable qu'il s'agit d'un gaz de la nature de ceux du groupe de l'argon. 'Je pense qu'il n'y a pas actuellement de raisons suffisantes pour admettre l'existence d'une emanation de matiere sous sa forme atomique ordinaire. Nous avons anterieurement, M. Debierne et moi, vainement cherche des raies nouvelles dans les gaz radioactifs extraits du radium. Enfin l'emanation disparait spontanement en tube scelle. Je considere aussi comme peu vraisemblable que les effets qui accompagnent l'existence de l'emanation aient leur origine dans une transformation chimique. On ne connait en effet aucune reaction chimique pour laquelle la vitesse de reaction soit independante de la temperature entre -180° et +450°'. Since the discovery of the thorium emanation, I have always taken the view that the emanation consists of matter in the radioactive state present in minute quantity in the surrounding gas. The experiments of Miss Brooks and myself ('Nature', p. 157, 1901) showed that the radium emanation mixed with air diffused very slowly. By comparing the rate of interdiffusion of the emanation into air with that of known gases, it was deduced that the emanation partic1es behaved like heavy gas moleeules of molecular weight probably lying between 40 and 100. I have long recognized that the electrical and other effects produced by the emanations can be manifested by an extremely minute quantity of radioactive matter in the gaseous state. For this reason I am not surprised that MM. Curie and Debierne have failed to obtain evidence by the spectroscope or balance of the existence of the emanation. At the same time I do not doubt that with sufficient quantity of active material the presence of the emanation will ultimately be detected by these means. I do not consider that the emanations remain permanently in the gaseous state, for it seems probable that the emanations gradually change into the matter responsible for excited activity, which is deposited on the walls of the containing vessel. Recent experiments by Mr. Soddy and myself show that the thorium emanation behaves chemically as an inert gas, and in this respect resembles the gases of the argon family. M. Curie has, apparently, not observed arecent paper by us (Proe. Chem. Soc. p. 219, 1902) in which it is shown that if the emanations of thorium or radium mixed with air, oxygen, or hydrogen are passed slowly through a spiral tube immersed in liquid air, the emanations are condensed in the tube, and the issuing gas is completely free from activity. On removing the spiral from the liquid air, the whole of the condensed emanation (allowing for the decay of activity in the interval) is released at a fairly definite temperature and appears .. Rutherford and Soddy, Phil. Mag., Nov. 1902.

Some Remm'ks on Radioacti\'ity

579

again in the stream of gas. A more detailed account of these investigations will shortly appear in this journal. These resuIts, in my opinion, conc1usively show that the emanations are gaseous in character, for it is very difficult to explain such phenomena except on a material hypothesis. In addition, I have recently shown that it is extremely probable that the greater proportion of the radiation from the emanation is material in nature, and consists of heavy charged bodies projected with great velocity, whose mass is of the same order as that of the hydrogen atom. In view of these results, which so strongly confirm the theory of the material nature of the emanation, the alternative theory proposed by M. P. Curie that the emanation consists of 'centres de condensation d'energie situes entre les moIecules du gaz et qui peuvent etre entraines avec lui', appears to me unnecessary. The interesting resuIt, obtained by M. Curie, of the exponentiallaw of decay of the radium emanation under all conditions, is only one of many others that have now been accumulated. I quite agree with M. Curie that such results cannot be satisfactorily explained on the laws of ordinary chemical change, but the difficulty disappears on the view already put forward by Mr. Soddy and myself (Phi!. Mag., Sept. and Nov. 1902) that the radioactivity of the elements is a manifestation of sub-atomic chemical change, and that the radiations accompany the change. There is no apriori reason to suppose that temperature would affect the rate of atomic disintegration; in fact the general experience of chemistry in failing to transform the elements is distinctly opposed to such a view. It is therefore not surprising that, if radioactivity is an accompaniment of subatomic change, the process should be independent of the ordinary chemical and physical agents at our disposal. These points, and many others bearing on the same question, are discussed in more detail in a joint paper with Mr. Soddy now in the course of publication. I am, Gentlemen, Yours very truly, E. RUTHERFORD McGill University, Montreal February 28, 1903

Condensation of the Radioactive Emanations by B.

RUTHBRFORD, M.A., D.se.,

Macdonald Professor of Physics,

McGill University, and F.

SODDY, M.A. (OXON)

From the Philosophical Magazine for May 1903, sero 6, v, pp. 561-76

IN a previous paper (Phi!. Mag., 1902, iv, p. 581 [po 505 of this volume]) we

have shown that the radioactive emanation from thorium passes in unchanged amount through a white-hot platinum tube and through a tube cooled to the temperature of solid carbon dioxide. The acquisition of a liquid air machine by the laboratory has enabled us to investigate the effect of lower temperature on the emanations from both thorium and radium. The result has been to show that both emanations condense at the temperature of liquid air, and possess sharply defined points of volatilization and condensation. If either emanation is conveyed by a slow stream of hydrogen, oxygen or air through a metal spiral immersed in liquid air, no trace of emanation escapes in the issuing gas. When the liquid air is removed and the spiral plunged into cotton-wool, several minutes elapse before any deflexion of the electrometer needle is observed, and then the condensed emanation volatilizes as a whole, and the movement of the electrometer needle is very sudden, especially in the case of radium. With a fairly large amount of radium emanation under the conditions mentioned, a very few seconds elapse after the first sign of movement before the electrometer needle indicates adeflexion of several hundred divisions of the scale per second. It is not necessary in either case that the emanating compound itself should be retained in the gas-stream. After the emanation is condensed in the spiral the thorium or radium compound may be removed and the gas stream sent directly into the spiral. But in the case of thorium under these conditions the effects observed are naturally small, owing to the very rapid loss of activity of the emanation with time, which experiment showed occurs at the same rate at the temperature of liquid air as at ordinary temperatures. As a matter of fact, in the case of radium the salt itself was seldom used. It was convenient to obtain the emanation from the solution and store it mixed with air in small gas-holders, the loss of activity during the course of a day's experiments being only a small part of the whole. If the radium emanation is condensed in a glass U-tube, the progress of the condensation can be followed by the eye by means of the fluorescence which the radiations excite in the glass. With a sufficiently slow gas stream

('ondensatiol1

(~l tlle

Radioactil'e Emanations

5!H

the fluorescencc is confined to the limb where the gas enters. If the ends of the tube are sealed and the temperature allowed to rise, the glow diffuses throughout the tube, and can be again concentrated at any point to some extent by application of a pad of cotton-wool soaked in liquid air. The U-tube can be made to impress its own image on a photographic plate through aluminium foil, and the MOTOR. impression is uniformly den se throughout the length of the tube. It LEADS TO LEADSTO retains its luminosity to a feeble MILLIVOLTMETER AMHETER extent after several days. The suddenness of the volatilization point of the condensed emanation is very remarkable considering 8 the minuteness of the actual amount of matter that must be involved. Arrangements were made to investigate the phenomena in an accurate quantitative manner. The method adopted was to condense the emanations in a spiral copper tube and to employ the latter as its own thermometer by determining its electrical resistance. For this purpose a constant current was maintained through the spiral, and the fall of potential between two fixed points on the spiral was determined by means of a Weston millivoltmeter. This method proved very reliable and convenient. A great number of preliminary experiments showed that to obtain accurate results two requirements must be satisfied. In the first place, since the temperature measured is the average temperature of the whole Fig. 1 spiral, the latter must be completely immersed in a bath of liquid kept weIl stirred. In the second place, in order to be sure that the spiral was not heated locally by the entering gas stream, it was necessary to subject the latter to a preliminary cooling to the desired temperature. This was accomplished by the apparatus represen ted in Fig. 1. The spiral and connecting tubes were made out of a continuous copper tube of length 310 cm., internal diameter 2 mm., and thickness of wall 0·32 mm. This was first wound into an inner spiral of 16 turns of mean diameter 1· 80 cm., which was soldered together into a

The Collected Papers 01 Lord Rutherford

582

compact cylinder. This effected the preliminary cooling of the gas stream. The tube was then wound back over the inner spiral into an outer spiral of 14 turns of mean diameter 2·90 cm. The turns of the outer spiral were separated from each other and from the inner spiral by air spaces. The outer spiral constituted the thermometer, and the potential leads were soldered on to the top and bottom and enc10sed in glass tubes. The system was supported inside a glass cylinder closed at the bottom, of height 41 cm. and diameter 3·5 CDl., by means of an airtight rubber cork fitting the open end, through which the ends of the spiral and the leads passed. A small stirrer at the bottom of the tube was driven by a central rod passing through a suitable bearing supported by the inner spiral, and through a glass tube /H'ILI./VOLTMETER

AMMETER.

RES

T

A CaCl2

E

B TU EARTH

Fig. 2 in the centre of the cork. This rod was operated by an electric motor supported above the cork. The spiral was kept from actual contact with the glass by a sheet of mica perforated into holes, and was given sufficient rigidity by means of an ebonite ring fitting tightly into the space between the upper part of the inside and outside spiral. A current was sent (Fig. 2) through the spiral by means of leads soldered above the cork from a storage battery, passing through a Weston ammeter and sliding resistance. By means of the latter the current was kept always constant at 0·900 ampere. The potential leads were connected with a Weston millivoltmeter which registered at ordinary temperature adeflexion of about 6 millivolts, or 60 divisions on the scale, at the temperature of liquid air rather less than 2 millivolts. This corresponds to a resistance at ordinary temperature of about 0·01 ohm, and it can readily be shown that the heating effect of the current in the spiral is negligible. No substance seems to be known which is liquid at the ordinary temperature and remains liquid at the temperature of condensation of the

Condensation

0/ the Radioactil'e Emanations

583

radium emanation (ab out - I 50°). Ethyl chloride most nearly fulfils this condition, but solidifies in the neighbourhood of -140°C. A bath of this substance, however. proved useful in one series of measurements with the thorium emanation. The rest of the determinations were carried out in a bath of liquefied ethylene. This boils at -103.5° and freezes at -169°, and gave just the range of temperature desired in these experiments. About 70 litres of the gas, purified by fractional distillation, was ordinarily used. This was sent into the apparatus by the tube A (Fig. 1), escaping by the tube carrying the stirring rod. The liquefied ethylene always covered the top of the spiral to the depth of several centimetres. The apparatus was surrounded by a tall copper cylinder well covered with lagging which contained the liquid air. Copper was preferable to glass for the purpose, for it ensured in the actual determinations a more uniform supply of heat to the apparatus.

Ca/ibration

0/ the Copper Thermometer

The readings of the voltmeter described were determined at the following temperatures: 100°, 0°, the boiling point of ethylene -103.5°, the freezing point -169°, and the temperature of liquid air. The ethylene employed was carefully fractionated for this purpose. 120 litres were condensed and the first and last 20 litres rejected, the determinations being made with the middle fraction. The temperature of liquid air is a variable, depending on the composition of the liquid, but if the latter is known the temperature can be fixed with great accuracy from the tables given by Baly. A sampie was therefore drawn off into agas-holder from beneath the surface of the liquid, at the time the temperature was read, and its composition determined by analysis. These constants were frequently redetermined throughout the course of the experiments and found to remain unaltered. The following table represents the results: Temperature

oe

Resistance (ohms)

Ratio

100 0 -103·5 -169 -192·2

0·00947 0·00701 0·00437 0·00262 0·00202

135·1 100 62·3 37·4 28·8

In the last column the ratio of the resistance is given, the value at 0° being taken as 100. In Fig. 3 the results are plotted with the resistance as ordinates and the temperature as abscissae. It will be observed that the curve is very nearly a straight line cutting the axis, if produced, at very nearly the absolute zero. For the particular thermometer used, therefore, the readings of the millivoltmeter may be taken without appreciable error to be proportional to the absolute temperature. The instruments employed were accurately

584

Tile Collected Papers

~f Lord

Rutile/lord

calibrated. The accuracy of the temperature determination by this method depends only on the sensitiveness of the millivoltmeter. At the temperature at which most of the observations were made. one division of the scale corresponded to about 4('\C, and the readings could be made to ]/10 of a division. The determinations were therefore accurate to within O· 5°, 40

100

~VD 40

~ ~

~~

40

~

~40

~~

;::;;

~20 ~

o

o

40

SO

120 160 200 TEMPRRArURl (CFNr. J

240

290

Fig. 3* which was sufficient for the purpose. The great advantage of the method is the ease and certainty with which a continually changing temperature can be followed. Experiments for the Radium Emanation with a Steady Current of Gas

Experiments with the radium emanation are much simpler than for that of thorium, since the activity does not decay appreciably over the time required for a complete series of observations, and much larger effects can be obtained. Fig. 2 represents the general arrangement of the apparatus for the determinabon of the volatilization temperature of the radium emanation m a steady current of gas. The latter (either hydrogen or oxygen) was conveniently obtained from a set of eight voltameters, arranged in series across the 110 volt circuit and capable of taking a current up to 3 amperes. The latter, measured bya Weston ammeter, furnished a measure of the number of cubic centimetres of gas passing per second. This enters the apparatus at A. The radium emanation mixed with air is stored in the gas-holder B. The exit of the copper spiral is connected with a testing cylinder T of the kind previously described, in which the ionization current through the gas, due ... Temperature in _oC. [Ed.]

('olldensatioll

0/ the Radioactive Emanations

585

to the rays from the emanation, is measured by the electrometer E. D is a drying tube. The ethylene bath was cooled by liquid air slightly below the temperature of condensation, and adefinite volume of the emanation was sent into the apparatus from B, and conveyed by a very slow gas stream for 10 minutes into the spiral so that it was condensed near the beginning of the spiral. The current of gas to be employed is then adjusted and the temperature of the bath allowed to rise slowly. One observer took the determinations of the temperature and the lapse of time, the other the readings of the electrometer. The temperature at which the electrometer needle commenced to move was recorded, and the rate of movement at succeeding temperatures. In most of the experiments the temperature rose at a rate of about 1· 6° to 2° per minute. The following table gives an illustration of the results obtained: Temperature

oe

Divisions per second of the electrometer

o

-160 -156 -154,3 -153,8 -152,5

o 1

21 24

The observed temperature of the first appearance of the emanation is subject to a correction for the time taken for the emanation to be swept into the testing vessel after volatilization. This depends on the volume of the spiral and its temperature, and on the current of gas, and can be approximately calculated for each experiment. Knowing the rate ofrise oftemperature in the experiment the actual temperature at which volatilization commenced could be deduced. In the table given above for agas stream of 1 . 38 cm. per second the application of this correction lowers the temperature of volatilization about o· 8°C. The following table inc1udes some of the corrected results for the radium emanation. Under the column Tl the temperatures at which the emanation began to volatilize are given, under T2 the temperatures at which one half of the total amount had been given off. Gas

c.c. per second

Hydrogen

0·25 0·32 0·33 0·92 1·38 2·3 0·34 0·33 0·58

Oxygen T*

Tl

-151,3 -153,7 -153,7 -152 -154 -162,5 -152,5 -152,5 -155

TI

-150 -151 -151 -151 -153 -162 -151,5 -151,5 -153

586

The Collected Papers ofLord Rutherford

With hydrogen streams from O· 25 c.c. to 1· 38 c.c. per second and for the slow stream of oxygen the results are in good agreement and give a mean value for Tl of -153°, for T2 -151'5°. But a well-marked difference appears when the stream of hydrogen is increased to 2· 3 c.c. per second. This corresponds to an initial velocity of 50 cm. a second through the spiral, and 20 cm. per second after the temperature of the bath has been obtained. The result is therefore to be expected, for in such a rapid stream the gas is not cooled down to the temperature of the spiral and the volatilization point of the emanation is, in consequence, apparently lower. For the same reason the temperature observed for the oxygen stream of only O· 58 c.c. a second is probably too low, for this gas is cooled with more difficulty than hydrogen. Even if the temperature of the spiral is ultimate1y attained, a rapid current of gas would tend to sweep out the emanation it had volatilized in its passage, without giving it time to be recondensed in the subsequent portions of the spiral. This effect, for reasons to be discussed later, would also be greater in oxygen than in hydrogen. Further determinations for the radium emanation by another distinct method are given later in the paper. At this stage some experiments may be mentioned that were performed with a much larger quantity of radium emanation to determine the amount that is volatilized at various temperatures. In one experiment at -154°, no escape of emanation was observed although less than 1/10000 part could have been detected. At - 152° about one half per cent and at -1500 considerably more than half of the total amount had come off. There is no doubt that in a bath, kept constant at the temperature of initial volatilization, all would volatilize if sufficient time were allowed, but it is probable that the time required would be considerable. There is, in fact, evidence that the condensed radium emanation possesses what corresponds to a vapour pressure in an ordinary substance. Experimentsfor the Thorium Emanation by the same Method

The rapid loss of the activity of the thorium emanation, which decays to half value in one minute, makes the determination of its volatilization point a more difficult task. In the first place, too slow agas stream cannot be employed or the ionization effects are too small. In the second place, the thorium compound must be retained all the time in the gas stream, unless the temperature of the spiral is made to rise so rapidly that its determination becomes impracticable. The results, therefore, bear a different interpretation from those given for radium, for in this case the point measured is the temperature at which some of the thorium emanation first escaped condensation, and not the point at which volatilization begins of that already condensed. In place of the T-tube and gas-holder of Fig. 2, a highly emanating thorium compound was placed in series with the gas stream, and the temperature T observed at which the emanation first began to make

londellsation

0/ the Radioactive Emanations

587

its appearancc in the testing vessel. The following are some of the results obtained: Gas

c.c. per second

Hydrogen

0·71 1·38 0·58 0·58 0·58

Oxygen

T

oe -ISS -159 -155·5 -156 -155·5

If these results are compared with those obtained for the radium emanation it will be seen that the temperatures with an equal gas stream in the two cases are very nearly the same. Thus, in the determinations in an oxygen stream of O· 58 c.c. per second, the radium emanation commenced to volatilize at -155°, and some of the thorium emanation escapes condensation at -155·5°C. It was at first thought that this result indicated that the condensation points of the two emanations were identical. It will be recalled t,llat no difference could be detected in the chemical properties of the two emanations, both being quite unaffected by the most powerful chemical reagents. Yet it seemed improbable that the two emanations could be materially identical on account of the completely distinct character of their radioactive properties. Not only are their rates of decay widely different, being 5000 times faster in the one case than the other, but the excited activities they give rise to are also completely different, not only in the rate of decay, but even in the number of changes through which they apparently pass before their activity disappears. In the course of further work a very distinct difference of behaviour in the condensation phenomena in the two cases was brought to light. It was observed that some of the thorium emanation was condensed at temperatures as much as 30° above the point of complete condensation. The curve, Fig. 5 is an example of the results obtained. The maximum ordinate taken as 100 represents the amount of emanation entering the testing vessel at temperatures far above the point at which condensation commences. This amount begins to decrease at about -120°, and becomes less than 1 per cent of its original value at -154°. This curve was obtained in a steady stream of oxygen of O· 38 c.c. per second. For faster currents the curve is displaced slightly to the right. When tested under the same conditions, the radium emanation showed no such behaviour. When the solution of radium chloride is retained in the gas stream, the temperature at which some of the emanation escapes condensation is only slightly above the point before found at which the condensed emanation commences to volatilize. Apart from the question of the actual condensation points themselves, there is, therefore, a well-marked distinction in the character of the phenomena in the two cases. To investigate this difference, a new method was devised which allowed determinations to be made with the two emanations under comparable conditions.

588

The Collected Papers oJ Lord RutherJorel Experiments by the Static M ethod

The use of a steady stream of gas through the spiral in which the emanation is condensed has many disadvantages, some of which have heen alluded to. By the use of a mercury pump, and by working in a partial vacuum, these disadvantages are avoided, and the conditions of experiments made more definite. Fig. 4 represents the arrangement employed. AMMETER

II#II.I.IV()I.TMETER

K

TfJ TESTING

VEJjEL

H

c F

E

A THORIUM

p

Ox/:DE

s Fig.4 100

~

~to

~

~60 ~,

ls

~40

I

~ ~

~ -110

l... ~

~-110

- 100

-110

-120

-130

·140

TEMPE~ATVltE (CGNT.)

Fig. 5

-150

-160

('olldensation

0/ the Radioactive Emanations

589

The Geissler pump P was connected with the copper spiral Sand the thorium compound in the tube A, and possessed a volume large in comparison with the volume of the whole of the rest of the apparatus. The small bulb V was first filled with hydrogen or oxygen entering at C, and the thorium tube-spiral and connecting tubes exhausted by the pump to apressure of a few millimetres of mercury. The three-way tap was then reversed, the tap E being open and F closed. In this way a quantity of the emanation was swept out of A into the spiral; E was c1osed, and the emanation allowed to remain in the spiral a definite time. This period varied in different experiments from 10 to 90 sec. At the expiration of this interval, the pump in whieh the mercury had been lowered was put into communieation with the spiral by the tap F, which was then c1osed, the mercury raised, and the emanation expelled into Hand carried on to the testing vessel by a steady eurrent of oxygen entering the tube at K, and kept eontinuously passing throughout the experiment. The pressures employed were deduced from the readings of the height of the mereury in the pump tube when the mereury was lowered, and the relative volumes of various parts of the apparatus. The various manipulations of the taps of the mereury pump were all timed throughout by a stop-watch, and the observations with the eleetrometer were always taken at the same interval after the commencement of the operations. In this way the decay of aetivity of the emanation was the same in eaeh experiment, and the results obtained in different experiments comparable with one another. One observer took charge of the manipulation of the apparatus, the other recorded the lapse of time, the temperature of the spiral at the instant the contents were drawn into the pump by opening the tap F, and the readings of the electrometer after the emanation had been sent into the testing-vessel. The latter eould be taken within 90 sec. from the time the emanation was removed from the thorium in those cases where it remained in the spiral for aperiod of 30 sec. The tap F was a three-way, so that the pump could be cut out of the circuit and experiments in a steady stream of gas carried on without alteration of the apparatus. The amounts of the emanation that remain uncondensed at different temperatures are shown graphically in Fig. 6. The different curves represent different series of observations with hydrogen and oxygen respectively, in which the time during which the emanation remained in the spiral was in some eases 30 sec., in others 90 sec. Curves A and B illustrate the difference in the condensation curves for hydrogen and oxygen under similar conditions. The press ure in the spiral after the removal of the uncondensed emanation corresponded to 19 mm. of mercury. The curves show that a greater proportion of the emanation is condensed for the same temperature in hydrogen than in oxygen. The curves C and D, which were obtained under different conditions of press ure and amount of emanation from curves A and B, show that a greater proportion is condensed in 90 sec. than in 30 sec. The proportion eondensed in a steady stream is less than in any of the experiments by the statie method. In all cases, however, eondensation commences at about the same

590

The Collected Papers

0/ Lord Rutherford

temperature, viz., -120°C, and there is no doubt that this amount must be taken as the real condensation point of the thorium emanation, and that the identity in the temperatures observed in the earlier experiments with a steady stream must be regarded as purely accidental. c:)

~'

~ ~

c:::i ~DU -110

~'It ~

-110 -110

'It 'It-110

~ 'It

~

"",40 ~

)0.",

~ ~

~2t

~

"",40

-110

-110

- /20

I

.. ISO

I

-140

TtWE~ATIIRE(CENT.J

I

-/50

Fig.6 Experiments with the Radium Emanation

The apparatus (Fig. 4) was slightly altered for the determination of the volatilization point of the radium emanation. The thorium tube was replaced by a drying tube of calcium chloride and the bulb V filled with air mixed with the radium emanation which was then allowed into the exhausted spiral kept below the temperature of condensation. The apparatus was then repeatedly exhausted by the pump, and after each exhaustion a bulb fuH of oxygen was sent into the spiral, the temperature of the bath being allowed to rise slowly during the exhaustion. The temperature at which the first trace of emanation escaped and the amounts at succeeding temperatures were noted as before. The following table is an example of the results obtained: Temperature

oe

-153 -151 -148,5 -146,5 -143 -139 -135

Divisions of electrometer per second

o o

0·74

5·3 5·1

0·7 0·08

Crmdensatioll of Ihe Radioaclire Emanations

591

The mean of several results gave -150° as the point at which the emanation first began to volatilize, and this is in good agreement with the result by the blowing method, that is, -153°. The difference is in the expected direction, for in the static method the mass of the gas employed is much smaller, and any emanation that is volatilized by the rush of heated gas in its passage through the spiral has time to be recondensed. The temperature -1500 e may therefore be taken with considerable confidence as being the true point at which the radium emanation first commences to volatilize. On the other hand, the table shows a somewhat less sudden volatilization than in the case of the blowing method, but this is inherent to the static method employed. The glass spiral connecting tube between the pump and the copper spiral had a greater volume than the latter itself, and at each exhaustion some of the volatilized emanation is left in this spiral. In the case of the thorium emanation this decays practically to zero before the next observation is taken, but in the case of the radium emanation it does not, and is added to the amount removed at the next exhaustion. The temperature of volatilization found in these experiments has almost exactly the value given by Ramsay and Travers for the boiling point of nitric oxide under atmospheric press ure -149' 9°e. A bath of liquefied nitric oxide was prepared, and used in place of the ethylene in former experiments. Its boiling point rose steadily until about one-fifth had boiled off. It then became constant at -151°e, as determined by the resistance of the cop per spiral, and remained so until the latter began to be no longer completely covered. The nitric oxide was not sufficiently pure to enable much weight to be put on this result as adetermination of temperature, for from its behaviour it obviously must have contained a considerable quantity of dissolved nitrogen, but it is of interest as being a completely independent check on the thermometer at almost the exact point of condensation, and shows that the value ascribed to the latter cannot be far from the truth. Such a bath of boiling liquid rising very slowly in temperature over the exact range in which volatilization takes place, afforded a means, however, of examining more exactly the progress of the volatilization after the initial point was reached. The latter occurred in this case at -155°C, a steady current of air being maintained through the spiral. In 4 minutes the temperature had increased to -153' 5°, and the amount volatilized was about four times as great as at -153°. In the next 5-!- minutes the temperature had increased to -152' 3°, and the whole amount practically had volatilized, which was at least fifty times the amount at the temperature of -153' 5°. Such a result would of course be explained by slight local inequalities in the temperature of the spiral, but since the latter was immersed in a rapidly boiling liquid it is difficult to believe that such could have been the case. It seems more reasonable to attribute it to a true vapour press ure possessed by the condensed emanation, although great refinement in the experimental methods would be necessary before such a conc1usion could be considered established.

592

The Collected Papers

0/ Lord Ru/her/ord

The Explanation o[ the Anomalous Behaviour o[ the Thorium Emanations

The results are satisfactory in so far as they show that the two emanations do not condense at the same temperature. The anomalous behaviour of the thorium emanation in condensation, which in the first experiments seemed to indicate that the two emanations condensed at the same temperature, has been shown to be due to an effect, not present in the case of the radium emanation, which depends on the nature of the gas, the concentration of the emanation, and the time that the latter has been left to condense. It remains to develop a view which gives a satisfactory explanation of this behaviour. In the first place the actual number of particles of emanation that are present must be almost infinitesimally small compared with the number of particles of the gas with which they are mixed. It is difficult to make an accurate calculation of the number actua1ly present, but an estimate of the order of the number present can be deduced from the following con~ siderations. The radiation from the thorium and radium emanations con~ sists, as far as it is known, entirely of oe rays. The view has recently been put forward that these, in the case of radium, consist of projected particles of the same order of mass as the hydrogen atom, carrying a positive charge, and travelling with a velocity about 1/10 the velocIty of light. It is extremely like1y that the radiations from the two emanations are quite comparable in character, and produce in their passage through the gas a similar number of ions. From results deduced from experiments made in an attempt to measure the charge carried by the oe rays there is no doubt that each of these projected partic1es produces at least 104, and possibly 106, ions in its path before being absorbed in the gas. For the present purpose, lOS ions will be taken as a probable value. The electrometer employed readily measured currents of 1O-3E.S. units per second. Taking the charge on an ion as 6 X 10- 10 E.S. units, this corresponds to a production of 1·7 X 106 ions per second, which would be produced by 17 expelled 'rays' per second. Each radiating particle cannot expelless than one ray, and may expel more, but it is likely that the number of rays expelled by a particle of the thorium emanation is not greatly different from the number expelled by a partic1e of the radium emanation. The view will be developed more generally in a sub~ sequent paper, * that the decay of activity of a radioactive substance is caused by the number of particles present diminishing owing to their changing into new systems, the change being accompanied by the expulsion of rays. From the law of rate of decay I t = Ioe- At, on this view AN particles change per second when N are present. Therefore to produce 17 rays per second AN cannot be greater than 17. Since in the case of the thorium emanation A equals 1/87, it follows that N cannot be greater than 1500. The electrometer used therefore detects the presence of about 1500 partic1es of the thorium emanation and since, in the static method, the volume of the condensing

* Phi!. Mag., sero 6, vol. v, p. 576.

('olldellsatiol/ 01' Ihe Radioaeti!'e Emal/ations

593

spiral was a bout 15 C.C., this corresponds to a concentration of about 100 particles per cubic centimetre. An ordinary gas at atmospheric pressure probably contains 1020 particles. On this estimate, therefore, the thorium emanation could have been detected jf it possessed a partial press ure in the condensing spiral of 10- 18 atmosphere. It is thus not surprising that the condensation point of the thorium emanation is not sharply defined. It is rather a matter of remark that condensation can occur at aIl with such sparse distribution of emanation partic1es in the gas, for in order for condensation to take place the partic1es must first approach within each other's sphere of influence. Now consider the case of the radium emanation, The rate of decay is about 5000 times slower than that of the thorium emanation and, consequently, the actual number that must be present to produce the same number of rays per second in the two cases must be of the order of 5000 times greater for the radium than for the thorium emanation. This conc1usion involves only the assumptions that the same number of rays are produced by a partic1e of emanation in each case, and that the rays expelled produce in the passage through the gas the same number of ions. The number of partic1es that must be present for the electrometer to detect them in this experiment must therefore be about 5000 x 1500, i.e. of the order of 10'. The difference of behaviour in the two cases is weIl explained in the view that for equal electrical effects the number of radium emanation partic1es must be far larger than the number of thorium emanation partic1es. It is to be expected that the probability of the particles coming into each other's sphere of influence will increase very rapidly as the concentration of the partic1es increases, and that in the case of the radium emanation, once the temperature of condensation is attained aIl but a negligibly smaIl proportion of the total number of particles present will condense in a very short time. In the case of the thorium emanation, however, the temperature might be far below that of condensation, and yet a considerable proportion remain uncondensed for comparatively long intervals. On this view the experimental results obtained are exactly accounted for. A greater proportion condenses the longer the time allowed for condensation under the same conditions. The condensation occurs more rapidly in hydrogen than in oxygen, owing to the diffusion being greater in the former gas. For the same reason the condensation occurs faster the lower the pressure of the gas present. FinaIly, when the emanation is carried by a steady gas stream, a less proportion condenses than in the other cases, because the concentration of emanation particles per cubic centimetre of gas is less under these conditions.

Decay 0/ Activity 0/ the Condensed Emanation Some experiments were made to test whether the rate of decay of activity of the thorium emanation was altered at the temperature of liquid air. The

594

The Collecled Papers 0/ Lord RU/her/ord

method employed was to pass a slow steady stream of the thorium emanation, mixed with hydrogen or oxygen, into the copper spiral immersed in liquid air for aspace of 5 min. The emanation was thus condensed in the spiral. The current of gas was then stopped, and after definite intervals, extending in successive experiments from 1 to 5 min., the spiral was rapidly removed from the liquid air, plunged into hot water, and the emanation present swept rapidly with a current of air into a large testing vessel. The results showed that the emanation lost its activity at the same rate at the temperature of liquid air as at ordinary temperature, i.e. its activity fell to half value in about 1 min. This is in agreement with results previously noted for other active products, showing that the rate of decay is unaffected by any physical or chemical agency.

Summary 0/ Results The results show that the thorium emanation begins to condense at about -120°C. The rapid rate of decay of its activity renders adetermination of the point at which the condensed emanation commences to volatilize experimentally impracticable. But under all conditions tried some of the emanation escapes condensation at temperatures much below the temperature of initial condensation. In a slow stream of gas the presence of the emanation is first observed at about -155°C. It is probable that -120° represents the true temperature of volatilization and condensation, and that the escape of emanation below this temperature is due to the extremely small number of condensing particles present. The radium emanation commences to volatilize at -153° in a steady stream of gas, and at -150° in a stationary atmosphere, and this latter value may be accepted with considerable confidence as being the true temperature of initial volatilization. In the case of radium there is no sensible difference between the temperature of volatilization and of condensation, and the whole of the emanation is condensed at temperatures only slightly below the initial point of volatilization. This difference of behaviour of the two emanations is explained on the view that the number of partic1es of emanation present for equal effects is probably many thousand times greater in the case of radium emanation than in the case of thorium emanation. All the radium emanation is volatilized within a very few degrees of the initial point, the rate of volatilization of course depending on the rate of rise of temperature. But with a very slowly rising temperature, practica1ly all of the emanation comes off very suddenly at a temperature not much more than one degree above that at which only 2 per cent has volatilized. The general indication of a1l the experiments, considered together, is to show that the condensed emanation possesses a true vapour pressure, and that the emanation commences to volatilize slowly two or three degrees below the temperature of rapid volatilization even when the process occurs in a stationary atmosphere. The emanations therefore possess the usual

(. 'olldensatioll

0/ file

Radioactil'e Emallatiolls

595

properties possessed by ordinary gaseous matter, in so rar as the phenomena of volatilization and condensation are concerned. It was shown in a recent paper that they also possess the property possessed by gases of being occ1uded by solids under certain conditions. These new properties, taken in conjunction with the earlier discovered diffusion phenomena, characteristic of the radioactive emanations, leave no doubt that the latter must consist of matter in the gaseous state. McGill University. Montreal March 9, 1903

Radioactive Change by E.

R UTHERFORD, M.A., D.se.,

McGill University, and

Macdonald Professor of Physics, F. SODDY, M.A. (OXON)

From the Philosophical Magazine for May 1903, sero 6, v, pp. 576-91

CoNTENTS

I. Tbe Products of Radioactive Change, and their Specific Material Nature. Tbe Synchronism between the Change and the Radiation. III. Tbe Material Nature of the Radiations. IV. Tbe Law of Radioactive Change. V. The Conservation of Radioactivity. VI. The Relation of Radioactive Change to Chemical Change. VII. The Energy of Radioactive Change and the Internal Energy of the Chemical Atom.

n.

§ I. The Products 01 Radioactive Change and their

Specijic Material Nature

IN previous papers it has been shown that the radioactivity of the elements radium, thorium, and uranium is maintained by the continuous production of new kinds of matter which possess temporary activity. In some cases the new product exhibits well-defined chemical differences from the element producing it, and can be separated by chemical processes. Examples of this are to be found in the removal of thorium X from thorium and uranium X from uranium. In other cases the new products are gaseous in character, and so separate themselves by the mere process of diffusion, giving rise to the radioactive emanations which are produced by compounds of thorium and radium. These emanations can be condensed by cold and again volatilized; although they do not appear to possess positive chemical affinities, they are frequently occ1uded by the substances producing them when in the solid state, and are liberated by solution; they diffuse rapidly into the atmosphere and through porous partitions, and in general exhibit the behaviour of inert gases of fairly high molecular weight. In other cases again the new matter is itself non-volatile, but is produced by the further change of the gaseous emanation; so that the latter acts as the intermediary in the process of its separation from the radioactive element. This is the case with the two different kinds of excited activity produced on objects in the neighbourhood of compounds of thorium and radium respectively, which in turn possess well-defined and characteristic material properties. For example, the thoriumexcited activity is volatilized at a definite high temperature, and redeposited

RadioQ('til'e Changt'

597

in the neighbourhood, and can be dissolved in some reagents and not in others. These various new bodies differ from ordinary matter, therefore, only in one point, namely, that their quantity is far below the limit that can be reached by the ordinary methods of chemical and spectroscopic analysis. As an example that this is no argument against their specific material existence, it may be mentioned that the same is true of radium itself as it occurs in nature. No chemical or spectroscopic test is sufficiently delicate to detect radium in pitchblende, and it is not until the quantity present is increased many times by concentration that the characteristic spectrum begins to make its appearance. Mme Curie and also Giesel have succeeded in obtaining quite considerable quantities of pure radium compounds by working up many tons of pitchblende, and the results go to show that radium is in reality one of the best defined and most characteristic ofthe chemical elements. So, also, the various new bodies, whose existence has been discovered by the aid of their radioactivity, would no doubt, like radium, be brought within the range of the older methods of investigation if it were possible to increase the quantity of material employed indefinitely. § 2. The Synchronism between the Change and the Radiation

In the present paper the nature of the changes in which these new bodies are produced remains to be considered. The experimental evidence that has been accumulated is now sufficiently complete to enable a general theory of the nature of the process to be established with a considerable degree of certainty and definiteness. It so on became apparent from this evidence that a much more intimate connexion exists between the radioactivity and the changes that maintain it than is expressed in the idea of the production of active matter. It will be recalled that all cases of radioactive change that have been studied can be resolved into the production by one substance of one other (disregarding for the present the expelled rays). When several changes occur together these are not simultaneous but successive. Thus thorium produces thorium X, the thorium X produces the thorium emanation, and the latter produces the excited activity. Now the radioactivity of each of these substances can be shown to be connected, not with the change in which it was itself produced, but with the change in which it in turn pro duces the next new type. Thus after thorium X has been separated from the thorium producing it, the radiations of the thorium X are proportional to the amount of emanation that it produces, and both the radioactivity and the emanating power of thorium X decay according to the same law and at the same rate. In the next stage the emanation goes on to produce the excited activity. The activity ofthe emanation falls to half-value in 1 min., and the amount of excited activity produced by it on the negative electrode in an electric field falls off in like ratio. These results are fully borne out in the case of radium. The activity of the radium emanation decays to

598

Tlre Collected Papers o[ Lord Ruther[ord

half-value in 4 days, and so also does its power of producing the excited activity. Hence it is not possible to regard radioactivity as a consequence of changes that have already taken place. The rays emitted must be an accompaniment of the change of the radiating system into the one next produced. Non-separable acli"vity. This point ofview at once accounts for the existence of a constant radioactivity, non-separable by chemical processes, in each of the three radio-elements. This non-separable activity consists of the radiations that accompany the primary change of the radio-element itself into the first new product that is produced. Thus in thorium about 25 per cent of the er; radiation accompanies the first change of the thorium into thorium X. In uranium the whole of the er; radiation is non-separable and accompanies the change of the uranium into uranium X. Several important consequences follow from the conc1usion that the radiations accompany the change. A body that is radioactive must ipso [acto be changing, and hence it is not possible that any of the new types of radioactive matter-e.g. uranium X, thorium X, the two emanations, etc.--can be identical with any of the known elements. For they remain in existence only a short time, and the decay of their radioactivity is the expression of their continuously diminishing quantity. On the other hand, since the ultimate products of the changes cannot be radioactive, there must always exist at least one stage in the process beyond the range of the methods of experiment. For this reason the ultimate products that result from the changes remain unknown, the quantities involved being unrecognizable, except by the methods of radioactivity. In the naturally occurring minerals containing the radio-elements these changes must have been proceeding steadily over very long periods, and, unless they succeed in escaping, the ultimate products should have accumulated in sufficient quantity to be detected, and therefore should appear in nature as the invariable companions of the radio-elements. We have already suggested on these and other grounds that possibly helium may be such an ultimate product, although, of course, the suggestion is at present a purely speculative one. But a closer study of the radioactive minerals would in all prob ability afford further evidence on this important question. § 3. The Material Nature o[ the Radiations

The view that the ray or rays from any system are produced at the moment the system changes has received strong confirmation by the discovery of the electric and magnetic deviability of the er; ray. The deviation is in the opposite sense to the ß or cathode ray, and the rays thus consist of positively charged bodies projected with great velocity (Rutherford, Phil. Mag., February, 1903). The latter was shown to be of the order of 2· 5 X 1()9 cm. per second. The value of e/m, the ratio of the charge of the carrier to its mass, is of the order of 6 X 103• Now the value of e/m for thecathoderayis about 107• Assumingthat the value of the charge is the same in each case, the apparent mass of the

Radioactil'e Change

599

positive projected particle is over 1,000 times as great as for the cathode ray. Now e/m = lQ4 for the hydrogen atom in the electrolysis of water. The particle that constitutes the IX ray thus behaves as if its mass were of the same order as that of the hydrogen atom. The IX rays from all the radioelements, and from the various radioactive bodies which they produce, possess analogous properties, and differ only to a slight extent in penetrating power. There are thus strong reasons for the belief that the IX rays generally are projections and that the mass of the particle is of the same order as that of the hydrogen atom, and very large compared with the mass of the projected particle which constitutes the ß or easily deviable ray from the same element. With regard to the part played in radioactivity by the two types of radiation, there can be no doubt that the IX rays are by far the more important. In all cases they represent over 99 per cent of the energy radiated, * and although the ßrays on account of their penetrating power and marked photographie action have been more often studied, they are comparatively of much less significance. It has been shown that the non-separable activity of all three radioelements, the activity of the two emanations, and the first stage of the excited activity of radium. comprise only IX rays. It is not until the processes near completion in so rar as their progress ean be experimentally traced that the ß or cathode ray makes its appearance. t In light of this evidence there is every reason to suppose, not merely that the expulsion of a charged particle accompanies the change, but that this expulsion actually is the change. § 4. The Law

0/ Radioactive Change

The view that the radiation from an active substance accompanies the change gives a very definite physical meaning to the law of decay of radioactivity. In all cases where one of the radioactive products has been separated, and its activity examined independently of the active substance which gives rise to it, or which it in turn produces, it has been found that the activity under all conditions investigated falls off in a geometrical progression with the time. This is expressed by the equation

Ir

10

= e-Ät

where 10 is the initial ionization current due to the radiations, Ir that after • In the paper in which this is deduced (Phil. Mag., September 1902, p. 329) there is an obvious slip of calculation. The number should be 100 instead of 1000. (Corrected in this volume, p. 470. [Ed.]). t In addition to the a: and ß rays the radio-elements also give out a third type of radiation which is extremely penetrating. Thorium as well as radium (Rutherford, Phys. Zeit., 1902), gives out these penetrating rays, and it has since been found that uranium possesses the same property. These rays have not yet been sufficiently examined to make any discllssion possible of the part they play in radioactive processes.

600

111e CoJ/ected Papers

01 Lord RutherJord

the time t, and >. is a constant. Each ray or projected particle will, in general, produce a certain definite number of ions in its path, and the ionization current is therefore proportional to the number of such particles projected per second. Thus nt -"AI -=e ,

no where nt is the number projected in unit of time for the time t and no the number initially. If each changing system gives rise to one ray, the number of systems Nt which remain unchanged at the time t is given by N,

=

f

CXl

I

n,.dt

no

=

Te-At.

The number No initially present is given by putting t = O.

no

N o =);

and

Nt

--e-"AI

No -

.

The same law holds if each changing system produces two or any definite number of rays. Differentiating dN dt = ->'N" or, the rate of change of the system at any time is always proportional to the amount remaining unchanged. The law of radioactive change may therefore be expressed in the one statement-the proportional amount of radioactive matter that changes in unit time is a constant. When the total amount does not vary (a condition nearly fulfilled at the equilibrium point where the rate of supply is equal to the rate of change) the proportion of the whole which changes in unit time is represented by the constant >., which possesses for each type of active matter a fixed and characteristic value. >. may therefore be suitably called the 'radioactive constant'. The complexity of the phenomena of radioactivity is due to the existence as a general rule of several different types of matter changing at the same time into one another, each type possessing a different radioactive constant. § 5. The Conservation

0/ Radioactivity

The law of radioactive change that has been deduced holds for each stage that has been examined, and therefore holds for the phenomenon generally. The radioactive constant >. has been investigated under very widely varied

RadioaClh'(' Change

601

conditions of temperature, and under the influence of the most powerful chemical and physical agencies, and no alternation of its value has been observed. The law forms in fact the mathematical expression of a general principle to which we havc been led as the result of our investigations as a whole. Radioactivity, according to present knowledge, must be regarded as the result of a process which lies wholly outside the sphere of known controllable forces, and cannot be created, altered, or destroyed. Like gravitation, it is proportional only to the quantity of matter involved, and in this restricted sense it is therefore true to speak of the principle as the conservation of radioactivity. * Radioactivity differs of course from gravitation in being a special and not necessarily a uni versal property of matter. which is possessed by different kinds in widely different degree. In the processes of radioactivity these different kinds change into one another and into inactive matter, producing corresponding changes in the radioactivity. Thus the decay of radioactivity is to be ascribed to the disappearance of the active matter, and the recovery of radioactivity to its production. When the two processes balance-a condition very nearly fulfilled in the case of the radio-elements in a c10sed space-the activity remains constant. But here the apparent constancy is merely the expression of the slow rate of change of the radio-element itself. Over sufficiently long periods its radioactivity must also decay according to the law of radioactive change, for otherwise it would be necessary to look upon radioactive change as involving the creation of matter. In the universe therefore, the total radioactivity must, according to our present knowledge, be growing less and tending to disappear. Hence the energy liberated in radioactive processes does not disobey the law of the conservation of energy. It is not implied in this view that radioactivity, considered with reference to the quantity of matter involved, is conserved under all conceivable conditions, or that it will not ultimateiy be found possible to control the processes that give rise to it. The principle enunciated applies of course only to our present state of experimental knowledge, which is satisfactorily interpreted by its aid. The general evidence on which the principle is based embraces the whole field of radioactivity. The experiments of Becquerel and Curie have shown • Apart from the considerations that follow, this nomenclature is a convenient expression of the observed facts that the total radioactivity (measured by the radiations peculiar to tbe radio-elements) is for any given mass of radio-element a constant under all conditions investigated. The radioactive equilibrium may be disturbed and the activity distributed among one or more active products capable of separation from the original element. But the sum total throughout these operations is at aß times the same. Far praclica/ purpases the expression 'conservation', applied to the radioactivity of the three radio-elements, is justified by the extremely minute proportion that can change in any interval over which it is possible to extend actual observations. But rigidly the term 'conservation' applies only with reference to the radioactivity of any definite quantity of radioactive matter, whereas in nature this quantity must be changing spontaneously and continually growing less. To avoid possible misunderstanding, therefore, it is necessary to use tbe expression only in tbis restricted sense.

602

The Collected Papers 01 Lord

Rulhe~fo,.d

that the radiations from uranium and radium respectively remain constant over long intervals of time. Mme. Curie put forward the view that radioactivity was a specific property of the element in question, and the successful separation of the element radium from pitchblende was a direct result of this method of regarding the property. The possibility of separating from a radio-element an intensely active constituent, although at first sight contradictory, has afforded under eloser examination nothing but confirmation of this view. In all cases only apart of the activity is removed, and this part is recovered spontaneously by the radio-element in the course of time. Mme. Curie's original position, that radioactivity is a specific property of the element, must be considered to be beyond question. Even if it should ultimately be found that uranium and thorium are admixtures of these elements with a small constant proportion of new radio-elements with correspondingly intense activity, the general method of regarding the subject is quite unaffected. In the next place, throughout the course of our investigations we have not observed a single instance in which radioactivity has been created in an element not radioactive, or destroyed or altered in one that is, and there is no case at present on record in which such a creation or destruction can be considered as established. It will be shown later that radioactive change can only be of the nature of an atomic disintegration, and hence this result is to be expected, from the universal experience of chemistry in failing to transform the elements. For the same reason it is not to be expected that the rate of radioactive change would be affected by known physical or chemical infl.uences. Lastly, the principle of the conservation of radioactivity is in agreement with the energy relations of radioactive change. These will be considered more fully in § 7, where it is shown that the energy changes involved are of a much higher order of magnitude than is the case in molecular change. It is necessary to consider briefl.y some of the apparent exceptions to this principle of the conservation of radioactivity. In the first place it will be recalled that the emanating power of the various compounds of thorium and radium respectively differ widely among themselves, and are greatly infl.uenced by alterations of physical state. It was recently proved (Phi!. Mag., April 1903, p. 453) that these variations are caused by alterations in the rate at which the emanations escape into the surrounding atmosphere. The emanation is produced at the same rate both in de-emanated and in highly emanating thorium and radium compounds, but is in the former stored up or occluded in the compound. By comparing the amount stored up with the amount produced per second by the same compound dissolved, it was found possible to put the matter to a very sharp experimental test which completely established the law of the conservation of radioactivity in these cases. Another exception is the apparent destruction of the thorium excited activity deposited on a platinum wire by ignition to a white heat. This has recently been examined in this laboratory by Miss Gates, and it

Radioactive ('hange

603

was found that the excited activity is not destroyed, but is volatilized at a definite temperature and redeposited in unchanged amount on the neighbouring surfaces. Radioactive 'lnduction'. Various workers in this subject have explained the results they have obtained on the idea of radioactive 'induction', in which a radioactive substance has been attributed the power of inducing activity in bodies mixed with it, or in its neighbourhood, which are not otherwise radioactive. This theory was put forward by Becquerel to explain the fact that certain precipitates (notably barium sulphate) formed in solutions of radioactive salts are themselves radioactive. The explanation has been of great utility in accounting for the numerous examples of the presence of radioactivity in non-active elements, without the necessity of assuming in each case the existence of a new radio-element therein, but our own results do not allow us to accept it. In the great majority of instances that have been recorded the results seem to be due simply to the mixture 0/ active matter with the inactive element. In some cases the effect is due to the presence of a small quantity of the original radio-element, in which case the 'induced' activity is permanent. In other cases, one of the disintegration products, like uranium X or thorium X, has been dragged down by the precipitate, producing temporary, or, as it is sometimes termed, 'false' activity. In neither case is the original character of the radiation at all affected. It is probable that a re-examination of some of the effects that have been attributed to radioactive induction would lead to new disintegration products of the known radio-elements being recognized. , Other Results. A number of cases remain for consideration, where, by working with very large quantities of material, there have been separated from minerals possible new radio-elements, Le. substances possessing apparently permanent radioactivity with chemical properties different from those of the three known radio-elements. In most of these cases, unfortunately, the real criteria that are of value, viz., the nature of the radiations and the presence or absence of distinctive emanations, have not been investigated, The chemical properties are of less service, for even if a new element were present, it is not at all necessary that it should be in sufficient quantity to bc detected by chemical or spectroscopic analysis. Thus the radio-lead described by Hoffmann and Strauss and by Giesel cannot be regarded as a new element until it is shown that it has permanent activity of a distinctive character. In this connection the question whether polonium (radio-bismuth) is a new element is of great interest. The polonium discovered by Mme. Curie is not a permanent radioactive substance, its activity decaying slowly with the time. On the view put forward in these papers, polonium must be regarded as a disintegration product of one of the radio-elements present in pitchblende. Recently, however, Marckwald (Ber. der D. ehem. Gesel., 1902, pp. 2285 and 4239), by the electrolysis of pitchblende solutions. has obtained

604

The Collected Papers of Lord Rutherlord

an intensely radioactive substance very analogous to the polonium of Curie. But he states that the activity of his preparation does not decay with time, and this, if confirmed, is sufficient to warrant the conclusion that he is not dealing with the same substance as Mme. Curie. On the other hand, both preparations give only cx rays, and in this they are quite distinct from the other radio-elements. Marckwald has succeeded in separating his substance from bismuth, thus showing it to possess different chemical properties, and in his latest paper states that the bismuth-free product is indistinguishable chemica1ly from tellurium. If the permanence of the radioactivity is established, the existence of a new radio-element must be inferred. If elements heavier than uranium exist it is probable that they will be radioactive. The extreme delicacy of radioactivity as a means of chemical analysis would enable such elements to be recognized even if present in infinitesimal quantity. It is therefore to be expected that the number of radio-elements will be augmented in the future, and that considerably more than the three at present recognized exist in minute quantity. In the first stage of the search for such elements a purely chemical examination is of little service. The main criteria are the permanence of the radiations, their distinctive character, and the existence or absence of distinctive emanations or other disintegration products. § 6. The Relation of Radioactive Change to Chemical Change

The law of radioactive change, that the rate of change is proportional to the quantity of changing substance, is also' the law of monomolecular chemical reaction. Radioactive change, therefore, must be of such a kind as to involve one system only, for if it were anything of the nature of a combination, where the mutual action of two systems was involved, the rate of change would be dependent on the concentration, and the law would involve a volume-factor. This is not the case. Since radioactivity is a specific property of the element, the changing system must be the chemical atom, and since only one system is involved in the production of a new system and, in addition, heavy charged partic1es, in radioactive change the chemical atom must suffer disintegration. The radio-elements possess of aU elements the heaviest atomic weight. This is indeed their sole common chemical characteristic. The disintegration of the atom and the expulsion of heavy charged particles of the same order of mass as the hydrogen atom leaves behind a new system lighter than before, and possessing chemical and physical properties quite different from those of the original element. The disintegration process, once started, proceeds from stage to stage with definite measurable ve10cities in each case. At each stage one or more cx 'rays' are projected, until the last stages are reached, when the ß 'ray' or electron is expeUed. It seems advisable to possess a special name for these now numerous atom-fragments, or new atoms, which result from the original atom after the ray has been expelled, and which remain in existence only a limited time, continually undergoing

605

Radioacth'e Change

further change. Their instability is their chief characteristic. On the one hand, it prevents the quantity accumulating, and in consequence, it is hardly likely that they can ever be investigated by the ordinary methods. On the other, the instability and consequent ray-expulsion furnishes the means whereby they can be investigated. We would therefore suggest the term metaboion for this purpose. Thus in the following table the metaboions at present known to result from the disintegration of the three radio-elements have been arranged in order. Uranium

Thorium

Radium

UraniumX

Thorium X

Radium Emanation

Thorium Emanation

Radium-Excited Activity I

Thorium-Excited Activity I

ditto 11

ditto 11

ditto III

The three queries represent the three unknown ultimate products. The atoms of the radio-elements themselves form, so to speak, the common ground between metaboions and atoms, possessing the properties of both. Thus, although they are disintegrating, the rate is so slow that sufficient quantity can be accumulated to be investigated chemically. Since the rate of disintegration is probably a million times faster for radium than it is for thorium or uranium, we have an explanation of the excessively minute proportion of radium in the natural minerals. Indeed, every consideration points to the conc1usion that the radium atom is also a metaboion in the full sense of having been formed by disintegration of one of the other elements present in the mineral. For example, an estimation of its 'life' goes to show that the latter can hardly be more than a few thousand years (see § 7). The point is under experimental investigation by one of us, and a fuller discussion is reserved until later. There is at present no evidence that a single atom or metaboion ever produces more than one new kind of metaboIon at each change, and there are no means at present of finding, for example, either how many metaboions of thorium X, or how many projected partic1es, or 'rays', are produced from each atom of thorium. The simplest plan, therefore, since it involves no possibility of serious error if the nature of the convention is understood, is to assume that each atom or metaboion produces one new metaboion or atom and one 'ray'.

§ 7. The Energy

0/ Radioactive Change, and the Internal Energy 0/ the Chemical Atom

The position of the chemical atom as a very definite stage in the complexity of matter, although not the lowest of which it is now possible to obtain

606

The Collected Papers of Lord Rutllelford

experimental knowledge, is brought out most clearly by a comparison of the respective energy relations of radioactive and chemical change. It is possible to calculate the order of the quantity of energy radiated from a given quantity of radio-element during its complete change, by several independent methods, the conclusions of which agree very weIl among themselves. The most direct way is from the energy of the particles projected, and the total number of atoms. For each atom cannot produce less than one Gray' for each change it undergoes, and we therefore arrive in this manner at a minimum estimate of the total energy radiated. On the other hand, one atom of a radio-element, if completely resolved into projected particles, could not produce more than about 200 such particles at most, assuming that the mass of the products is equal to the mass of the atom. This consideration enables us to set a maximum limit to the estimate. The IX rays represent so large a proportion of the total energy of radiation that they alone need be considered. Let m = mass of the projected particle, v = the velocity, e = charge. Now for the

IX

ray of radium

= 2·5

v

e

-m =

X

109,

6 X 103•

The kinetic energy of each particle ~mv2

1m

= -2 -e v2e =

5

X

1014e.

J. J. Thomson has shown that e

=

6 X 10-10 B.S. Units = 2 X 10-20 Electromagnetic Units.

Therefore the kinetic energy of each projected particle = 10-5 erg. Taking 1()2° as the probable number of atoms in one gram of radium, the total energy ofthe rays from the latter = 1015 ergs = 2·4 X 107 gram-calories, on the assumption that each atom projects one ray. Five successive stages in the disintegration are known, and each stage corresponds to the projection of at least one ray. It may therefore be stated that the total energy of radiation during the disintegration of one gram of radium cannot be less than 108 gram-calories, and may be between 109 and 1010 gram-calories. The energy radiated does not necessarily involve the whole of the energy of disintegration and may be only a small part ofit. 108 gram-calories per gram may therefore be safely accepted as the least possib1e estimate of the energy of radioactive change in radium. The union of hydrogen and oxygen liberates approximately 4 x 103 gram-calories per gram of water produced, and this

Radioactil'e Change

607

reaction sets free more energy for a given weight than any other chemical change known. The energy of radioactive change must therefore be at least twenty-thousand times, and may be a million times, as great as the energy of any molecular change. The rate at which this store of energy is radiated, and in consequence the life of a radio-element, can now be considered. The order of the total quantity of energy liberated per second in the form of rays from 1 gr. of radium may be calculated from the total number of ions produced and the energy required to produce an ion. In the solid salt a great proportion of the radiation is absorbed in the material, but the difficulty may be to a large extent avoided by determining the number of ions produced by the radiation of the emanation, and the proportionate amount of the total radiation of radium due to the emanation. In this case most of the rays are absorbed in producing ions from the air. It was experimentally found that the maximum current due to the emanation from 1 gr. of radium, of activity 1,000 compared with uranium, in a large cylinder filled with air, was 1· 65 X 10-8 electromagnetic units. Taking e = 2 X 10-20 , the number of ions produced per second = 8·2 X 10 11 • These ions result from the collision of the projected partic1es with the gas in their path. Townsend (Phi!. Mag., 1901, vol. i.), from experiments on the production of ions by collision, has found that the minimum energy required to produce an ion is 10-11 ergs. Taking the activity of pure radium as a million times that of uranium, the total energy radiated per second by the emanation from 1 gr. of pure radium = 8,200 ergs. In radium compounds in the solid state, this amount is about 0·4 of the total energy of radiation, which therefore is about 2· 1Q4 ergs per second, 6·3 X 1011 ergs per year, 15,000 gram-calories per year. This again is an under-estimate, for only the energy employed in producing ions has been considered, and this may be only a small fraction of the total energy of the rays. Since the oe radiation of all the radio-elements is extremely similar in character, it appears reasonable to assurne that the feebier radiations of thorium and uranium are due to these elements disintegrating less rapidly than radium. The energy radiated in these cases is about 10-6 that from radium, and is therefore about 0·015 gram-calories per year. Dividing this quantity by the total energy of radiation, 2·4 X 107 gram-calories, we obtain the number 6 X 10-10 as a maximum estimate for the proportionate amount of uranium or thorium undergoing change per year. Hence in one gram of these elements less than a milligram would change in a million years. In the ca se of radium, however, the same amount must be changing per gram per year. The 'life' of the radium cannot be in consequence more than a few thousand years on this minimum estimate, based on the assumption that each partic1e produces one ray at each change. If more are produced the life becomes

608

The Collected Papers 0/ Lord Ruthelford

correspondingly longer, but as a maximum the estimate can hardly be increased more than fifty times. So that it appears certain that the radium present in a mineral has not been in existence as long as the mineral itself, but is being continually produced by radioactive change. Lastly, the number of 'rays' produced per second from 1 gr. of a radioelement may be estimated. Since the energy of each 'ray' = 10-5 ergs = 2·4 X 10-13 gram-calories, 6 X 1010 rays are projected every year from 1 gr. of uranium. This is approximately 2,000 per second. The IX radiation of one milligram of uranium in one second is probably within the range of detection by the electrical method. The methods of experiment are therefore almost equal to the investigation of a single atom disintegrating, whereas not less than 1014 atoms of uranium could be detected by the balance. It has been pointed out that these estimates are concerned with the energy of radiation, and not with the total energy of radioactive change. The latter, in turn, can only be a portion of the internal energy of the atom, for the internal energy of the resulting products remains unknown. All these considerations point to the conclusion that the energy latent in the atom must be enormous compared with that rendered free in ordinary chemical change. Now the radio-elements differ in no way from the other elements in their chemical and physical behaviour. On the one hand they resemble chemically their inactive prototypes in the periodic system very closely, and on the other they possess no common chemical characteristic which could be associated with their radioactivity. Hence there is no reason to assume that this enormous store of energy is possessed by the radio-elements alone. It seems probable that atomic energy in general is of a similar high order of magnitude, although the absence of change prevents its existence being manifested. The existence of this energy accounts for the stability of the chemical elements as weH as for the conservation of radioactivity under the influence of the most varied conditions. It must be taken into account in cosmical physics. The maintenance of solar energy, for example, no longer presents any fundamental difliculty if the internal energy of the component elements is considered to be available, i.e. if processes of sub-atomie change are going on. It is interesting to note that Sir Norman Loekyer has interpreted the results of his spectroseopic researches on the latter view (Inorganic Evolution, 19(0) although he regards the temperature as the cause rather than the effect of the process. McGill University Montreal

The Amount of Emanation and Helium from Radium From Nature, 68,1903, pp. 366-7

IN connection with the very striking experiments described by Sir William Ramsay and Mr. Soddy in NATURE of August 13, in which they have observed the presence of helium in the gases obtained from radium bromide and also the production of helium by the emanation of radium it may be of interest to give some calculations of the probable amount of emanation and of helium produced by radium on the disintegration hypothesis, recently put forward by Mr. Soddy and myself to explain the phenomena of radio-activity. A method of calculation has already been indicated by us (Phi!. Mag., May), but the data on which it was based are somewhat imperfect. A more accurate estimate can be made from the data of the amount of heat liberated by radium, recently measured by Curie and Laborde. I have shown that the (X or easily absorbed rays from radium consist of a stream of positively charged bodies, of mass about twice that of the hydrogen atom, projected with a velocity of about 2· 5 X 109 cm. per sec. These results have been recently confirmed by Des Coudres. These (X bodies are expelled from every part of the mass of radium, but in consequence of the ease with which they are absorbed, only a small proportion of them escapes into the air. This self-bombardment of the radium probably gives rise to a large proportion of the heat which keeps the radium at a temperature above that of the surrounding atmosphere. Assuming for the moment that all of the heat is supplied by this continuous bombardment, an estimate can readily be made of the number of (X bodies projected per second from one gramme of radium. The kinetic energy of each projected body is 5 x 10- 6 ergs. Since this energy is transformed into heat in the mass of radium, the number of bodies projected to give an emission of heat of 100 gr. cals. per hour-the amount determined by Curie and Laborde-can be shown to be 2·4 x 10 11 per second. Now Townsend has shown from experimental data that Ne = 1·22 X 1010, where N is the number of atoms in 1 C.c. of gas at standard pressure and temperature, and e is the charge carried by an ion. The latest value of e, found by J. J. Thomson, is 3·4 x 10- 1°, so that N = 3·6 X 10 19. If the (X bodies after expulsion can exist in the gaseous form, the volume of the gas produced (at standard pressure and temperature) is thus 2·4 x 1011 .6 X 1019 =6·7 X 10-9 c.c. per sec., or 0·21 c.c. per year. Allowing a wide margin for the possibility that only one-tenth of the heat emitted by radium is due to the kinetic energy of the projected bodies, the volume of u

610

The Collected Papers ofLord Rutherford

the cx partic1es should lie between 0·021 c.c. and 0·21 c.c. per year for each gramme of radium. The determination of the mass of the cx body, taken in conjunction with the experiments on the production of helium by the emanation, supports the view that the cx partic1e is in reality helium. In addition, the remarkable experiment of Sir William and Lady Huggins in which they found that the spectrum of the phosphorescent light of radium consisted of bright lines, some of which within the limit of error were coincident with the lines of helium in the ultra-violet, strongly supports such a view. For as a consequence of the violent expulsion of the cx partic1e it is to be expected that it would be set into powerful vibration and give its characteristic spectrum. In the experiments of Sir William Ramsay and Mr. Soddy 30 milligrammes of radium bromide, probably about four months old, were used. If the cx body is helium, the amount of helium liberated by solution of the radium in water must have been between 0·00017 and 0·0017 c.c., assuming that all of the helium produced was occluded in the mass of the substance. There is evidence of at least five distinct changes occurring in radium, each of which is accompanied by the expulsion of an cx partic1e. One of the products of these changes is the radium emanation. It is of interest to calculate the volume of the emanation occluded in radium when in astate of radioactive equilibrium. Taking as the simplest hypothesis that one cx partic1e is projected at each change, the number of atoms of the emanation produced per second is 1/5 of the number of cx partic1es, i.e. 1·3 X 10- 9 c.c. When radio-active equilibrium is reached, it has been shown that 463,000 times the amount of emanation produced per second is stored up in the radium. This corresponds to 6 X 10- 4 C.C. The maximum amount of emanation to be obtained from one gramme of radium thus probably lies between 6 X 10- 5 c.c. and 6 X 10- 4 C.C. The radium emanation is the active principle of radium, for about i- of the activity of radium is due to it. Thus a large proportion of the radiations from radium is a direct result of the changes occurring in the very minute amount of matter constituting the radium emanation. If ever 1 C.c. of the radium emanation can be collected at one spot, it will exhibit some remarkable properties. The powerful radiations from it would heat to a red heat, if they would not melt down, the glass tube which contains it. This very rapid emission of energy, in comparison with the amount of matter producing it, would continue for several days without much change, and would be appreciable after a month's interval. The very penetrating rays from it would light up an X-ray screen brilliantly through a foot of solid iron. E. RUTHERFORD Bettws-y-Coed August 15

Heating Effect of the Radium Emanation From Nature, 68, 1903, p. 622

IN connection with the discovery of P. Curie and Laborde that radium continuously emits heat at a rapid rate, an interesting question arises as to whether the heat emission is directly connected with the radio-activity of that element or is independent of it. To settle this point we have performed the following experiments. The heating effect of 30 milligrammes of pure radium bromide was first measured in a differential air calorimeter. The radium bromide was then heated to a sufficient temperature to drive off the emanation, and the latter was condensed by passing through a short glass tube immersed in liquid air, and then the tubes were sealed off. On testing the de-emanated radium, the heating effect diminished rapidly during the first few hours, and fell to a minimum corresponding to about 30 per cent. of the original value and then slowly increased again. On substituting the emanation tube in the calorimeter, the heating effect at first increased for a few hours to a maximum corresponding to about 70 per cent. of the original heat emission of the radium and then slowly decayed with the time. At any time after removal of the emanation, the sum of the heating effect of the de-emanated radium and of the emanation was found to be the same as that of the original radium. Experiments are still in progress to determine the rate of recovery and loss of heating power of the de-emanated radium and the separated emanation respectively, but so far as the observations have gone, the curves of decay and recovery are the same as those for the corresponding IX radiation. It has been shown (Rutherford and Soddy, Phi!. Mag., May) that, if the emanation is removed from radium, the activity of the radium decays in the course of a few hours to about 25 per cent. of its original value. This residual activity consists entirely of IX rays. The solid radium compound regains its original activity after the lapse of about one month. Immediately after the separation of the emanation the activity (tested in a sealed vessel) rises to about twice its original value, due to the production of excited activity on the walls of the vessel, and then slowly decays with the time, falling to half value in ab out four days. At any time after removal of the emanation the sum total of the activity of the radium and the emanation has a value equal to that of the original radium. There is thus an exact parallel between the variation in radiating power (measured by the IX rays) and the heating effect. In order to be sure how much of the emanation was removed by heating, control experiments were made

612

TIle Collected Papers 0/ Lord Rutlleliord

on the y rays from the radium and the separated emanation. This was tested by observing the rate of discharge of an electroscope after the rays had passed through 5 cm. of lead. In some preliminary experiments by one of us last year it was found that the y rays from radium appeared at the same time as ß rays, and were always proportional to them. From these results it was deduced that all but 6 per cent. of the emanation was removed by the heating. It is thus seen that the heating effect of radium directly accompanies the cx radiation from it, and is always proportional to it, and that more than two-thirds ofthe heating effect is not due to the radium at all, but to the radioactive emanation which it produces from itself. This result accounts for the variation of heat emission with age observed by the euries, an account of which was given by Prof. Dewar at the British Association. The amount of emanation from 30 milligrammes of radium bromide, when collected in the tube, was sufficient to cause a bright phosphorescence in the tube, but it was too small either to measure or weigh. The amount of heat emitted from the radium emanation is thus enormous compared with the amount of matter involved. It seems probable that the greater part of the heating effect of radium is a direct consequence of the expulsion of cx rays. It still remains to be shown in what proportion the radiated energy is distributed between the projected cx particles and the systems from which they are expelled. The results given here are at once explained on the disintegration hypothesis (Rutherford and Soddy, Phil. Mag., May), in which the heat is considered to be derived from the internal energy of the atom. On the view held by some that radium gains its heat from an external source, it would be necessary to suppose that less than a third of the heat is due to the radium itself, and that the other two-thirds are due to the radium emanation which is being continuously produced, and the power of which of absorbing energy from an external source decays with the time. E. RUTHERFORD H. T. BARNES McGill University, Montreal October 16

Heating Effect of the Radium Emanation From Nature, 69, 1903, p. 126

IN a letter to NATURE of November 5, Prof. Schuster has made some remarks on a letter published by us the week previously, containing abrief account of some experiments to show that the heating effect of radium is temporarily reduced by the removal of the emanation, and that the tube containing the emanation separated from the radium shows a considerable heating effect. The difficuIty feIt by Prof. Schuster apparently arose from the fact that we included in the heating effect of the emanation not only that due to the emanation itself, but also that due to the secondary products to which the emanation gives rise. It was an oversight on our part to have omitted in the sentence 'more than two-thirds of the heating effect is not due to the radium at all, but to the radio-active emanation which it produces from itself', the words 'together with the secondary products to which the emanation gives rise'. We were fully aware that the heating effect was in part due to the 'excited activity' produced by the emanation. We specially mentioned the gradual decay of the heating effect of radium to a minimum in the course of a few hours, and the increase of the heating effect of the emanation tube during the same period. These effects are connected with the gradual decay and rise, respectively, ofthe excited activity produced by the radium emanation. The results would have little meaning ifwe believed the heating effect was due to the emanation alone, for, as Prof. Schuster quite correctly points out, the heating effect in such a case should at once drop to a minimum after removal of the emanation, and the heating effect of the tube containing the emanation should not at first increase. On account of the rapid rise of the excited activity in a tube containing the radium emanation, the separation of the heating effect of the emanation from the complicated secondary changes which result from it is a difficult experimental problem. Our letter was merely a preliminary announcement of the results of our experiments. It is not possible to discuss the consequences to be deduced from the experiments without ente ring into a detailed description of the measurements. We hope to publish shortly a full account of our work on the various heating effects. E. RUTHERFORD H. T. BARNES McGill University November 20

Radioactive Processes by

PROF. E. RUTHERFORD

[Abstract]

From Proceedings ofthe Physical Society, 18, 1903, pp. 595-7 Read June 5,1903

THERE are three distinct types of radiation spontaneously emitted from radioactive bodies, which may be called the oe, ß, and y rays. The oe rays are prominent in causing the conductivity of agas, they are easily absorbed by metals and are projected bodies, not waves. These bodies are about the size of a hydrogen atom, they are positively charged and travel with about one-tenth of the velocity of light. The ß rays are similar in all respects to the cathode rays produced in a vacuum-tube. The y rays are probably like Röntgen rays, but ofvery great penetrating power. The oe rays are by far the most important. In addition to these rays two of the radio-elements give off radioactive 'emanations', which are in all respects like gases. The radiations from these emanations are not permanent, but fall off in a geometrical progression with the time. The radiation of the thorium emanation falls to half value in one minute, that from radium in four days. They have alt the properties of gaseous matter in infinitesimal quantity. Their coefficients of diffusion can be measured, the order of their molecular weights is 100, they are occluded by solid compounds producing them and may be condensed at low temperatures. The radium emanation condenses sharply at -150° C., the thorium emanation between -120° C. and -150° C. The two emanations excite on objects, with which they come in contact, two kinds of temporary radioactivity, that from the radium emanation decaying much faster than that from the thorium emanation. The latter decays in a G.P. with the time, falling to half value in eleven hours. These effects appear to be produced by solid matter in invisible and unweighable quantity, which can be dissolved off in some acids but not in others. On evaporating the solutions, the radioactivity is obtained unchanged in the residue. The experiments of Crookes and Becquerel in separating by chemical treatment the matter responsible for the activity of uranium, called uranium X, were referred to, together with the latter's observation that the separated activity had completely decayed after the lapse of a year, by which time the uranium itself had completely recovered its activity.

Radioaf'{ h'C' P/'Of'('ssC's

615

The work of Rutherford and Soddy on thorium was then discussed in detail. Thorium precipitated in solution by ammonia retains only 25 per cent of its activity. lf the solution is evaporated and ignited the remaining 75 per cent is found in the extremely small residue left, which, by reason of its separation, is different chemically from thorium and was called Thorium X. Left to themselves the thorium gradually recovers its activity and the ThX loses it. The activity of the latter falls in a G.P. with thorium the half value being reached after four days. At any time the sum-total of the two activities is a constant. This would occur if the ThX were being continua11y produced by the thorium, and this was shown to be the case by precipitating thorium at definite intervals after its separation from ThX. The ThX, and not thorium, pro duces the thorium emanation. The production of ThX by thorium, of the emanation by ThX, and of the matter causing the excited activity by the emanation, are a11 changes of the same type, although the rates of change are distinct in each case. The change of uranium into uranium X is also similar, being the slowest of a11. Twenty-two days elapse before uranium freed from UrX recovers one half of its activity. In radium the radium emanation is the first product produced, and since this in asolid is almost completely occluded, the activity of a radium salt after it has been obtained from its solution rises after precipitation to several times its original value, due to the occ1usion of the emanation. In a11 three radio-elements a part of the radioactivity is non-separable, and this part consists only of IX rays. The ß rays only result at the last stages of the process that can be experimenta11y traced. In a11 cases the radiation, from any type of active matter, is a measure of the amount of the next type produced. Thus the radioactivity of ThX at any period throughout its life is always a measure of the amount of emanation it produces. These results find their explanation if it is supposed that the IX partic1es projected form integral portions of the atom of the radioactive element. Thus ThX is thorium minus one or more projected IX particles. The emanation similarly is ThX less a further IX particle, and so on. The non-separable activity is due to the atoms of the original radio-element disintegrating at a constant rate. The whole of the processes take place unaltered in velocity, apparently under a11 conditions of temperature, state of aggregation, and chemical combination. This is to be expected of a subatomic change in which one system only is involved at each change. On this view the spontaneous heat-emission of solid radium-salts, discovered by Curie, is explained by the internal bombardment by the IX particles shot off and absorbed in the mass of the substance. The amount of energy given out in these subatomic changes is enormous, and from Curie's experiments it can be deduced that each gram of radium gives out 109 gram-calories during its life, which is sufficient to raise 500 tons a mile high. It seems probable that the internal energy of atoms in general is of a simiJar high order of magnitude.

616

The Collected Papers 0/ Lord Ruthelford Discussioll

Sir Oliver Lodge congratulated Prof. Rutherford and Mr. Soddy upon the field of research which had opened up to them and upon the way they were pursuing it. Referring to the fact that the temperature of a piece of radium is above that of its surroundings, he said that of Prof. Rutherford's discovery that atoms of matter were thrown out by radium at one-tenth the speed of light would account for energy effects and he thought it necessary that when energy passed from an unrecognized atomic to an irregular molecular form, that previously non-existent heat would be produced. The important point in Prof. Rutherford's work was that he bad established the fact that one kind of matter is thrown off by another kind of matter, and bad also measured the velocity and weight ofthe particles thus thrown off. We were accustomed to electrons being shot out by matter, but the proof that light atoms were thrown off from heavy atoms was the evolution or transrnutation of matter experimentally demonstrated. The heavy atoms of radium appear to collapse and throw off atoms of low atomic weight; the remainder is unstable, and more matter is thrown off, the original atom getting, it is to be presumed, lighter and lighter-according to the view of the author of the paper. It might be thought that this hypothesis about the degradation and the instability of the atoms was mere speculation, but it was the most reasonable explanation of observed phenomena. And the reason he was thus cordially willing to accept it as a working hypothesis was because he had been indistinct1y looking for some such effect, being guided thereto by pure theory. And that was the chief point he wished to bring forward. According to an electric theory of matter, i.e. on the view that an atom contained electrons with rapid interatomic movements obeying laws like astronomica1laws, this instability ought to exist. Taking a formula by Prof. Larmor for the radiating power of an accelerated electric charge, Sir Oliver Lodge gave a rough proof of the latter statement (reproduced in Nature for lune 11th). He stated, in conclusion, as a working hypothesis that we must not suppose that atoms are permanent and eternal, and suggested that we may possibly find a rise and decay in ordinary matter, and find tbat the history of an atom may be written-in accordance with the analogy of solar systems and cosmic configurations generally. On the electric theory of matter the falling together of electrons might produce the electric aggregate called an atom, and its subsequent gradual decay or separation into other forms would be accompanied by epochs of radioactivity. Prof. W. E. Ayrton, after congratulating the author, said he would like to ask a question. Prof. Rutherford had shown us the discharge of an electroscope when a thick plate of metal was placed between a piece of radium and the instrument. Would the electroscope be affected if a piece of radium completely surrounded by an earthed metal casing was brought near to it? Again, if a piece of radium was insulated inside a completely closed metal box, would an electroscope, metallically connected with the outside of the

Radioactil'e Processes

617

box-the whole thing being placed in a very good vacuum-be charged, and if so, what would be the sign of the charge. Prof. Everett found it difficult to believe that there was a sufficient store of energy in the atom to account for the effects observed. He asked if the phenomena were not due to resonance, the radium atoms being shaken asunder by vibrations in the ether to which they were responsive. These scattered atoms might knock asunder some of the atoms of neighbouring bodies, and so produce the ß rays. Prof. S. P. Thompson pointed out that the notion of energy within the atom was based upon the statement that the temperature of a piece of radium was above the temperature of its surroundings. Without doubting the work of Curie, he thought that critical experiments should be made on this point before it could be looked upon as an accepted fact or such large generalizations be based upon it. Dr. Lowry called attention to the slender evidence on which the theory of atomic degradation had been based. The behaviour of thorium and ThX was precisely analogous to that of a substance like 7r bromo-nitrocamphor, which exists in an inactive 'normal' form as a neutral substance and a dielectric, but also in an active 'pseudo' form in which it is a strong acid and an electrolyte. Both forms are fairly stable in the solid state, but when dissolved are exceedingly labile, and the merest trace of impurity is sufficient to bring about an isometric change which leads in the course of a few days or hours to a condition of equilibrium in which there are about 94 parts of the inactive normal to 6 parts of the active pseudo form. The curves representing the rate of recovery and decay of (electrolytic) activity are mathematically identical with those shown for the recovery and decay of radioactivity by the two modifications of thorium and uranium. There is, therefore, the strongest evidence in favour of the view that the conversion of thorium into ThX is a reversible change, and that the emanations and excited activity are merely physical phenomena. Prof. Rutherford, in reply to Prof. Ayrton, said that in the first case discharge would take place, and that the result in the second case would depend upon the thickness of the walls, and therefore upon the nature of the rays which could pass through. He also pointed out that Dr. Lowry's explanation involved hirn in perpetual motion. It did not explain radioactivity, and it was difficult to see how it could account for the phenomena occurring in stages. Moreover, there was no reversible chemical action that went on at the same rate at a red heat as in liquid air. With regard to Prof. Everett's suggestion, it seemed that such a process would be even more wonderful than that suggested. The phenomena go on unchanged even when the substance is shielded from external influence by many inches of lead. He feIt confident that Curie's observation of heat emission would prove to be well founded.

u·

Does the Radio-activity of Radium depend upon its Concentration? From Nature, 69, 1904, p. 222

SOMB experiments have recently been made to test whether the radio-activity of radium is influenced by the continuous bombardment to which it is subjected by its own radiations. In an artic1e in this Journal on radium (April 30, 1903) Prof. J. J. Thomson suggested that the radio-activity of radium may possibly depend upon its degree of concentration, and that a given quantity of radium, diffused throughout a mass of pitchblende, may be less than when concentrated in a small mass. In order to test this point, measurements of the radio-activity of radium bromide were made when in the solid state and when diffused throughout the mass of a solution more than a thousand times the volume occupied by the radium compound. Two tubes, c10sed at one end, were taken, in one of which was placed about a milligram of pure solid radium bromide and in the other a solution of radium chloride. The tubes were connected near the top by a cross tube, and the open ends were then sealed by a blowpipe. Measurements of the radio-activity of the radium were made by means of an electroscope. The tubes, fixed on astand, were placed in a definite position near an electroscope and the rate of discharge 0bserved. This was due to the ßand y rays emitted by the radium, since the oe rays were completely absorbed in the walls ofthe tube. By placing a lead screen 6 mm. thick between the tubes and the electroscope the rate of discharge was due to the y rays alone. After measurements of the activity had been made, the glass apparatus was tilted so as to allow the radium chloride to flow into the arm containing the radium bromide. This dissolved the radium, and part of the emanation was released and distributed itself throughout the tubes. No appreciable change of the radio-activity of radium was observed over a month's interval. If the rate of production of the emanation, or the excited activity caused by it, had varied during the interval, a corresponding change would have been observed in the rate of discharge due to the y rays, for other experiments have shown that the amount of y rays is proportional to the amount of emanation present, provided measurements are made several hours after the introduction of the emanation into avessei, in order to allow the excited activity to reach a maximum value. The rate of discharge due to the y rays was somewhat diminished, but this was due to an increased absorption of the ß rays by the solution, and not to a change in the rate of emission of these rays. On account of the great penetrating power of the

nors the Radio-actii'il)' of Radium depend upon ils COl1centration? 619 y rays, the illcreased absorption due to the presence of the solution was negligible. Since, after solution, the radium bromide was diffused through a mass of solution at least 1000 times the bulk of the solid radium bromide, we may conc1ude that a distribution of the radiating matter over a thousand times its original volume has no appreciable influence on its radio-activity. This experiment shows that, over the range investigated, the radio-activity of radium is not influenced by its own intense radiations. This result is in agreement with previous observations, for neither the radio-activity of any active product nor the rate of 10ss of its activity has been found to be affected by its degree of concentration. It is thus improbable that the energy given out by radium is due to an absorption of an unknown external radiation which is similar in character to the radiations which are emitted. Experiments are in progress to test whether still further dilution of the radium solution produces any alteration in its radio-activity. E. RUTHERFORD McGill University, Montreal December 18, 1903

Heating Effect of the Radium-emanation by E. R UTHERFORD, F.R.S. Macdonald Professor of Physics, McGill University, Montreal

From the Transactions 01 the Australasian Association lor the Advancement 01 Science Dunedin, January 1904, pp. 87-91

ESTIMATES have been made at various times of the energy radiated from the active substances in the form of IX and ß rays, and it was known that the amount of this energy was considerable. Rutherford and McClung* deduced that a thin layer of radium chloride, of activity 100,000 times uranium, emitted energy into the gas in the form of IX rays at the rate of 3,000 gramcalories per year. Taking the latest estimate, 1,500,000, as the activity of pure radium chloride compared with uranium, this corresponds to an emission of energy from pure radium chloride in the form of IX rays of 45,000 gram-calories per year. P. Curie and Labordet recently drew attention to the striking result that a radium compound continuously kept itself at a temperature of several degrees above that of the surrounding atmosphere. In addition to the energy emitted in the form of the ionizing radiations, there is thus a continuous emission of energy in the form of heat from the radium compound. Curie and Laborde determined by two distinct methods that 1 gr. of radium emits heat continuously at the rate of 100 gram-calories per hour, or 876,000 gram-calories per year. These results have since been confirmed by other observers. Giesel found that a small thermometer placed in a tube containing 1 gr. of radium bromide indicated a rise of temperature of SOC above that of the surrounding atmosphere. As far as observations have at present gone, this rate of heat-emission of radium is constant. In the course of a few years 1 gr. of radium would thus radiate an enormous amount of energy. It seems very improbable that such a large emission of heat is the result of an ordinary chemical change taking place in the radium. The union of hydrogen and oxygen to form a gram of water gives out only about 4,000 gram-calories, and more heat is evolved in this reaction than in any other known to chemistry. In order to explain the rapid and continuous heat emission of radium two general theories have been proposed. On the one view, radium is supposed to absorb known or unknown types of radiation from the atmosphere, and acts as a mechanism to transform this borrowed energy into the t Comptes Rendus, 136, p. 673, 1903. • Phil. Trans., A., p. 25, 1901.

lIeating tl/ect of the Radium-emanation

621

form of heat and the peculiar radiations emitted by radium. On the other view, the energy is supposed to be derived from the energy latent in the radium atom. The radium atom is supposed to be undergoing spontaneous disintegration, accompanied by the emission of rays, and the energy emitted in the form of heat and of IX, ß and y rays is a result of this breaking-up of the atom. Tbe view that the radium acts as a transformer of borrowed energy has no experimental evidence to support it. As far as experiments have gone the rate of heat emission of radium is independent of external conditions. Curie and Laborde found that the radium gave out the same amount of heat when immersed in an ice calorimeter as under ordinary conditions. J. J. Thomson (Nature, p. 601, 1903) has pointed out that it is impossible to suppose that under such conditions the radium borrowed the heat from the surrounding aIr. P. Curie,· however, observed that the heating effect of radium varied with the age of radium compound. It was small when first prepared, but increased to a maximum value in the course of one month, and did not change appreciably in the following two months. The explanation of this result will be seen later. On the disintegration theory advanced by Rutherford and Soddy (Phi/. Mag., May 1903), it is supposed that adefinite small proportion of the radium atoms (about 1 in every 1010 will suffice) break up per second. The disintegration of each atom is accompanied by the expulsion of an IX ray or partic1e with great velocity. I have recently shown that the IX rays of radium consist of positive1y charged bodies of mass about twice that of the hydrogen atom, projected at a speed of 20,000 miles per second. The expulsion of an IX partic1e willleave the atom lighter than before, and will have changed its chemical and physical properties. The radium atom minus one IX partic1e on this view constitutes the atom of the radium emanation. This emanation behaves like a chemically inert gas of high molecular weight, which obeys the laws of diffusion, and can be condensed by extreme cold. The atom of the emanation is again unstable, and in turn breaks up with the expulsion of another IX particle. The IX particles expelled from the emanation constitute the radiation from the emanation. After the expulsion of an IX partic1e from the emanation, the residue behaves like asolid, and attaches itself to the surface of bodies, giving rise to the phenomena of 'excited' or 'induced' activity. This matter in turn breaks up, and, after a succession of well-marked changes, a final product is reached which is not radioactive. Each of the products, like the emanation and the matter which causes excited activity, break up at a definite rate, which is uninfluenced by any chemical or physical agency. The reactions that give rise to the series of well-marked products in radium are of a different character from those observed in chemistry, for no reaction is known which proceeds at the same rate at the temperature of liquid air as at a red heat. • Societe de Physique, Paris, 1903.

622

The Collected Papers of Lord Rutherford

On the disintegration theory, the heat emitted from radium arises partly from the rays emitted and partly from the systems from which the rays are expelled. Of the three types of rays emitted by radium the oe, or easily absorbed rays, are by far the most important. More than 99 per cent of the energy radiated in the form of ionizing rays is due to them. The ß and 'Y rays, in comparison, are of far less significance. These oe rays are very easily absorbed in matter, and in a pellet of radium a large proportion of the oe partic1es which are expelled is absorbed by the mass of the radium itself. The radium should thus be heated above the temperature of the surrounding air by its self-bombardment. There is no reason, however, to suppose that all the energy emitted by radium is due to this self-bombardment. The expulsion of an oe partic1e from a system must set it into violent vibration. The component parts of the system left behind will tend to arrange themselves so as to form a stable or temporarily stable system. During this rearrangement, which probably entails a condensation of the parts of the atom, energy will be emitted, which will be manifested in the form of heat. It remains for experiment to decide how much of the heat emitted is due to the oe partic1es and how much to the resulting rearrangement of the parts of the atom. If these views are correct, a proportion of the heat emitted by radium should be due to the active products which arise from the radium, viz. the emanation and the matter which causes excited activity, and the heat emission should be directly connected with the radioactivity of radium. This point has been recently investigated by Professor Barnes and myself (Nature, October 29, 1903). The heating effect of 30 milligrams of radium bromide was first measured in a special form of differential air calorimeter: The radium bromide was then heated in a glass tube to a sufficient temperature to drive off the emanation occluded in it, and this emanation was condensed in a short glass tube immersed in liquid air. The tubes containing the radium and the emanation were then sealed off, and the heating effect due to each tested separately at definite intervals. It was observed that the heating effect of the de-emanated radium decreased for the first few hours to a minimum corresponding to about 30 per cent of the original heat emission. At the same time the heating effect of the emanation tube increased for the first few hours to a maximum corresponding to about 70 per cent of the rate of heat emission of the radium. The radium was found to spontaneously regain its rate of heat emission with time, and after a month's interval the rate of heat emission was the same as at first. At the same time the heating effect of the emanation tube, after reaching the maximum, steadily decayed with time, falling to half value in about four days. At any time the sum total of the heat emission of the de-emanated radium and the emanation tube was equal to that of the original radium. These results show that the heating effect of radium is directly connected with its radioactivity. It has long been known that radium when heated loses for the time 75 per cent of its activity measured by the oe rays. On leaving the radium the

I/('ating r.f!f.'CT nJ 111(' Radill117-('mmWTinl1

(i~3

lost activity is spontaneously recovered in the course of a month's interval. This recovery of activity is due to the gradual production of the emanation by the radium. This emanation is occ1uded in the radium compound, and, together with the excited activity produced by it, adds its radiations to that of the radium proper. When, however, the emanation is removed from a radium compound by heat, the matter which causes excited activity is left behind. The activity due to this matter gradually decays, and in the course of a few hours practically disappears. At the same time the emanation in the c10sed vessel produces excited activity on the walls of the containing vessel. The gradual decay of the heating effect of the radium to a minimum after removal of the emanation, and the gradual rise of the heating effect of the emanation tube, is connected with this decay and rise, respectively, of the excited activity produced by the emanation. The results indicate that a large proportion of the heat emitted from the emanation tube is due to the matter which causes excited activity. It is a difficult experimental problem to isolate the heating effect produced by the emanation from that due to the secondary products which arise from it. The conc1usion, however, may be drawn that more than two-thirds of the heat emission of radium is not direct1y due to the radium itself, hut is due to the radioactive emanation and the secondary products which result from it. I have indicated that probably three distinct changes occur in the matter which causes excited activity. The experiments have not yet been pushed far enough to decide how the rate of emission of energy is divided between the emanation and these three changes. With the amount of radium so far available, the presence of the emanation of radium has not been detected, either by its volume or its weight. The amount of emanation from 30 milligrams of radium must have heen extremely minute, but yet it produced a strong phosphorescence in the containing tube, and gave out a large quantity of heat. I have calculated from severallines of evidence (see Nature, p. 367, 1903) that 1 gr. of radium in astate of radioactive equilibrium contains from 6 X 10-5 to 6 X 10-4 C.C. of emanation, measured as agas at standard pressure and temperature. Since 1 gr. of radium emits 100 gram-calories per hour, the experiments show that the changes occurring in the emanation from it give rise to 70 gram-calories per hour. I C.c. of the emanation would thus emit energy at a rate lying between 1· 2 X 105 and 1·2 X 106 gram-calories per hour. This rate of emission of energy would suffice to heat to a red heat, if not to melt down, the gl ass tube containing the emanation. It is of interest to make a rough estimate of the amount of energy emitted per second from 1 Ib. weight of the radium emanation. The emanation behaves as agas of heavy molecular weight, probably Iying between 100 and 200. Taking the molecular weight as 200, I lb. of the radium emanation corresponds to about 25,000 C.c. of gas. The rate of emission of energy from 1 lb. of the emanation thus lies hetween 3 ~ 109 and 3 X 10 10 gram-calories per hour. This corresponds to

624

The Collected Papers o[ Lord Ruthelford

an initial rate of emission of energy of about 5,000 to 50,000 horse-power. Since the heating effect of the emanation falls to half value in about four days, ,t can be readily deduced that the energy emitted from 1 lb. of the emanation while its activity lasts lies between 30,000 and 300,000 horse-power-days. Quite independently of any hypothesis, it can be calculated from the experiments that the amount of heat evolved by 11b. of the emanation is of a similar order of magnitude, for it is known that an unmeasurable and unweighable quantity of emanation emits energy at a readily measurable rate. On the disintegration theory this enormous emission of energy is derived from the latent energy stored up in the complex radio atom. This energy is released by the spontaneous disintegration of the atom in several successive stages. It is very difficult to explain the experimental results on the view that the radium acts as transformer of energy borrowed from the atmosphere, for it would be necessary to postulate that most of the heat emission of radium is not due to the radium at all, but is due to the radium emanation, which is produced by itself. It is also necessary to suppose that the property of the emanation, and of the products to which it gives rise, of absorbing energy from external sources is not constant, but decays with time. It has been pointed out (Rutherford and Soddy, Phi!. Mag., May 1903) that the radio elements have no special chemical characteristics except their high atomic weight which distinguish them from the other chemical elements. It is thus probable that the energy resident in the atoms of the elements is enormous compared with that released or absorbed in chemical reactions. This energy has not been observed on account of the difficulty of breaking up the atoms by the physical and chemical processes at our disposal.

Heating Effect of the Radium Emanation by E. RUTHERFORD, F.R.S. andH. T. BARNES, D.se.

Professors of Physics, McGill University, Montreal From the Philosophical Magazine for February 1904, sero 6, vii, pp. 202-219 (A short account of the preliminary results was published in Nature (October 29, 1903, p. 622). Read before the American Physical Society, St. Louis, December 29, 1903)

P. CURIE and Laborde* first observed the rapid rate of heat emission of radium, and deduced that 1 gram of radium emitted heat at the rate of about 100 gram-calories per hour. In a later paper, P. Curiet found that the rate of emission of heat depended upon the age of the radium preparation. The heating effect for freshly prepared radium compound was small at first, but gradually increased to a maximum after a month's interval, and remained constant over a further interval of two months. The present experiments were undertaken with the view of seeing how the heat emission of radium is connected with its radioactivity. It has been shown by Rutherford and Soddyt that the radiation emitted from a radium compound in astate of radioactive equilibrium may be divided into three parts: (1) A non-separable radiation consisting entirely of oe rays and constituting about 25 per cent of the total radiation. (2) The radiation from the emanation occluded in the radium, also consisting entirely of oe rays. (3) The excited radiation produced by the emanation in the mass of the radium, and consisting of oe, ß, and y rays. (2) and (3) together constitute about 75 per cent of the total radiation. Some experiments have been recently made to find how much of the activity of radium is supplied directly by the emanation occluded in it. The saturation current, between parallel plates, due to a radium preparation spread uniformly over a platinum plate, was determined by means of an electrometer. The platinum plate was then heated rapidly to a temperature sufficient to completely drive off the emanation and the saturation current due to the radium immediately measured. There was a decrease observed corresponding to 18 per cent of the total. The gradual decay of the excited

* Comptes Rendus, cxxxvi, p. t Societe de Physique, 1903.

673 (1903).

t Phi!. Mag., April 1903.

626

The ('ollected Papers

0/ Lord Ruthellord

activity left behind in the radium after the removal of the emanation is shown graphically in Fig. 5, curve A. We may thus conclude that the emanation supplies 18 per cent, the nonseparable activity 25 per cent, and the excited activity 57 per cent of the total activity of radium. The excited activity produced on bodies has been shown to be due to a deposit of radioactive matter on their surface. The term 'excited activity' refers only to the radiations from this active matter. It is convenient to have a definite name for the matter itself. It is suggested that the name 'emanation X' be given to it, since the matter which causes excited activity is produced direcdy from the emanation. This name is given from analogy to the products UrX and ThX, which are produced directly from uranium and thorium respectively. On this nomenclature, the radium produces the emanation at a constant rate and this, in turn, is transformed into the emanation X. The matter of emanation X of radium itself undergoes at least three and probably four successive changes. The nature of these changes and their connexion with the radioactivity will be discussed later. On heating or dissolving a radium compound in an open vessel, the emanation is released and can be entirely removed by a current of air. The emanation X, which is non-volatile, is left behind with the radium, and it at once commences to lose its activity. In the course of a few hours the activity due to it has practically disappeared. The ß and y rays which are produced only by emanation X disappear from the radium at the same time, and there then remains a non-separable activity of radium consisting entirely of IX rays. At the same time that the emanation X, left behind in the radium, is undergoing change, fresh emanation X is being produced by the separated emanation, and at such a rate that the activity at any time due to the emanation X left in the radium, together with that due to the emanation X formed afresh by the emanation, is equal to the original activity of the emanation X stored up in the radium. Since fresh emanation is being continually produced by the radium and occluded in it, the activity ofthe radium after falling to its minimum gradually rises again, and in the course of about a month has nearly reached its original constant value. The experiments which will now be described were undertaken to see if tbe heat emission of radium varied in tbe same way as its activity when tbe emanation was removed. For tbis purpose, the heating effect ofthe radium was first determined. Tbe emanation was then removed from it and collected by condensation in a small glass tube, and the distribution of the heating effect between the emanation and emanation X and the radium was determined, and also the variation with time of the heating effect of both the emanation and the radium from which it was separated.

lIearing Eflerf o/the Radium Emanation

627

Description 01 Apparatus As only about 30 milligrams of pure radium bromide were available in the experiments, special methods were devised to measure with accuracy the small heating effects involved. The maximum heat emission from 30 milligrams of radium bromide is not more than 6 x 10-4 gram-calories per second, and it was necessary to measure with certainty a heat emission of at least -h of this amount. In the following experiments a simple form of differential air calorimeter was employed. In Fig. 1 is shown a simple sketch of the apparatus. It

Fig. 1. Differential Air Calorimeter consisted of two I-litre flasks with rubber stoppers through which passed three-way glass cocks, and tubes into which the radium or the emanation tube could be lowered. The glass cocks connected the air in the flasks either to the outside air or to a manometer tube which registered the differences in pressurc. Some little difficulty was encountered at first in selecting a suitable liquid with sufficient mobility and low vapour pressure for the manometer tube. We finally used xylene, which proved in every way satisfactory. The difference in level of the xylene standing in the two arms was observed with a microscope provided with a micrometer eyepiece mounted on a cathetometer stand. The two flasks were immersed in a constant-temperature water bath, and remained very steady throughout the work. Readings were taken of the position of the xylene in the manometer tube with the radium or emanation tube in one flask, and then again when transferred to the other flask, allowing in each case ample time for thc attainment of steady

628

The Collected Papers 0/ Lord Rutherford

conditions. To calibrate the readings of the micrometer scale direcdy in gram-calories per hour, two colls of manganin wire were constructed of about 50 ohms each. One was made in a small compact shape of about the same volume as the 30 milligrams of radium bromide introduced, while the other was made of many turns of wire on a frame of approximately the same shape as the emanation tube. Currents of known strength were sent through the heating coils, and the corresponding differences of pressure observed on the manometer. Thus two calibration curves were obtained, which differed a little but represented as nearly as we could arrange the two sources of

Fig. 2. Differential Inner-coll Platinum Thermometers in Water-bath heat. Although the differential air calorimeter did very weIl for the first observations over the whole range of several weeks, and was very suitable for measuring the quantity of heat emitted, on account of its large mass and the volume of air affected, it did not respond quickly enough to obtain the important initial changes. Two pairs of sensitive inner-coil platinum thermometers were therefore constructed, having a lag of not more than 6 or 7 min. The two pairs enabled us to work with the radium and the emanation tube at the same time, which was important for the first changes. Each thermometer consisted of 35 cm. of fine platinum wire wound carefu1ly on the inside of a thin glass tube 5 mm. in diameter. To secure the wire, the tube was gently warmed to a low red heat until the wire stuck fast to

Heating Effect

0/ the Radium Emanation

629

the g]ass, forming a coil about 3 cm. long. The ends of the coll were gold soldered to heavy platinum-wire leads. The tube containing the radium, as weIl as the tube containing the emanation, was selected so as to slide easily into the interior of the coil, the wire thus being in direct contact with the g1ass envelope containing the source of heat. Each thermometer was fitted centrally in a larger glass tube, passing through a long narrow water bath (Fig. 2). Astirrer in the water bath kept the temperature always uniform. The changes in the resistance of the thermometers when the radium or emanation tube was transferred from one to the other were measured on a Callendar type of compensated resistance box constructed with great care. This box has already been described by one of us* in another place. The differences were observed on a scale-and-vernier reading to the l~o of a millimetre. The greatest difference observed amounted to somewhat over 7 mm., and corresponded to a difference of temperature of about ~oC, and readings could be made with certainty down to one-hundredth ofthis. We did not aim to obtain quantitative measurements of the heat evolved, as we were directly concerned with the initial portions of the curves, for which relative values were all that were required. As a check, however, we calibrated one pair of the differential thermometers, and found a very c10se agreement for the heat emission of 30 milligrams of pure radium bromide, with the value obtained by the air calorimeter. This amounted to a little over 100 gram-calories per hour for I gram of pure radium, which agrees very closely with the values given by Curie and Laborde, and later by Runge and Precht. t M easurements 01 Radioactivity

In the experiments, the emanation was driven off from the radium by heating it, and the emanation was then condensed in a small g1ass tube immersed in liquid air. It was important to know how mueh of the emanation was removed by the heating, and how mueh of the emanation, whieh was driven off from the radium, was colleeted in the condensing tube. This was done in the following way. The tube containing the radium was plaeed behind a lead sereen 5 em. thick, placed near a cylindrical metal eleetroscope, in which the gold-Ieaf system was insulated by a sulphur bead, after the manner first employed by C. T. R. Wilson. The rate of movement of the charged gold-Ieaf was observed by means of a mieroscope with a micrometer eyepiece. A rapid rate of movement of the gold-Ieaf was observed due to the y rays from the radium after passing through the lead sereen 5 em. thick. Preliminary experiments by one of us showed that the y rays always accompanied the ß rays from radium and were proportional in amount to them. If the radium was completely de-emanated, after sufficient time has elapsed for the excited • H. T. Bames, Phi!. Trans., vol. cxcix, p. 185 (1902). 38 (1903).

t Sitz. Akad. Wiss. Berlin. No.

630

The Collected Papers of Lord RutherJord

activity to decay, the 'Y rays, as weIl as the ß rays, almost completely disappeared. If the radium, four or five hours after de-emanation, showed more than 1 or 2 per cent of the original rate of discharge measured by the 'Y rays, it was concluded that a portion of the emanation had not been removed. * The proportion of the emanation left behind was in such a case directly proportional to the r rays, since the excited activity and consequently the rays left behind, after the lapse of four or five hours, is practically proportional to the emanation present. If no emanation had been lost, the 'Y rays from the radium and the emanation tube together should, at any time, be equal to that from the original radium. If this were not the case, it showed that a portion of the emanation had escaped into the pump, and the amount so lost would be calculated from the difference in the rates of discharge of the electroscope. These measurements could be made with rapidity and accuracy, and served as a guide to the amount of emanation present in a vessel. It must be borne in mind that the 'Y rays emitted only serve as a measure of the amount of emanation present, when sufficient time has elapsed (3 to 4 hours is enough) for the emanation and its product emanation X to reach astate of approximate radioactive equilibrium. Immediately after the introduction of the emanation into avesseI no ß or 'Y rays are observed, but the intensity of the ß rays and 'Y rays reaches about half their maximum value 45 minutes later.

Description of Experiments The heating effect of 30 milligrams of radium bromide, enclosed in a narrow glass tube, was first determined in the air calorimeter. The radium tube R (Fig. 3) was then connected through a phosphorus-pentoxide tube A to a short narrow glass tube T, about 3 cm. long, connected to a mercury pump P. The tube T passed through a small liquid-air vessel, made of ebonite, in order to condense any emanation passing through it. Liquid air was placed in the vessel, and the tubes partially exhausted. The radium was then heatedt with a spirit lamp to drive off the emanation. The water-vapour given off was absorbed in the tube P and the emanation was condensed in the tube T, and the whole emanation given off was condensed in the tube T by slowly working the pump. The tube T and the radium tube were then sealed off, and the heating effect of each tested as soon as possible afterwards in the air calorimeter. The results obtained are shown in Fig. 4. The heating effect of the radium, when first tested, • In the first heating of the radium bromide to a temperature considerably below that of a red heat, the emanation was completely released. A second heating, after the radium bad recovered its activity, only removed 75 per cent of the occluded emanation. t When the radium was heated to about a red heat a very bright phosphorescence was produced in the radium compound. Tbe luminosity was bright enough to observe in ordinary daylight and persisted for several days. Its spectrum was kindly photographed fOT us by Dr Schenck and was found to be continuous.

lIeating Effect

0/ the

631

Radium Emanation

had fallen considerably, and continued to do so for about three hours, when it reached a minimum corresponding to about 30 per cent of the original value. At the same time the emanation tube, when first tested, showed considerable heating effect. This increased for about three hours, when it reached a maximum. After reaching the minimum, the radium

A

LlctUIDA

~o

P

LlctUID AIR

RADIUM

P

Fig. 3. Method of Condensing the Emanation from the Radium 2~4

~

~

~

RadiulI

....t:t... ,.&

~

t3

~

~

~

.~

~

'8

.~

::: ~ ~" -tJ.

o

80

2+D

1&0

Hours Fig.4

310

PUMP

632

The Collected Papers 0/ Lord Rutherford

gradually recovered its heating effect, rising to its original value, after about a month's interval. At the same time the heating effect of the emanation tube gradually diminished according to an exponentiallaw with the time, falling to half value in about four days. Within the limit of experimental error the sum total of the heating effect of the radium, together with that of the emanation tube, was, over the whole course of the experiment, always equal to that of the original radium. Measurements of radioactivity showed that about 6 per cent of the emanation in the above experiment was not released from the radium by the heating. We may conc1ude from these results that about 75 per cent of the heating efl'ect observed from radium is not direct1y due to the radium, but to the emanation and emanation X, which it produces from itself. There is a c10se connexion between the variation of the radioactivity of the radium and its rate of heat emission. After separation of the emanation, the activity of the radium falls to a minimum of about 25 per cent in the course of a few hours, and then gradually increases again. At the same time, the activity due to the emanation (observedin a closed vessel) increases with the time, on account of the excited activity produced by the emanation on the walls of the vessel. The curves of recovery of the heating effect of radium, and the gradual decrease of the heating effect of the emanation, are almost exact1y the same as the corresponding curves for the recovery of activity of the radium and the loss of activity with time of the separated emanation. The rate of heat emission of the emanation, like its activity, falls to about half value in about four days. Half of the lost heating effect, as well as half of the lost activity of the radium, is spontaneously recovered during the same interval. The curve of diminution of the heating effect of the emanation tube is thus expressed by the equation:

Q, Qo

= e-ÄI

'

where Q, is the rate of heat emission at any time t and Qo the maximum rate. The rate of heat emission Q, of the radium, at any time t after reaching its minimum value, is expressed by the equation:

~: = 0·25 + 0·75(1 -

e-A'),

where Qo is the maximum rate and ,\ the same constant as before. The same numerica1 constants as well as the same equation are obtained for the recovery curves of activity of de-emanated radium, measured by the IX rays (see Rutherford and Soddy, Phi!. Mag., April 1903). These results, as far as they go, are in agreement with the view that the heating effect of radium is, at any time, proportional to its activity measured by the IX rays; or, in other words, that the heat emission of radium is an accompaniment of the expulsion of IX particles. The heat emission is not

Heating EfJect

0/ file Radium Emanation

633

proportional to the ß or y rays; for in the experiment the amount of ß and y rays remaining in the radium, after partial de-emanation, was only 6 per cent of the total, while the heating effect was about 30 per cent of the total. These experiments are not, however, sufficient to justify us in concluding that in all cases the heat emission accompanies the expulsion of a; particles and is proportional to the number expelled. It will be necessary, in addition, to show that the emanation and emanation X each supplies an amount of heat proportional to its activity, measured by the a; rays, and also that the heating effect of each of the succession of changes of emanation X is always proportional to its activity, measured by the a; rays.

Decrease 0/ Heating Effect of Radium immediately after De-emanation The air calorimeter, already described, is not suitable for rapid observations of the variation of the heat emission of the de-emanated radium or the emanation immediately after the latter has been removed. For this purpose the differential platinum thermometers, previously described, were used. The emanation was condensed in a small glass tube 3 cm. long, 3 nun. internal diameter. As the object of these experiments was to determine the

100%

l" ~ ~

~ ... 10

CI. Q IIC Cl

I§

""

60

21

40

50

10

Fig.5

IDO

1'20

"ND

...,uLnJ

o

MlNJ!OAtTIVlni

'ZD

MINI!' HfATING EFFE&T

40

14D MINuns

634

The Collected Papers

0/ Lord Ruthelford

initial drop of the heating effect of the radium, only a few minutes were occupied in heating the radium and condensing the emanation. In consequence of this, a small amount of the emanation was carried over into the pump and was not condensed in the tube. The radium tube was rapidly removed and placed inside the air bath, and observations made on its heating effect, as soon as its temperature was steady. The results are shown in Fig. 5, curve B. The heating effect of the radium about 10 minutes after removal of the emanation had fallen to about 45 per cent of the original value. It then decayed more slowly to a minimum corresponding to about 25 per cent of the original value. The gradual decay of the heating effect to a minimum is connected with the removal of the emanation and consequent decay of activity of the emanation X left behind. The curve of decrease of the heating effect to a constant minimum should thus be the same as the curve of decrease to zero obtained for the emanation tube, when the emanation is removed from it. The emanation was allowed to stand four or five hours in the emanation tube, so that the excited activity on the walls of the tube had reached a practical maximum. The ends of the tube were then broken, and the emanation rapidly withdrawn. The heating effect of this tube was then determined at regular intervals. The results are shown in Fig. 6, curve B. The heating

lOGt 80

a.

~

z:

&u

CI CI

60

.,.

::I

4D

20

a

20

4D

&0

10

Fig.6

100

110

140 Mn,uns

Healing 1:.JIecl al/he Radium Emanation

635

cffect dropped rapidly during the first ten minutes after the removal of the emanation, then more slowly, and finally diminished approximately according to an exponential law with the time, falling to half value about every 30 minutes. The curve of decrease of the heating effect to zero is about the same as the curve of decrease of the heating effect of the radium to a constant minimum.

Increase 01 Heat Emission after the Introduction of the Emanation into a Tube If the total rate of heat emission is not altered by the transfer of the emanation from the radium to the emanation tube, the curve showing the rise of heating effect of the emanation tube should be complementary to the curve showing the decrease to a minimum of the heating effect of the radium after the removal of the emanation. This was found to be the case. The results are shown in Fig. 6, curve A. * As the emanation tube, during the condensation, was cooled to the temperature of liquid air, it was not feasible to take observations on its heating effect after such a short interval as in the case when the emanation was withdrawn without changing its temperature. The results, however, showed that the heating effect of the emanation tube was about 75 per cent of its final value, after the emanation had been introduced into the tube 35 minutes. Over the range of observations, the curve of rise of the heating effect of the emanation tube and curve of decrease of the heating effect of the emanation tube, after withdrawal of the emanation, are complementary to one another. If the curves are plotted on the same scale the sum of the ordinates of the two curves, at any time, is constant.

Discussion

0/ Results

It has been shown in Figs. 5 and 6 that there is a very sudden drop of

the heating effect observed for both the emanation tube and the radium after removal of the emanation. For the purpose of comparison, the decay curve of the activity of radium after removal of the emanation is plotted alongside the curve of decrease of the heating effect of radium (see Fig. 5, curve A). It is seen that immediately after removal of the emanation there is a sudden drop of the activity to 82 per cent of the maximum value. This shows that the radiation from the emanation supplies about 18 per cent of the total activity of radium, measured by the IX rays. There then follows a fairly rapid decrease for 6 to 8 minutes, and then a more gradual decay to a minimum of 25 per cent. This rapid variation of the excited activity for the first few minutes is due to the fact that the first change in emanation X takes place very rapidly. The effect of tbis first change is clearly shown when a rod is • Three sets of curves were obtained, of the increase of the heating effect after introduction of the emanation, and of the decrease of the heating effect after removal of the emanation. The curves in all cases were in good agreement.

636

The Collected Papers ofLord Ruthelford

exposed for a very short interval in the presence of the radium emanation. On removal of the rod the activity decreases at first very rapidly, falling to about half value in about 3 minutes. After falling below 20 per cent of the maximum value the activity varies very little for aspace of about 25 minutes, when there follows a gradual decay to zero (see Rutherford and Miss Brook, Phi!. Mag., July 1902). P. Curie and Danne* have shown that the decay of activity of a body exposed for a long interval in the presence of the emanation is given by the equation I 10 = ae-A•t -(a - l)e-A1t, where I t is the activity at any time t, and 10 the activity immediately after removal; Al = 1/2420, A2 = 1/1860, where a second is taken as the unit of time; the numerical constant a = 4· 20. Curie and Danne state that this equation holds accurately over the whole period of decay; there is no doubt, however, that there is a rapid initial drop for ab out the first ten minutes after removal which is not expressed by this equation. The equation will, however, hold if the initial observations of the activity start from aperiod about ten minutes after removal, as the first change is almost completed by that time. The equation will also hold, with the same limitations, for the decay of the activity of the emanation X, left behind in the radium after removal of the emanation. An analysis of the decay curves of excited activity, produced for different intervals of exposure in the presence of the emanation, shows that there are three well-marked changes occurring in emanation X of radium. In the first change, half the matter is transformed in 3 minutes; in the second, half in 34 minutes; and in the third, half in 28 minutes. A full account of the analysis of these changes and their peculiarities will be given by one of us in a later paper. The first change is accompanied only by cx rays, the second change is not accompanied by cx rays at an, and the third change by cx, ß, and 'Y rays. While the curves of decrease of the rate of heat emission and of the activity of radium are very similar in shape, it is not possible to deduce directly from the result how much of the comparatively sudden drop of the heating effect observed is due to the emanation and how much to the first change in emanation X, on account ofthe rapid first change in the latter. The activity of the first change of emanation X (half value in 3 minutes) will have been reduced to about 10 per cent of its original value in 10 minutes. The first certain observation of the heating effect of the radium in the emanation tube could not be made until about 10 minutes after removal of the emanation. It is essential to allow an interval of at least 5 minutes in order to allow the glass envelope to take up the temperature corresponding to the source of heat inside it. • Comptes Rendus, cxxxvi, p. 364 (1903).

637

I/eating Elfcrt oJ the Radium Emanation

On account of the rapid initial change in emanation X, we can only conclude from the experiments that about 41 per cent of the total heating effect of radium is due to the emanation together with the first change in emanation X. The other results all indicate that the heating effects accompany the emission of oe rays; for example, the minimum heating effect due to deemanated radium is 25 per cent of the total, and the activity is also 25 per cent of the total activity, and consists entirely of oe rays. Both the radiation from the emanation and that from the first change in emanation X consist entirely of oe rays; and there can be Httle doubt that the emanation does give rise to a heating effect of the same order of magnitude as is observed in the other changes, which are accompanied by oe rays. Further experiments are, however, in progress to separate, if possible, the heating effect of the emanation from that due to the first change in emanation X. The decrease of the heating effect due to the excited activity after the first sudden drop is not very rapid at first, and, in this respect, c10sely resembles the decay of the excited activity. It thus appears probable that the second change in emanation X, which does not give rise to oe rays, is also not accompanied by the same emission of heat as in the other changes. Further experiments are in progress to settle this point. The rate of he at emission of radium, when in astate of radioactive equilibrium, is thus made up as folIows: Active products

Nature of rays

Radium (freed from active products) ~

Emanation

Percentage Percentage proportion of proportion of activity measured total heating by oe rays effect

oe

rays

25

oe

rays

25

~

Emanation X (first change)

oe

~

Second change ~

Third change

rays

No oe rays oe,

ß, and

y rays

25

25

25

25

25

25 25 25

Although the experimental results are not yet complete enough to settle definitely whether the heating effect of each of the products is proportional to its activity, measured by the oe rays, the results, as far as they have gone, indicate that this is approximately the case. The heating effect is certainly directly connected with the radioactivity of radium, and the time variation of the heating effect of each of the radioactive products is the same as the time variation of the activity. This result shows that the heat emission of

638

The Collected Papers of Lord Rutherford

radium 1S an accompaniment of the successive changes occurring In radium. It still remains to be shown how much of the heat emission of radium is due to the kinetic energy of the cx partic1es, and how much to the atomic systems from which they are expelled. In a mass of radium nearly all the cx rays emitted from it are absorbed in the radium itself. The radium is thus subjected to an intense bombardment by the cx particles projected from its own mass. There is no doubt that a proportion of the heating effect of the radium is due to this self-bombardment, but probably, also, apart is due to the energy emitted consequent upon the rearrangement of the components of the atoms from which the cx partic1es are expelledo It is not to be expected that the division of heat between the two systems wou1d be the same for each active product, and, in consequence, that the heating effect of each product shou1d be accurately proportional to its activity measured by the cx rayso In the experiments made on the heating effect, the cx rays, in all cases, were absorbed in the glass envelope, and thus added their heating effect to that given out by the systems from which they were expelled. A large proportion of the ß rays and practically all the 'Y radiation escaped. It has, however, been shown* that, in all probability, the energy emitted in the form of ß and 'Y rays is only a small fraction of the energy emitted in the form of cx rays.

Amount 01 Heat from the Emanation It has been shown that the heating effect from the emanation, together with that of the active product to which it gives rise, is equal to about 75 per cent of the total heat emission of radium. Since 1 gram of radium emits heat at a rate of 100 gram-calories per hour, the emanation from 1 gram ofradium in astate of radioactive equilibrium, a few hours after its removal, radiates at the rate of75 gram-calories per houro Since the rate ofheat emission at any time (after it has reached a maximum value) is given by

Q, Qo

=

e-At '

the total amount of heat given out from the emanation released from 1 gram of radium is equal to

I° oo

Qoe-At dt

= Qo.

,\

Since the heating effect of the emanation decays to half value in about 3·73 days ,\ = O· 0077 when an hour is taken as the unit of time. The total amount of heat derived from the emanation from 1 gram of radium is thus about 10,000 gram-calories. Now it has been shown (Rutherford, Nature, August 20, 1903, p. 366) that the volume of the emanation released from • Rutherford and Grier, Phil. Mag., September 19020

lIeating Ellect

oI thc

Radium Emanation

639

1 gram of radium probably lies between 6 X 10-4 and 6 X 10- 5 C.C., at standard pressure and temperature. The amount of heat liberated per hour from 1 c.c. of the emanation would thus He between 1· 25 X 105 and 1· 25 X 106 gram-calories. This amount of heat from 1 c.c. of the emanation would probably be sufficient to raise to a red heat, if not to melt down, the glass tube containing it. The emanation behaves as if it were agas of heavy molecular weight. Assuming for the purpose of calculation that the moleeule of the emanation is 100 times as heavy as the moleeule of hydrogen, it can readily be deduced that 1 gram of the radium emanation, in its succession of changes, would radiate an amount of energy lying between 2 X 1()9 and 2 X 1010 gramcalories. One pound weight of the emanation would initially radiate energy at the rate of 104 to lOS horse-power, and, while the heating continued, would emit an amount of energy between 6 X 104 and 6 X lOS horsepower-days. Quite independently of any assumptions, a result of the same order of magnitude can be deduced from the observed heating effect of the emanation, and the fact that the emanation has not, so far, been detected either by its volume or its weight. There is thus no doubt that matter under special conditions is capable of emitting an amount of energy enormous compared with that released in the most intense chemical reactions. On the disintegration hypothesis (Rutherford and Soddy, Phi!. Mag., May 1903) this energy is derived from the energy latent in the radium atoms, and is released in the successive stages of their disintegration. McGiIl University, Montreal December 22, 1903

Nature of the

y

Rays from Radium

Frorn Nature, 69,1904, pp. 436-7

THE interesting results recorded by Mr Eve in the preceding letter on the relative conductivity of gases for very penetrating Röntgen rays removes the strongest objection that has been urged against the common belief that the y rays are an extremely penetrating type of Röntgen rays. All the experimental evidence so far obtained is now in agreement with the view that the y rays are very penetrating Röntgen rays which have their source in the atom of the radio-active substance at the moment of the expulsion of the ß or kathodic particle. For example, I have found that the y rays from radium always accompany the ß rays, and are always proportional in amount to them. In radium the ß and y rays appear only in the third change occurring in the radio-active matter which causes 'excited activity,' i.e. in the fourth of the chain of radio-active products which result from the disintegration of the radium atom. In addition, as Mr Ashworth pointed out in arecent letter to this Journal (January 28), the fact that the amount of y rays from radium is independent of its degree of concentration points to the conclusion that the y rays arise from the disintegrated atom, and are not secondary rays set up by the bornbardment of the radium as a whole by the ß rays. On the theory of the nature of Röntgen rays, developed by the late Sir George Stokes and Prof. J. J. Thomson, it is to be expected that Röntgen rays would be set up at the sudden starting as weH as at the sudden stopping of the electron or ß particle. As a result of the sudden expulsion of the ß particle from radium, it is to be expected that a narrow electromagnetic pulse, i.e. a 'hard' or penetrating type of Röntgen rays, would be generated. In addition, on account of the great speed of the ß particle, it is able to penetrate through a considerable thickness of matter before it is stopped. A broad pulse or 'soft' Röntgen rays should thus arise at the point of incidence of the ß rays. E. RUTHERFORD McGill University, Montrea1

February 18

DIFFERENTIAL AIR-CALORIMETER FOR DETERMINING THE HEA TING EFFECT OF RADIOACTIVE EMANATION

The similar lead cylinders surrounding the glass tubes are visible in each flask. See: 'Heating Etfect of the Radium Emanation' (H. T. Barnes), page 625, and 'Heating Effect of the V Rays from Radium' (H. T. Barnes), page 792.

This page intentionally left blank

The Radiation and Emanation of Radium Part I by

Macdonald Professor of Ph)'sics, McGill Unil'el'sity, Montreal

E. RUTHERFORD, M.A., D.Se., F.R.S.,

From Technics, July 1904, pp. 11-16 (Address to the Royal Institution, March 1904)

SCARCELY more than seven years have elapsed since Becquerel discovered that uranium possesses the property of spontaneously emitting radiations, capable of passing through substances opaque to light, of acting on a photographic plate, and of discharging electrified bodies. Nevertheless our knowledge of the subject has grown with surprising rapidity, and a very large amount of information has now been coUected, not only concerning the radiations from the radio-active bodies, but also of the complicated processes occurring in them. A theory has also been advanced which serves to connect in an intelligible manner the remarkable se ries of phenomena observed. Soon after Becq uerel's discovery of the radio-active properties of uranium, the rare element thorium was found to be radio-active. An examination ofthe mineral pitchblende has revealed the presence of aseries of new radio-active bodies, which possess the property of radio-activity to a very intense degree compared with uranium and thorium. Ofthese bodies, the substance actinium, discovered by Debierne, will probably prove to be a new element of activity comparable with that of radium; while the radio-active elements present in the polonium, discovered by Mme. Curie, and thc radio tellurium , discovered by Marckwald, have not yet been chemically isolated. The substance radium, discovered in 1898 by M. and Mme. Curie in pitchblende, has attracted the greatest amount of attention, partlyon account of the fact that it has been found to be a new element with adefinite spectrum and atomic weight; but chiefly on account of its surprising radio-active properties: for its radiations are nearly two million times as intense as those emitted from an equal weight of uranium. Much of our present knowledge of radio-activity has been derived from a study of the comparatively feeble radio-active bodies, uranium and thorium. The radio-active properties of thorium are qualitatively similar to those of radium, but differ very widely in intensity. In this article, for brevity, the properties of radium alone will be considered, although the explanation advanced for radium in many cases applies also to the other radio-active bodies. x

642

The Collected Papers of Lord Rutherford

The characteristic radiations from radium are invisible to the eye, but are, in part, transformed into light when they fall on certain fluorescent substances. An ordinary X-ray screen of barium platino-cyanide is rendered luminous in a dark room when the rays from radium fall upon it. The mineral willemite (zinc silicate) lights up brightly under the rays from a few milligrams of radium bromide; it appears quite translucent, emitting light of a beautiful greenish colour. Another mineral, kunzite, exhibits a beautiful rose colouration under the rays, but the intensity is not so marked as in the case of willemite or crystals of the platino-cyanides. The radiations from radium are very complex in character, and comprise three kinds of rays, called the a:, ß and y rays. The greater part of the rays emitted bear no resemblance to ordinary light waves, but consist of flights of material partic1es projected with enormous velocity. The nature of the ß rays was first determined, on account of the ease with which they are defiected by a magnet. If a small quantity of radium bromide is placed at the base of a smalliead cylinder, the cone of rays issuing from the open end produces a luminous patch on a willemite or X-ray screen brought near it. The luminosity observed is mainly due to the ß rays. On bringing up a magnet near the radium, the light patch is observed to be defiected and also to be greatly broadened. By reversal of the magnet, the direction of movement is reversed. The broadening of the cone of rays in a magnetic field shows that the rays emitted are complex in character, some of them being more readily deflected by a magnetic field than others. In a similar way, it has been found that the rays are defiected in passing between two parallel plates kept charged to a high difference of potential by an electric machine. In these and other respects, the ß rays are identical with the rays which are shot off from the cathode when a strong electric discharge is passed through a vacuum tube. By measuring the amount of deflection of the cathode rays, in passing through both a magnetic and electric field of known strength, J. J. Thomson has shown that they consist of a Hight of particles, carrying negative charges of electricity, and projected with a velocity of about 50,000 miles a second. The particle shot off from the cathode is the smallest body known to science, for apparently its mass is only about 10100 of that of the hydrogen atom. The ß partic1e shot out from radium is identical in size with the cathode ray particle, but is projected with a much greater average speed. Kaufmann has shown that some of the particles possess an initial velocity of over 170,000 miles per second, i.e., a velocity very nearly equal to that of light. The y rays from radium are of an extraordinarily penetrating character, for their presence can readily be detected through several inches of lead or a foot of iron. They are not deflected by a magnetic or electric field, and there is now little doubt that they are a type of very penetrating Röntgen rays. According to the views of Stokes and J. J. Thomson, Röntgen rays are very short transverse waves or pulses in the ether, which are set up when the cathode ray particle is suddenly stopped by striking an obstac1e. It is to be

The Radiatio1l ami Emanation oJ Radium. J

643

expeded that Röntgen rays should be set up at the sudden starting as weIl as at the sudden stopping of a charged particle, and the present evidence points 10 the conclusion that the y rays arisc at the moment of the expulsion of the ß particle from the radium atom. Thus, unlike the ß rays, the y rays are not corpuscular in character, but are more akin to very short light waves. The IX rays have very slight penetrating power and are absorbed in their passage through a few centimeters of air or by a sheet of paper. They were for a long time thought to be incapable of deflection by a magnetic or electric field; but in 1902, the writer found that they could be bent from their path by the application of a very intense magnetic or electric field. The direction of deflection was, however, opposite to that observed for the ß rays, which showed that the IX rays carried with them, not a negative, but a positive charge of electricity. By observing the amount of bending of the rays in passing through a magnetic and an electric field of known strengths, it was deduced that the IX particles are projected with a velocity of about 20,000 miles a second, and have a mass about twice that of the hydrogen atom. If the IX particles consist of any known kind ofmatter, these measurements pointed to the conc1usion that they were either hydrogen or helium atoms, for next to hydrogen, helium has the lightest atom known to science. The significance of this result is seen in the light of the production of helium from radium, which will be considered later. The fundamental phenomenon of radio-activity consists in this continuous and spontaneous expulsion from radium of heavy particles of matter, atomic in size, with enormous velocity. Each partic1e has a velocity about 40,000 times as great as a rifle bullet, so that its energy of motion is enormous compared with the mass of the body in motion. The rays from radium are very similar to those produced when a strong electric discharge is sent through a vacuum tube. The ß rays are like the cathode rays, the y rays are similar to the Röntgen rays, while the oe rays are very analogous to the 'canal' rays discovered by Goldstein. In order to produce these rays in a vacuum tube, a large expenditure of electric energy is required. Radium, on the other hand, gives them out spontaneously, without, so far as is known, the action of any external exciting cause. The effect of a magnetic field in P P P P separating out the rays of radium is P shown diagrammatically in Fig. 1. The radium R is placed in a small lead vessel, and the oe, ß and y rays are P projected as a cone of rays from the opening. On applying a magnetic field P P at right angles to the plane of the paper, P and in a direction downwards through Fig. 1 the paper, the ß rays are bent to the

644

The Collected Papers oJ Lord Ruthe,ford

right, the Cl rays to the left, while the y rays are not bent at all. The amount of bending of the 0(, compared with the ß rays, is very much exaggerated in the figure. The comparative size, velocity and energy of motion of the Cl and ßparticles is shown graphically in Fig. 2. The energy of motion of the Cl particle is much greater than that of the ß particle. There is, in addition, evidence to show that four Cl partic1es are projected for every ß particle, so that nearly all the energy emitted from radium, in the form of rays, is carried off by the 0( particles. Not only are the Cl partic1es of more importance from the point of view of emission of energy, but they also play a far more important part in the complicated processes which occur in the radium atom. It can be calculated that every gram of radium projects about one hundred thousand million 0( particles per second, and yet there is such an enormous number of atoms in a gram of matter that MASS

VELOCITY

ENERGY

0{

fJ Fig. 2. The velocity is represented by the length cf a line and the mass and energy by spheres the process could continue for years before an appreciable proportion of the matter had been fired away. The rate of expulsion of the Cl or ß partic1es is not appreciably influenced by any physical or chemical agency under our control. At the temperature of liquid air, it proceeds at the same rate as at a red heat. The continuous expulsion of Cl particles from radium is very well illustrated by a beautiful experiment devised by Sir William Crookes. If a trace of radium is brought near to a zinc sulphide screen, the surface of the screen is rendered luminous by the Cl rays. On examining the screen with a magnifying glass, the luminosity is found to lack uniformity, consisting of a multitude of sparks of light coming and going with great rapidity. The action brings vividly before the observer the idea that the screen is subjected to an intense bombardment, the impact of every projectile as it strikes the target being marked by a flash of light. The massive Cl particles have such an enormous energy of motion, compared with their size, that they produce, at impact, an alteration in the crystals of zinc sulphide-probably a c1eavage-which is accompanied by a flash of light. The most important property possessed by the radiations from the radioactive bodies is that of discharging electrified bodies. This property has been

7h(' Radiation ami Emanation o( Radium. I

645

utilised as a means of accurate measurements of the radiations, and is by far the most delicate method of detecting the property of radio-activity in bodies. If radium exists in a gram of inactive matter, only to the extent of one part in ten thousand million, its presence can readily be detected by its power of discharging a gold leaf electroscope. This discharging action of the rays is due to the production of positively and negatively charged particles, or ions, as they are termed, in the electrically neutral gas through which the radiations pass. The oe and ß particles have such a large energy of motion that they produce ions by collision with the gas molecules in their path: each projected partic1e is capable of producing many thousands of ions before its energy is dissipated. These ions travel in the electric field and cause the loss of charge observed. If a trace of radium bromide is placed on a plate A (see Fig. 3) near a plate B, connected metallically with a charged gold leaf electroscope E, the

A ~IUM

A Earth

E

A

Fig.3 charge is very rapidly dissipated, and the leaves L rapidly collapse. Most of this discharging action is due to the oe rays. If a sheet of paper is placed over the radium, the oe rays are almost completely intercepted by it, and the discharging action is then due to the ß and 'Y rays. If the radium is completely covered with a lead screen ab out 5 millimeters thick, the ß rays are almost alt stopped, and the discharge in the electroscope is thus caused by the action of the 'Y rays alone. Speaking gene rally, the oe rays are far more effective in causing discharge, and producing luminosity in bodies, than the more penetrating ß and 'Y rays. On the other hand, the ß rays are the most active in darkening a photographic plate. Curie and Laborde recently showed that radium possesses the striking property of continuously keeping itself at a temperature of 4° or 5" F. above the surrounding air. This shows that radium, in addition to its radiations, continuously emits energy in the form of heat. The amount of heat from I gram of radium corresponds to about 100 gram-calories per hour. Apound of radium would emit per year about as much heat as is given out by the combustion of one hundred pounds of coal. This is a very remarkable result, for at the end of a year thc radium shows no apparent sign of alteration, and

646

The Collected Papers

0/ Lord Ruther/ord

still continues to give out heat at an undiminished rate. The rate of heat emission has been shown, by Curie and Dewar, to be unaltered when the radium is kept at the temperature of liquid hydrogen. This emission of he at is, in reality, directly connected with the radioactivity of radium, and is not an independent phenomenon. There is little doubt that the heat emission is a direct consequence of the continuous expulsion of IX partic1es from the radium. The IX partic1es are readily absorbed in their passage through matter; and in a pellet of radium, only a small percentage of the total number emitted are able to escape into the surrounding gas. The rest are absorbed in the mass of the radium itself. The radium is thus subjected to an intense and continuous bombardment by the IX particles projected from its own mass. Just as a target is heated by the impact of bullets upon it, so the radium becomes hot in consequence of its selfbombardment. A part also of the he at emitted probably results from the explosion in the atom, which results in the ejection of the IX partic1e.

,sIow ClJrrent o./oir

.B

'Thorium Hytoxide

ElECTR9SC9PE

Fig.4 Not content with emitting three kinds of rays and continuously giving off heat at a rapid rate, radium also produces from itself a radio-active 'emanation' or gas. This emanation is produced in minute quantity; but compared with the amount of matter, is an extraordinarily powerful radiator of energy. This property of giving off an emanation was first observed by the writer in the radio-active element thorium. If a slow current of air is passed through a tube T (Fig. 4) containing some powdered thorium oxide, or still better, thorium hydroxide; and is then passed through several yards of tubing opening into the electroscope E, the gold leaves of the electroscope are observed to rapidly collapse. The emanation, mixed with the air, is conveyed into the electroscope, and the radiation from it ionises the gas and causes the collapse of the leaves. If the current of air is stopped, the rate of discharge rapidly decreases with the time, falling to half value in an interval of one minute. If a radium solution is substituted for the thorium, a similar action is observed; but on account of the enormous activity of radium, the effects are far more intense. The emanation obtained from one millionth 0/ a gram of radium bromide causes the leaves to collapse in a few seconds, while the emanation from one ten thousand millionth 0/ a gram is readily measurable. This property of radium, of giving off an emanation, affords the most delicate

Thr Radiation and EmaIlation of Radium. I

647

and certain method of detection and estimation of minute quantities of radium. The emanation of thorium is very readily distinguished from that of radium, on account of the slow 10ss of activity of the latter. The radiating power of the emanation of radium falls to half value in four days, and is still appreciable after it has been stored up in a closed vessel for a month. The emanations of thorium and radium have been the subject of a large amount of investigation. They have been shown to diffuse through air like heavy gases, and to be unacted on by any known chemical reagents. Mr. Soddy and the writer found that the emanations of thorium and radium could be condensed, like gases, by the action of extreme cold. The thorium emanation condenses at -120 c., and the radium emanation at -150 C. With a large amount of radium emanation, the process of condensation can be followed by the eye. A simple experiment for this purpose is shown in Fig. 5. A quantity of emanation is collected in the small gasometer by 0

0

'Rc!dium &nanafj~

Willemite

U'luidAir:

Fig. 5. Condensation of radium emanation bubbling air through a radium solution. The emanation, mixed with air, is then passed through a gl ass U tube, immersed in liquid air contained in a Dewar fIasko In order to show the presence of the emanation, the U tube is partly filled with small pieces of willemite. As the emanation is slowly carried into the U tube, it condenses at the point of the tube just below the surface of the liquid air. The radiations from the condensed emanation render the willemite luminous; and, with a considerable quantity of emanation, the luminosity is very brilliant. With a very slow current of air, the luminosity is concentrated at the portion of the U tube where the air enters, while the rest of the tube is not luminous at all. On closing the ends of the tube and removing it from the liquid air, the emanation volatilises as soon as the temperature rises above -150'" C., and the whole tube becomes luminous. The light from the tube slowly diminishes, as the enc10sed emanation loses its activity, but is still appreciable after a month's interval. If the emanation, after being allowed to remain in the U tube for several hours, is blown out by a current of air, the luminosity of the willemite does not at onee disappear, but continues for several hours. This continued

648

The Collected Papers

0/ Lord RUfhel:{ord

luminosity is due to a remarkable property possessed by the emanation of manufacturingfrom itself a neu.' kind of radio-active matter which is not gaseous, hut behaves like asolid, and is deposited on the surface of bodies. A body

which has been exposed some time in the presence of the emanation, behaves as if it were eovered with an invisible film of intensely radio-active matter. This deposit is soluble in acids, is driven off at a white heat, and generally aets like a substance with definite ehemieal properties. The deposited matter eontinues to radiate for several hours. This power of 'exciting' or 'inducing' aetivity in neighbouring bodies is possessed by the emanations of both thorium and radium. The exeited aetivity produced by the thorium lasts much longer than that produced by the radium emanation, and is appreciable several days after the emanation has been removed. If the tube in whieh the radium emanation is stored is filled with layers of different substanees whieh are rendered luminous by the rays, the differences in eolour and intensity of the luminosity can be very c1early seen. Willemite shines with a beautiful green colour, and kunzite with a dark red colour. Zine sulphide shows a white luminosity. Crystals of barium platino-cyanide are at first as luminous as willemite; but the colour and luminosity of the crystals rapidly change, due to a ehemical action of the rays upon them. The luminosity of willemite and zinc sulphide is mostly due to the oe rays, while kunzite is more sensitive to the ß and y rays. A tube containing the radium emanation has temporarily all the properties of radium. It gives out ß and y rays, and also emits heat at a rapid rate. The emanation, when collected in a concentrated form, is luminous in the dark and causes glass to phosphoresce. It rapidly blackens the walls of the glass tube containing it, and is able to produce marked chemieal action in some substanees. If colleeted over water, it manufaetures hydrogen and oxygen at a rapid rate. All these effeets are produeed by an extraordinarily small amount ofthe emanation. Sir William Ramsay and Mr. Soddy have reeently succeeded in isolating the emanation from radium, and in determining its volume. They calculated that one gram of radium eontains about one cubic millimeter of the emanation, when measured at atmospheric pressure and temperature. The emanation has all the ordinary properties of agas. It has adefinite spectrum, but it differs from all other gases, inasmueh as its volume diminishes with time: it is not permanent, but is continuously ehanging into asolid substance, whieh is deposited on the walls of the containing vessel. But the most remarkable feature of the emanation is its enormous power of radiating compared with its weight. Dr. Barnes and the writer recently showed that three-quarters of the total heating effect of radium is due to the small amount of emanation stored up in it. The emanation was removed from the radium by heating it, and then collected by eondensation in a small glass tube immersed in liquid air. The tube containing the emanation glowed in the dark and gave off heat at a rapid rate, while the radium from whieh the emanation was separated lost three parts of its heating effeet. The emanation lost its heating effect at the same rate as it loses its activity, i.e., the heating

Ihe Radia/ion ami Emanation

0/ Radium. 1

649

ctTcct diminished to half value in four days, and had nearly disappeared in the course of a month. The radium at the same time spontaneously regained its power of heat emission, and after a month's interval gave out heat at the same ratc as before the experiment. Not only does the emanation itself give out heat, but the matter which is deposited from it on the walls of the vessel does so also. The heating effect was found to depend upon the emission of IX partic1es, and is direct1y connected with the radio-activity. The greater portion of the heating effect of the emanation is due to the continuous bombardment of the walls of the tube and the contained gas by the oe partic1es, which are thrown off by the emanation, and the other substances which are produced from it. If one cubic inch of the emanation were collected, the heat energy from it would be sufficient to melt down the walls of the glass tube containing it. It can readily be deduced that, weight for weight, the emanation emits about a million times more energy than that produced in the most violent chemical reaction. Apound weight of the emanation, immediately after its separation, would give out energy at a rate of about 10,000 horse-power. The rate of emission of energy would decrease to half value in four days, and would be appreciable after a month's interval; but thc emanation, during the time its activity lasts, would give out an amount of energy equivalent to an engine working at 10,000 horse-power for six days. Jf it should ever be possible to obtain a large quantity of this emanation and to utilise its energy, a few pounds of it would suffice to provide enough power to drive a liner across the Atlantic. As it would probably require about fifty tons of radium to produce one pound of the emanation, the outlook for the utilisation of the emanation as a source of power is not at present very promising.

x'"

The Radiation and Emanation of Radium Part II by E. R UTHERFORD, M.A., D.se., F.R.S., Macdonald Professor 0/ Physics, McGill University, Montreal

From Technics, August 1904, pp. 171-5 (Address to the Royal Institution, March 1904)

THE effeet of the eontinuous produetion of the emanation on the radioaetivity of radium will now be eonsidered. Solid radium bromide, in a dry atmosphere, gives off very little emanation into the air. This is not due to stoppage of the produetion of the emanation, but to its inability to eseape from the mass of radium. The emanation is stored up or occluded in the solid, and can only be released by heating or dissolving the radium. Suppose that a small quantity of solid radium bromide, some time after it has been made, is heated in order to drive off the emanation, whieh is then drawn off and mixed with air in a elosed vessel. The aetivity of the radium is at onee diminished, and in the course of four or five hours reaches a minimum value-the aetivity, measured by the (X rays, possessing only one-quarter of its original value. This gradual deeay of the aetivity to a minimum is due to the dying away of the exeited aetivity produced in the mass of the radium by the emanation stored in it. At the same time, the radium has almost eompletely lost, for the time, its power of emitting ß and 'Y rays. This loss of aetivity by the radium is not permanent, for the radiating power of the radium spontaneously inereases again, and, after a month's interval, has nearly reaehed its original value. Let us now eonsider the emanation whieh was separated by heating the radium. In a c10sed vessel, its power of discharging an electrified body at first inereases for several hours. This is due to the excited aetivity produeed by the emanation on the walls of the eontaining vessel. The emanation loses its aetivity in a geometrieal progression with the time. The aetivity falls to half value in four days, a quarter value in eight days, and so on. The eurve of deeay of aetivity of the emanation is shown graphically in Fig. 6. While the emanation is losing its aetivity, the radium from which it was separated is reeovering its lost aetivity. This is due to the eontinuous production, by the radium, of emanation which is retained as already described: sinee the emanation itself radiates, the activity of the radium

The Radiation ami Emanation

0/ Radium.

lJ

651

increases with the time. The curve of recovery of activity, due to the growth of fresh emanation, is shown in the same figure. The two curves of decay and recovery are complementary to each other. When the emanation has lost half its activity, the radium has spontaneously regained half of its lost activity. The sum total of the activity of the separated emanation, together with that of the radium from which it has been removed, is always equal to that of the original radium. lt would appear as if the separated emanation and the radium were connected by some subtle and intricate mechanism, so that any decrease of the radiating power of the one is compensated by an equal increase in the other. The connection, however, becomes c1ear if it is remembered that radium is always manufacturing fresh emanation at a constant rate, and that the emanation is always losing its activity in

RADIUM EMANATION

8

8

12

DAYS

16

20

Fig. 6. Curves showing decay of activity of the emanation, and recovery of activity of radium consequence of changes occurring in it. When the emanation is driven off from the radium, fresh emanation is produced and stored up in the radium. If the emanation did not lose its power of radiating with the time, it would be expected that the activity of radium would continue to increase steadily with the time. Since, however, the emanation is always changing and losing its activity, a stage must be reached where the production of fresh emanation just compensates the loss of activity of the emanation stored up in the radium. Just as the population of a country remains constant when the number of births is equal to the number of deaths, so the activity of radium reaches a steady limiting value when the number of partic1es of emanation produced per second is exactly equal to the number of particles which lose their power of radiating in the same time. In this continuous production of new kinds of active matter, which only radiate for a limited time, lies the key to the explanation which has been advanced to account for the phenomena of radio-activity. What process, occurring in radium, can account for its remarkable behaviour, for the

652

The Collected Papers

0/ Lord Ruthelfol'd

spontaneous and unceasing projection of IX and ß partic1es, the rapid emission of heat, and the production of agas endowed with new and surprising properties? It is obvious that some sort of chemical explanation is necessary to account for the appearance of new kinds of matter. But, on the other. hand, the laws which control the production of this matter are very different from those of ordinary chemical change. Temperature, which has such a marked infiuence on ordinary chemical reactions, has no appreciable effect in changing the processes occurring in radium. The activity or heating effect is not affected by plunging radium into liquid air; and, so far as observation has gone, the rate of decay of activity of the emanation is unaltered by the most drastic physical and chemica1 treatment. If, however, it is supposed that the changes occur in the atoms of the radio-element itself, it is not to be expected that temperature would have much influence, for the experience of chemists in failing to break up the atoms into simpler forms, shows that wide changes of temperature have little effect in altering the stability of the chemical atom. The theory that the phenomena of radio-activity are due to the disintegration of the radio-atoms was advanced more than a year ago by Mr. Soddy and the writer. This theory accounts in a simple way for all the complicated phenomena manifested by radio-active bodies, and welds together aseries of apparently disconnected facts. Let us, for brevity, consider the application of this theory to explain the properties of radium alone. It is supposed that a very small number of the radium atoms-about one in every hundred thousand million will suffice-become unstable every second and break up with explosive violence. Apart of the atom-the IX particle-is expelled with great velocity. The expulsion of a particle which has a mass about twice that of the hydrogen atom, leaves the radium atom lighter than before, and must change its chemical and physical properties. The radium atom minus the expelled IX partic1e, on this view, constitutes the atom of the 'radium emanation'. The atom of the emanation is also unstable; and, on an average, half of the total number produced break up in four days, each atom as it breaks up expelling another IX partic1e. The emanation as a result is changed into another type of matter-emanation X, as it has been named-which behaves as asolid. This, in turn, is unstable, and the process of disintegration goes on from stage to stage till in the fifth change, in addition to the IX partic1e, the ß partic1e is thrown off with its accompaniment, the y ray. The different substances produced as a result of the disintegration of the radium atom, together with the nature of the rays emitted at each change, are shown diagrammatically in Fig. 7. At least five distinct substances are produced as a result of the disintegration of the radium atom. The emanation is a chemically inert gas, while the succeeding products behave like metallic substances which are readily soluble in some acids, and are volatilised by heat. Each of these substances differs from an ordinary chemical element, inasmuch as it is not

Thc Radiation and Emanation

0/ Radium.

653

If

permanent, but is continuously and rapidly changing into another kind of matter. The products of the disintegration of the radium atom may thus be considered as transition elements which have a very limited life. Each of the products is transformed according to adefinite law, and at a perfectly definite rate. After any interval t, the number of atoms, N, of any given kind 0(

0(

0(

ß I"

0(

x

0(

X

FINAL

XI 3 "4PRODUCT RADIUM EMANATION XI Fig. 7. Diagram to represent the disintegration of a radium atom of matter which remains unchanged, is given accurately by the equation, N = NOE- A1 , when No is the initial number of atoms present, ,\ is the constant of change, and E is the base of the Naperian logarithms. The time, T, which elapses before half the amount of each transition element is transformed is shown in the following table:Name of Substance

Time, T

Remarks

4 days

1st product

3 minutes

2nd product

Emanation X (2nd

21 minutes

3rd product

Emanation X (3rd

28 minutes

4th product

Emanation X (4th

very slow

5th product

Radium Emanation Emanation X (Ist change)

Final product The transformations of each of the products of radium, with the exception of the third, are accompanied by the emission with great velocity of oe partic1es alone (Fig. 7). It is remarkable that the ß and y rays are only emitted during the changes in the fourth product, i.e., during the last rapid change which takes place in the radium atom.

654

The Collected Papers 01 Lord Ruthelford

The fifth product of radium differs from the others in its extremely slow rate of change. The evidence, so far obtained, points to the conc1usion that probably several hundred years would be required before half this matter is transformed. Since this substance is being continuously produced by the radium, and changes very slowly, it should graduaJly collect in some quantity in matter which contains radium. There is at present a good deal of evidence that the radio-active substance, separated from pitchblende by Marckwald, and called by hirn radio-tellurium, is in reality the fifth product of the disintegration of the radium atom. Since the radium products are unstable and rapidly breaking up, we cannot regard them as identical with any known kind of stable matter. Each of the radio-active products is found to possess distinctive chemical and physical properties which serve to distinguish it, not only from the preceding and succeeding products, but also from the parent element. The emanation is a heavy gas, which can be condensed by the action of extreme cold: like the gases of the helium-argon family, it is chemically inert. On the other hand, the solid products derived from the emanation are soluble in some acids; they are volatile at high temperatures, and can be partially separated from each other by their difference in volatility or by e1ectrolysis. With the exception of the emanation, none of the products of radium has been collected in sufficient quantity to be examined by direct chemical methods. The amount of each product to be obtained from a given quantity of radium depends on the rate at which it breaks up. When astate of radioactive equilibrium is reached, the number of atoms of any product appearing in unit time must equal the number disappearing; but, since the products break up at different rates, the number of atoms must be greater in the case of the more slowly changing product. It can be calculated that the weight.of emanation (half of which breaks up in four days) obtainable from 1 gram of radium bromide is about 1/100 of a milligram; while the weight of the fourth product (half of which breaks up in twenty-eight minutes) is about 3/100,000 of a milligram-a quantity too smal1 to be detected by the chemical balance. Thus it appears that the radio-active products resulting from the disintegration of the radium atom can never be collected in any great quantity on account of their limited life. The inactive products, however, must increase in quantity so long as there is any radium present, and we may hope to find them in radium ores in measurable quantities. The inactive products of radium are the expelled (X partic1es and the final product. Now the mineral pitchblende, in which radium occurs, contains in small quantity a large proportion of the known elements; the presence of the rare gas helium is noteworthy. Helium is only found associated with the radio-active minerals, and its presence in them has always been a matter of surprise. Mr. Soddy and the writer suggested, in 1902, that helium might prove to be a disintegration product of the radio-elements. This hypothesis

The Radiation ami Emanation oI Radium. Il

655

received strong support from measurements of the mass of the projected O!: particles of radium, for, within the limit of experimental error, these appear to have about the same mass as the helium atom. This suggestion has been verified in a brilliant manner by the recent experiments of Sir William Ramsay and Mr. Soddy. They found that helium always appeared in a closed tube in which the radium emanation had been stored for some time. The results indicate that helium is produced from the emanation as it gradually breaks up and disappears. There has been a tendency to assume that helium is a final product of the disintegration of the atom of the emanation, but the evidence so far obtained rather points to the conclusion that helium is in reality the expelled IX partic1e. If this is the case, helium should be produced from each of the radium products which emit O!: rays; but the experimental difficulties in an investigation of this character are so great that progress must necessarily be slow. The production of helium by the radium emanation is of extreme importance, as it is the first well-authenticated case of the transmutation of one element into another. This process of transmutation is of a very special character, for it takes place spontaneously, and at a rate that is independent of our control. In order to explain the production of helium from radium on strictly chemical lines, it has been suggested that helium is not a true element, but is in reality an unstable compound of helium with some known or unknown element, and that this compound is steadily breaking up with the liberation ofhelium. It must be borne in mind, however, that this postulated compound is very unique in character, for it is necessary to suppose that, unlike any other molecular compound, it breaks up with the expulsion of charged particles moving with enormous ve]ocity, and that the energy liberated during these changes is about one million times greater than the energy liberated in the most violent chemical reaction. In addition, it is necessary to suppose that the process by which helium is liberated is unaffected by wide changes of temperature-a result never before observed in any chemical reaction. So [ar as observations have yet gone, radium is a true element in the ordinary accepted chemical sense. It has adefinite spectrum and atomic weight, and in chemical behaviour is closely allied to the well-known element barium. On the disintegration theory, the helium and the radio-active products appear as a consequence of the disintegration of the radium atoms. The difference in the two points of view is, to a large extent, one of nomenclature alone. The chemical atom is defined as the smallest chemical unit which enters into combination with other substances, and which cannot be broken up by the action of physical and chemical forces at our disposal. This is true, so far as we know, in the ca se of radium; for the breaking up that does occur is spontaneous, and cannot be accelerated or retarded by chemical or physical agencies. Taking into account the novel character of the changes occurring in radium, and the enormous emission of energy, it appears more reasonable

656

The Collected Papers

0/ Lord Ruthe/lord

to suppose that these appear as the result of changes of quite a new eharaeter in matter-a breaking up of the chemical atom rather than of the ehemical molecule. Sinee radium is continuously breaking up with the expulsion of oe particles and the production of new kinds of matter, a given quantity of radium must, in the course of time, disappear as such, and be transformed into inaetive substances. It ean be calculated that, in a gram of radium bromide, about half a milligram is transformed per year. In the course of about 1500 years half the radium present has been changed. If the whole world had been initially composed of pure radium, the amount of radium remaining 30,000 years later would not be more than one part in a million, i.e. about the amount observed to-day in a good specimen of pitchblende. Since there is every reason to believe that the earth's erust is very much older than this, we are forced to the eonclusion that radium must, in some way, be continuously produeed from the materials of the earth. In looking for a possible parent of radium, the elements uranium and thorium both suggest themselves, for both fulfil the neeessary eondition of having an atomic weight greater than that of radium, and both are always found in minerals from which radium is derived. In addition, both of these elements have a very long life compared with radium. Sinee the activity of uranium and thorium is less than one-millionth of that of radium, its life should be about one million times longer, i.e., a length of time of over 1,500 million years would be required before half the uranium present has been transformed. In some respects uranium seems the most likely parent of radium, for minerals rich in uranium are found to contain the most radium, while minerals rich in thorium often eontain little radium. It remains for future work to give adefinite answer to this question. If radium is produced from uranium, radium would oecupy the same position in regard to it as the radium emanation does to radium, the only differenee being that radium has a much longer life than its emanation. After the lapse of a few thousand years, the quantity of radium present in the mineral would reaeh a eonstant value, the production of radium balancing the loss by disintegration. The quantity of radium present in a mineral should, on this view, be always proportional to the amount of the parent element. Much attention has recently been directed to the distribution of radioactive matter in the earth's erust and atmosphere. Its presence and amount have been determined by observations on the rate of diseharge of an electroscope-a method transcending in delicacy even spectrum analysis. The experiments of Elster and Geitei, J. J. Thomson, Strutt, and others, have shown that radio-active matter, while found in the greatest quantity in the mineral pitchblende, is widely diffused in nature. Radio-active emanations occur everywhere in the atmosphere, in weIl water, in hot springs, and in surface water. Elster and Geitel have found that the earth's crust is all more or less radio-active. If air is sueked up through a pipe let down into the ordinary garden soil, it is found to be impregnated with

Tl!e Radiation alUl Emanatioll of Radium. Tl

657

emanations. The amount of radio-active matter varies in different localities, but is most marked in c1ays and in the mud obtained from hot springs. Since the presence of radio-activity is an indication that matter is breaking up with an enormous emission of energy, it is natural to ask what part radio-active substances play in cosmical physics. On the assumption that the earth was former1y a molten mass, Lord Kelvin has estimated that to cool down to its present state it would require aperiod of from twenty million to one hund red million years. This calculation places a definite maximum limit to the time that has e]apsed since the earth became cool enough to support life. In Thomson and Tait's 'Natural Philosophy', Appendix E, the foUowing sentence occurs, after a discussion of the probable age of the sun and earth: 'As for the future, we may say, with equal certainty, that the inhabitants of the earth cannot continue to enjoy the light and heat essential to their life for many million years longer, unless sources of heat-now unknown to us-are prepared in the great storehouse of creation.' In the light of the discovery of radium and other radio-active substances, which in their changes are able to emit an enormous amount of energy, this remark seems almost prophetie. Assuming the average temperature gradient (1' 50° Fahrenheit per foot descent from the earth's surface) and the conductivity of the rocks of the earth taken by Lord Kelvin, it can be calculated that the amount of heat conducted to the earth's surface each year and lost by radiation could be supplied by the presence of radium (or an equivalent amount of other radio-active matter) to the minute extent of about five parts in ten thousand million by weight. The observations of Elster and Geitel show that the amount of radio-active matter present in the soil is of this order of magnitude. Thus it does not appear improbable that the temperature gradient observed in the earth may be due to the heat liberated by the radio-active matter distributed throughout it. If this be the case, the present temperature gradient may have been sensibly constant for a long interval of time, and Lord Kelvin's computation may only supply the minimum limit to the age of this planet. Thus the earth may have been at a temperature capable of supporting anima I and vegetable life for a much longer time than estimated by Lord Kelvin from thermal data. Similar considerations apply to the question of the sun's heat; for the presence of radium in the sun, to the extent of about four parts in one million by weight, would of itself account for the present rate of emission of heat. The discovery of the radio-active elements, which in their disintegration liberate enormous amounts of energy, thus increases the possible limit of the duration of life on this planet, and allows the time claimed by the geologist and biologist for the process of evolution.

Slow Transformation Products of Radium by E. RUTHERFORD, F.R.S.

Macdonald Professor of Physics, McGill University, Montreal From the Philosophical Magazine for November 1904, sero 6, viii, pp. 636-650 Communicated to the International Electrical Congress, St. Louis, Sep~berI6,

1904

IT has been previously shown* that radium undergoes disintegration through aseries of weIl marked stages. The radium first of all produces the radium emanation, and this in turn is transformed into an active deposit, which behaves as asolid, and gives rise to the phenomena of excited activity. I have recently shown that this active deposit undergoes three further rapid transformations. t For convenience, the products in the active deposit will be termed Radium A, Radium Band Radium C, respectively.t The change from A to B is accompanied by a: rays alone, the change B into C is a rayless change, while the change C into D gives rise to a:, ß and 'Y rays. The time T for each of the products of radium to be half transformed is shown in the following table:

Radium Emanation Radium B Radium B Radium C

Rays a: rays

T

Radium

Active deposit

4 days

a:

rays

3 minutes

a:

rays

21 minutes

no rays

28 minutes

a:,

ß,

'Y rays

The changes in radium are not, however, completed at this stage, for it will be shown that there is very strong evidence that there are at least two more slow transformations. M. and Mme Curie§ observed that a body exposed in the presence of • Rutherford and Soddy, Phi!. Mag., April and May 1903.

t Bakerian Lecture, Roy. Soc., London, 1904.

t Note. The term Emanation X, which I previously employed to designate the matter Radium A, is not very suitable, and I have discarded it in favour of the present nomenclature, which is simple and elastic. § Theses presenlees cl la Faculte des Sciences, Paris, 1903, p. 116.

.'''/011' Transformation Producl.'I of Radium

659

radium emanation did not, after removal, completely lose all its activity. A residual activity always remained, which they state was of the order of 1/20,000 of the initial activity. It will be seen later, however, that the magnitude of this residual activity depends not only on the amount of emanation to which the body has been exposed, but also on the time of exposure. For an exposure of several hours, the residual activity is less than one millionth of the activity immediately after removal. Giesel* also observed that a platinum wire, after exposure to the emanation, showed residual activity which, he states, consists only of oe rays. An account will now be given of some investigations made by the writer on the nature of this residual activity and the chemical properties of the active matter itself. It is first of all necessary to show that the residual activity arises in consequence of a deposit of radioactive matter, and is not due to some action of the intense radiations to which the body made active has been subjected. The inside of a long glass tube was covered with equal areas of thin metal, inc1uding aluminium, iron, copper, silver, lead and platinum. A large amount of radium emanation was introduced into the tube, and the tube c1osed. After seven days the metal plates were removed, and, after allowing two days to elapse for the ordinary excited activity to disappear, the residual activity of the plates was tested by an electrometer. The activity of the plates was found to be unequal, being greatest for copper and silver, and least for aluminium. The activity of copper was twice as great as that of aluminium. After standing for another week, the activity of the plates was again tested. The activity of each had diminished in the interval to some extent, but the initial differences observed had to a large extent disappeared. After reaching a minimum value the activity of each plate slowly but steadily increased at the same rate. After a month's interval the activity of each of the plates was nearly the same, and over three times the minimum value. The initial irregularities in the decay curves of the different metals are, in all probability, due to slight but different degrees of absorption of the radium emanation by the metal plates, the absorption being greatest for copper and silver and least for aluminium. As the occ1uded emanation was slowly released or lost its activity, the activity of the met al fell to a limiting value. The absorption of the radium emanations by lead, paraffin and caoutchouc was some time ago observed by Curie and Danne. t The residual activity on the plates comprised both oe and ß rays, the latter being present in all cases, in a very unusual proportion. The equality of the activity and the identity of the radiation emitted from each plate shows that the residual activity is due to changes of some form of matter deposited on the plates, and that it cannot be ascribed to an action of the intense radiations ; for if such were the case, it would be expected that the activity produced on the different plates would vary not only in quantity, but also • Berichte d. D. Chem. Ges., p. 2368 (1903). t Comptes Rendus, cxxxvi, p. 364 (1903).

The Collected Papers 0/ Lord Ruthel/ord

660

in quality. This result is confirmed by the observation that the active matter can be removed from a platinum plate by solution in sulphuric acid, and has other distinctive chemical and physical properties. The variation of the residual activity with time will first be eonsidered. A platinum plate was exposed in the presence of the radium emanation for seven days. The amount of emanation initially present was equal to that obtained from about 3 milligrams of pure radium bromide. The plate immediately after removal gave a saturation eurrent, measured between parallel plates by a galvanometer, of 1· 5 X 10- 7 ampere. For some hours after removal the aetivity decayed according to an exponentiallaw with the time, falling to half value in 28 minutes. Three days after removal the active plate gave a saturation-eurrent, measured by an electrometer, of 5 X 10-13 ampere, i.e. 1/300,000 of the initial activity. The aetivity was observed to increase steadily with the time. The results are shown in Fig. 1, where the time is reckoned from the middle of the time of exposure to the emanation. 100

80 ~

~ .~ ~

60

~

'ts

~40

His, 111a r~y' ,u:;t/VI't-!I witIJ time

~

'i:::

"" ZO

10

20

30

Time /n

'~'y$

40

50

6D

Fig.l The curve is a straight line passing through the origin. The activity increases uniformly with the time for the interval of two months over which the observations have extended. Some results indicate that this steady increase with time continues for at least nine months. The emanation from 30 milli grams of radium bromide was condensed in a glass tube, which was then sealed. After a month's interval the tube was opened, and dilute sulphuric acid introduced. The acid dissolved

S/Oll'

Tran.~f()/,/1/afion

Products of Radium

661

thc acti"c rcsiduc dcpositcd in thc tube. On driving off the sulphuric acid hy heat. a radioactivc deposit was obtained. The first determination of the activity of this residue was made about six weeks after the introduction of the emanation. The activity eight months later was found to be about seven times the initial value. Thc results could not be very accurately obtained, as a portion of the activity had been removed in the interval by a bismuth rod placed in a solution of the active matter. The result. however, indicated that the activity had stcadily increased over aperiod of nine months. Radiations [rom the Active Matter

The residual activity consists of both oe and ß rays, the latter being present initially in an unusually large proportion. The proportion of oe to ß rays from the platinum plate, one month after rem oval, was at the most one-fiftieth of that from a thin film of radium bromide in radioactive equilibrium. UnHke the oe ray activity, the activity measured by the ß rays remains constant, and, in consequcncc, the proportion of oe to ß rays steadily increases with the time. The experiments showed that the intensity of the ß rays did not vary much, if at aU, over aperiod of nine months. The want of proportionality between the oe and ß rays shows that the two types of rays arise from different products. This conc1usion is confirmed by experiments, now to be described, which show that the products giving rise to oe and ß rays can be temporarily separated from one another by physical and chemical means. EjJect o[ Temperature on the Activity

An active platinum plate was exposed to varying temperatures in an electric furnace, and the activity tested after exposure at atmospheric temperature. Four minutes' exposure in the furnace, at first at 430°C, and afterwards at 800°C, had little, if any, effect on the activity. After 4 minutes at lOOO°C the activity decreased about 20 per cent, and a further exposure of eight minutes at a temperature of about 1050°C almost completely removed the oe ray activity. The activity recovered by the ß rays was, on the other hand, not appreciably changed by the high temperature. Further healing, however, at a still higher temperature, caused a decrease of the activity measured by the ß rays, showing that the ß ray product was also volatile. These results show that the active matter consists of two kinds. The part which emits ß rays is non-volatile at lOOO°C, but the other part which emits Ot: rays is almost completely volatilized at that temperature. Separation o[ the Constituents by means o[ a Bismuth Plate

The active matter of slow decay was obtained in solution by introducing dilute sulphuric acid into a glass tube in which the emanation from 30 milligrams of radium bromide had been stored for a month. The solution showed

662

The Collected Papers of Lord Rutherford

strong activity and gave out both IX and ß rays, the latter, as in other cases, being present in an unusually large proportion. When a polished bismuth disk was kept for some hours in the solution, it became strongly active. The active matter deposited on the bismuth gave out IX rays, but no trace of ß rays. After several bismuth disks had been successively left in the solution, the active matter, which emits IX rays, was almost completely removed. This was shown by evaporating down the solution after treatment. The ß ray activity remained unchanged, but the IX ray activity had been reduced to about 10 per cent of its original value. The active matter deposited in the bismuth does not appreciably change in activity in the course of one month, and some observations point to the conc1usion that there is not much change in five months. The observations in the latter case were, however, not precise enough to be sure that there was not a small percentage variation during that time. Experiments are now in progress to examine, with accuracy, the activity of the bismuth plate from time to time, and it is hoped that observations extending over the ensuing year will fix the rate of decay of this product, provided the rate of change is rapid enough to be measurable in a year's interval. The results obtained in this way are in agreement with those deduced by heating the active deposit to a high temperature. The active deposit contains two kinds of matter, viz.: (1) A product giving out only ß rays, which is soluble in sulphuric acid, but non-volatile at 1OOOoe, and which is not deposited on bismuth. (2) A product giving out only IX rays, which is soluble in sulphuric acid, volatile at 1OOOoe, and is deposited from a solution on bismuth. Explanation 01 the Results

We have seen that the IX ray activity increases if the ß ray product is present, but remains sensibly constant or changes very slowly if the IX ray product is removed from the ß ray product by the action of a bismuth plate. The ß ray activity remains sensibly constant independently of whether the IX ray product is present or not. These results show that the ß ray product is the parent of the IX ray product. The amount of residual activity from the radium emanation depends on the amount of emanation present and the time of exposure to the emanation. These results show that the active deposit of slow decay is a decomposition product of the emanation, and, since the first three transition products of the emanation, vize radium A, radium Band radium e, have been carefully analysed and shown to be consecutive, it is natural to suppose that the matter of slow rate of change is a product of the last rapid change in radium e. Following the nomenc1ature suggested, radium e gives rise to the ß ray product, which will be called Radium D, while radium D changes into the IX ray product, which will be called Radium E. The product radium D gives out only ß rays. The transition products of the disintegration of radium are shown diagrammatically in Fig. 2.

663

.\/ml' Transformation Products o.f Radium

o a pt.

No further changes have so far been observed. The active solution of radium D and E was tested to see if an emanation were present. A trace of the radium emanation was always observed, but this was probably due to a slight trace of radium carried over into the emanation vessel. This point is. however, under further investigation.

RADIUM

RAD. iM.

RAD. A

RAD.B

RAD.C

--'

RAD·E

RAD.D

ACTIVG DEPOSIT

ACTIVE DEPOSIT

tJF

OF

IlAPID CHANSE

SLOWCHAN"

Fig.2 Theory ofTwo Successive Changes

In all cases of radioactive change that have been examined, the amount of unchanged matter N, present at any time

t

is given by

~=

e-'J..t, where

No is the amount initially present, and A is the constant of change. Differen.. dN, \ . a1ways proportlOna . 1 to the tlatmg, dt = - I\N" or the rate 0 f change IS

amount present. Suppose that Po particles of the product radium D are deposited during the time of exposure to the emanation. This time is supposed to be so short that the amount of change of radium D during the time of exposure is very smalI. Let P = number of particles of matter radium D present at any time.

Q .0:: number of particles of radium E present at any time. AI = constant of change of radium D. A2 = constant of change of radium E. Then P = Poe-'J..1'. As the matter D changes into E, the value of Q at first increases. The increase dQ in the time dt is given by the difference between the number of particles of E (AlP) supplied by the change of D into E and the number of E (A 2 Q) which change into F.

664

The Cullectcd Papers uf Lurd Rutherfur"

Then

dQ = >"IPdt - >"2Qdt, dQ dt = >"tPOe-Alt - >"2Q.

and

The solution of the equation is of the form Q = ae-Alt + be-Alt. Since Q = 0 when t = 0, we havc a= -b

and

Q

=

=

>.., ·po >"2 - >"1 '

>",Po (e-Alt _ e- A•t ).

>"2 - >"t

For small values of t, Q = >"tPo.t, Le. the value of Q increases proportionately with the time. The value of Q passes through a maximum at a time T, given by e(A.-AI)T =

~~,

and then decreases with the time.

The initial increase of the er; ray activity with the time is thus in agreement with the view that radium E (which emits only a; rays) is produced from radium D. The time of observation (two months) has not yet been long enough to obtain more than the initial part of the er; ray curve. The results, however, show that an interval of two months is very short compared with the time required for the product D or E to be half transformed. Although the times of observation have been too short to determine experimentally either the value of >"1 or >"2, it is possible, on certain assumptions, to form a rough estimate of these values. It has been experimentally observed that each of the products of radium wh ich emit er; rays supplies about an equal proportion of the activity of radium when in radioactive equilibrium. Since, when equilibrium is reached, the same number of particles of each of the successive products must break up per second, this is an expression of the fact that every atom of each product breaks up with the expulsion of an equal number (probably one) of er; particles. Now radium Dis direct1y derived from radium C, and since the rate of change of D is very slow compared with that of C, the number of particles of D initially present must be very nearly equal to the number of particles of radium C which break up during the time that radium D is being formed. Now, suppose that each atom of radium C and D emits one ß particle with the same velocity . The ionization produced by each particle will be the same under the same experimental conditions, and the integrated value of the saturation current due to the ß rays over the time that the body is exposed to radium C must equal the corresponding integrated value for the ß rays during the life of radium D. Suppose, for example, that a quantity of emanation is introduced into a glass tube and left to stand for a month. During that interval tbe emanation has nearly all been transformed. The activity due to the ß rays from it will reacb a maximum several hours after

,)'/011' TJ'al/.~/;mllatiol1

Produc/s

(~f'

Radium

665

the introductioll nf Ihe emanation. and will then decay with the time falling to half value in fOllT days. Let i l be the maximum saturation current due to the ß rays, measured in a suitable testing vessel. The total quantity QI of electricity passing between the plates of the testing vessel during the life of the emanation is approximately given by

QI

f i dt

fes:,

oc,

=

0

=

0

ile-Afdt =

.

~•

where A is the constant of change of the emanation. In a similar way, if h is the initial current due to the ß rays from the radium D deposited in the tube (measured under identical experimental conditions) the corresponding value of Q2 =

~~ where

AI is the constant

of change of D. Since by hypothesis

QI

=

Q2. i2

AI T=~'

The ratio

~ 11

is determined experimentally, and since A for the emanation

is known, AI is determined. The details of the experiments by which the ratio i2/i l was determined need not be given here. It was deduced, on the above assumption, that half of the matter of radium D should be transformed in 40 years. In a similar way, the total number of oe partic1es expel1ed from radium C during the time radium D was being deposited must equal the number of oe partic1es expelled from radium E during its life, supposing that there is only one change which gives rise to oe rays. Assuming for the moment that the oe ray activity, observed for the active deposit ten months old, decayed from that time according to an exponential law, it was calculated that the period of the change could not be longer than 80 years. If the oe ray change has aperiod short compared with the ß ray change, the oe ray activity will finally decay at the same rate as the ß ray activity. These two computations will agree, if it is supposed that the oe ray activity increases to twice the value observed after an interval of ten months, and then decays with the time according to the period of the ß ray change. This would fix the period of the oe ray change at about one year. When the oe ray activity reaches its maximum value, it is to be expected, on this view. that the ratio ~ should be the same as for the product radium C. This is, however. somewhat at variance with experiment; for the ten months old deposit has about the same ratio ~ as radium C, while on the above computation the ratio ~ should be one half of that value. This difference

666

The Colleeted Papers of Lord RlIthel:ford

may possibly be due to radium E undergoing a further change, giving rise to oe rays. which has not so far been detected. In these calculations, it has been assumed that the oe and ß particles given out in these slow changes produce the same ionization as the corresponding particles from radium C. There is no doubt, however, that this is not realized in practice. The ß rays of radium D are slightly less penetrating than those of radium C, while the oe rays of radium E have only about half the penetrating power of those of radium C. Our knowledge of the mechanism of absorption in matter is, however, too imperfect to correct for these differences with any certainty. The above methods of calculation, though somewhat complicated, certainly serve to give the right order of magnitude of the periods of the two changes. It will be shown, too, that the calculated periods agree approximately with the amounts of radium D and E present in old sampIes of radium. The chief uncertainty in the methods of calculation lies in the difficulty of ascertaining the relative electrical effect produced by the oe and ß particles compared with those emitted from radium C. The time T required for each transition product of radium to be half transformed is shown in the following table: Transition products of radium

Time Tto be half transformed

Radium

about ]000 years

Emanation

4 days

RadiumA

3 minutes

Radium B

21 minutes

Radium C

28 minutes

Radium D

about 40 years

Radium E

about one year

Experiments with Old Radium Since the substance radium D is produced from radium at a constant rate, the amount present mixed with the radium will increase with its age. I had in my possession a small quantity of my first specimen of impure radium chloride, kindly presented to me by Professors Elster and Geitel four years ago. The amount of radium D present in it was tested in the following way: The substance was dissolved in water and kept continuously boiling for a period of about six hours. Under these conditions the emanation is removed as rapidly as it is formed, and the ß rays from the radium, due to the product

S/Oll' li"am/ormatioll Products

oI Radium

667

radium C, practically disappear. A newly prepared specimen of radium bromide under these conditions retains only a fraction of 1 per cent of its original ß radiation. The old radium, however, showed (immediately after this treatment) an activity measured by the ß rays of about 8 per cent of its original amount. The activity could not be reduced any lower by further boiling or aspiration of air through the solution. This residual ß ray activity was due to the product radium D stored up in the radium. It could not have been due to ß rays from radium C, since there was a distinct difference in penetrating power for the two kinds of ß rays. The ß ray activity due to radium D was thus about 9 per cent of that due to radium C. Disregarding the differences in the absorption of the ß rays, when the activity of the product D in radium reaches a maximum value, the ß ray activity due to it should be the same as that due to C. Since Dis halftransformed in forty years, the amount present in the radium after four years should be about 7 per cent of the maximum amount, i.e. it should show a ß ray activity of about 7 per cent of that due to radium C. The observed and calculated values (7 and 9 per cent respectively) are thus of the same order ofmagnitude. The amount of ß rays from radium D present in pure radium bromide about one year old was about 2 per cent of the total. The amount of radium E present in old radium was measured by observations of the activity imparted to a bismuth disk left for several days in the solution. Radium E is not deposited to an appreciable amount on the bismuth from a water solution of radium bromide. If, however, a trace of sulphuric acid is added to the solution, the radium E is readily deposited on the bismuth. The addition of sulphuric acid to the radium solution practically effected aseparation of radium D and E from the radium proper; for the latter was precipitated as sulphate and the products D and E remained in solution. After filtering, the solution contained a greater proportion of the products D and E and very little radium. The ratio a./ß for the old radium was found to be about twice that observed for radium C. This result is in agreement with the deductions made in the calculations of the periods of the changes; for it can be theoretically shown that the amounts of D and E in the radium continue to be approximately proportional to one another after five years' production, assuming the periods of the changes are forty years and one year, respectively. The amounts of radium D and E observed in the old radium are thus in good agreement with the results deduced from other data. Variation

0/ the Activity 0/ Radium with Time

It has long been known that the activity of freshly prepared radium increases at first with the time and reaches a maximum value after an interval of about one month. The results already considered show that there is a further slow increase of activity with the time. This is the case whether the activity is measured by the oe or ß rays. After a lapse of about 200 years the amount

668

The Collected Papers ofLord Ruthelford

of the products radium D and E will have practically reached a maximum value. The same number of atoms of each of the products C and D will then break up per second. If each atom of these products in disintegrating throws off an equal number (probably one) of ß partic1es, the number of ß partic1es thrown off per second will be twice as great as from radium a few months old. The number will increase at first at the rate of about 2 per cent a year. Similar considerations apply to the oe ray activity. Since, however, there are four other products of radium besides radium itself which expel oe partic1es, the number of oe particles emitted per second from old radium will not be more than 25 per cent greater than the number from radium a few months old. The activity measured by the oe rays will thus not increase more than 25 per cent and probably still less, as the oe partic1es from radium E probably produce less ionization than the oe particles expelled from the other radium products. It is probable that half of the radium itself is transformed in about 1000 years. The activity of radium will consequently rise to a maximum after 200 years and then slowly die away with the time.

Products in Pitchblende The products radium D and E must be present in pitchblende in amounts proportional to the quantity of radium present, and should be capable of separation from the mineral by suitable chemical methods. The radioactive properties of these substances, if obtained in the pure state, are summarized below. Radium D. The product immediately after separation should emit only ß (and probably y) rays. The ß ray activity should decay to half value in about forty years. In consequence ofthe change of D into E, the latter ofwhichgives out oe rays, the oe ray activity will increase for a few years, pass through a maximum, and then decrease with the time and fall to half value in about forty years. Since the rate of change of D is about 25 times as fast as radium itself, the activity on separation, measured by the number of electrons expelled per second, should be 25 times as great as from an equal weight of radium. The oe ray activity produced in it should at any time be capable of separation by adding a bismuth plate to a solution of the substance. Radium E. The substance should emit only oe rays, and its activity should fall to half value in about one year. Since its rate of change is about 1000 times as great as radium, the substance, weight for weight, should emit about 1000 times as many oe particles as freshly prepared radium, and about 250 times as many as from radium about one month old. The activity measured by the electric method will probably be about 100 times as great as that of pure radium. It is now necessary to consider the question whether the substances radium D and E have been previously separated from pitchblende, and are known by other names. In regard to radium D, there is some doubt

S'/Oll' Trall.~r(}rmatiol/ Prodllcls

oI Radium

669

whcthcr it has heen previously separated. It is possible that it is the radioactive constituent present in the radio-lead of Hofmann, for he states that this substance emits a large amount of ß rays. On the other hand, the radio-lead prepared by other observers lost its activity rapidly with the time. In regard to radium E, I think there is little doubt that it is the radioactive constituent present in the so-called radio-tellurium of Marckwald. * It will be recalled that Marckwald obtained a deposit of radioactive matter on a bismuth plate introduced into a solution of pitchblende. This active bismuth gave out only "1 e

=

-Alt

(7),

Q = .--~ (>.., e- A2t _ >"1 - >"2 >"2

R = no (ae- Alt

e-].It)

(8),

+ be- A2t + ce-Alt)

(9),

where

a=

>"2 .----.-.. b=- >.., >"'>"2 - c--(>"1 - >"2)(>"1 - >"3)' (>"1 - >"2)(>"2 - >"3)' - >"J (>"1 - >"3)(>"2 - >"3)'

100

Rela.tive a.mounts

R

I

of ma.tte~ A.B.C.

pre5en~ at anH in5ta.nt ~fter JI30

lang e:Jf05Ure.

80 c

0 ~

C

eS

~

er:

60

'+0

Q

~40

'e; s:

.c

~

.CI

..5: 2.0

A 0

.c P

A I5

B 30

60 45 Time in Minutes.

Fig.9

75

90

105

The Succession of ('hanges in Radioact ire Bodi('s

(jR7

Tlle relative numbers of atoms of P, Q, R existing at any time are shown graphically in fig. 9, curves A, B, and C respectively. The number of atoms R o is taken as 100 for comparison, and values of "" are taken corresponding to the 3, 21, and 28-minute changes in the active deposit of radium. A comparison with fig. 8 for a short exposure brings out very clearly the variation in the relative amounts of P, Q, R in the two cases. The amount of R initially decreases very slowly. This is due to the fact that the supply of C due to breaking up of B at first, nearly compensates for the breaking up of C. The values of Q and R after several hours decrease exponentially, reaching half value every 28 minutes. 12. Any Time oJ Exposure.-Suppose that a body is exposed in the presence of a constant supply of emanation, and that no particles of the matter Aare deposited each second. After a time of exposure T, the number of particles PT of the matter A present is given by

"2' "3

PT = no

I

T e-/'II

dt

n

=

~ (l -

"1

o

e-/'I T ).

At any time f, after rem oval of the body from the emanation, the number of particles P of the matter A is given by P = PTe- I ' 11 ==

X; (I

-

e-/'I T ) e- i ' lt .

Consider the number of partic1es nodt of the matter A deposited du ring the interval dt. At any time t later, the number of particles dQ of the matter B, which results from the change in A, is given (see equation 4) by dQ = -- no~_~_ (e- A2t Al - .1 2

e- All )

-

•

elt . --=- nof(r) elt

. (10).

After a time of exposure T, the number of particles QT of the matter B present is readily seen to be given by QT = no [J(T) df

IoJ(t) elf.

-+- J(T -

elf) dt

+ ....... +/(0) dt]

T

-=-=

no

If the body is removed from the emanation after an exposure T, at any time f later the number of particles of B is in the same way given by T+t

Q = noJJ(t)dt. I

It will be noted that the method of deduction of QT and Q is independent of the particular form of the functionf(t).

688

The Collected Papers of Lord RUfhel!ord

Substituting the particular value of I(t) given in equation (10) and integrating, it can readily be deduced that Q

ae- A2t -

QT

where

a

be-Att

(11),

a-b

b = 1 - e- AtT Al

= 1 - e- A2T

A2

In a similar way, the number of partic1es R of the matter C present at any time can be deduced by substitution of the value of I(t) in equation (5). These equations are, however, too complicated in form for simple application to experiment, and will be omitted. 13. Changes in the Active Deposit Irom Thorium.-If the variation of the activity, imparted to a body exposed for a short interval in the presence of the thorium emanation, is due to the fact that there are two successive changes in the deposited matter A, the first of which is a 'rayless' change, the activity I t at any time t after removal should be proportional to the number Ql of partic1es of the matter B present at that time. Now, from equation (4), it has been shown that

_ Aln , e ( -A2t Qt-, I -

_

1\2

-Alt) e .

The value of Qt passes through a maximum QT at the time T when

A2/AI =

e-(AI- A2)T.

The maximum activity I T is proportional to

QT

and

It Qt e- Azt - e- Att I T = QT = e- A2T _e- AIT

. (12).

It will be shown later that the variation with time of the activity, imparted to a body by a short exposure, is expressed by an equation of the above form. It thus remains to fix the values of Ah A2' Since the above equation is symmetrical with regard to Ah A2 , it is not possible to setde from the agreement of the theoretical and experimental curve which value of Arefers to the first change. The curve of variation of activity with time is unaltered if the values of Al and A2 are interchanged. It is found experimentally that the activity 5 or 6 hours after removal decays very approximately according to an exponentiallaw with the time, falling to half value in 11 hours. This is the normal rate of decay of thorium for all time of exposure, provided measurements are not begun until several hours after the removal of the active body from the emanation. This fixes the value of the constants of one of the changes. Let us assurne for the moment that this gives the value of Al'

Tllc Successioll 0/ Changes in Radioactil'e Bodies l'hcn ", =

1·75

Ä

689

10- 5 (sec)--'.

Since the maximum activity is reached after an interval T = 220 minutes (see fig. I, curve C), substituting the values of "-1 and T in equation (12), the value of comes out to be

"2

"2 =

2·08

X

10- 4 (sec)-l.

This value of "-2 corresponds to a change in which half the matter is transformed in 55 minutes. Substituting now the values of "-h "-2' T, the equation (12) reduces to

I,/IT = 1· 37 (e-)'2' - e-)'I'). The agreement between the results of the theoretical equation and the observed values is shown in the following table:Time in minutes

15 30 60 120 220 305

IellT

Observed value of I,fIT

0·22 0·38 0·64 0·90 1·00 0·97

0·23 0·37 0·63 0·91 1·00 0·96

Theoretical value of

It is thus seen that the curve of rise of activity for a short exposure is explained very satisfactorily on the supposition that two changes occur in the deposited matter, of which the first is a rayless change. Further data are required in order to fix which of the time constants of the changes refers to the first change. In order to setUe this point, it is necessary to isolate one of the products of the changes and to examine the variation of its activity with time. If, for example, a product can be separated whose activity decays to half value in 55 minutes, it would show that the second change is the more rapid of the two. Now PEGRAM* has examined the radioactive products obtained by electrolysis of thorium solutions. The rates of decay of the active products depended upon conditions, but he found that, in several cases, rapidly decaying products were obtained whose activity fell to half value in about I hour. Allowing for the probability that the product examined was not completely isolated by the electrolysis, but contained also a trace of the other product, this result would indicate that the last change which gives rise to rays is the more rapid of the two.

*

'Phys. Rev.', p. 424, December, 1903.

690

Tlte Collected Papers 0/ Lord Rutherford

The results obtained by VON LERCH* in the electrolysis of solution of the active deposit also admit of a similar interpretation. Products were obtained on the electrodes of different rates of decay, but which lost half their activity in times varying from about 1 hour to 5 hours. This variation is possibly due to the admixture of the two products, but further experiment is necessary to settle thls point with certainty. The evidence, as a whole, thus supports the conclusion that the active deposit from thorium undergoes two successive transformations as follows:-

=

1·75 X 10- 5 , i.e., in which half the matter is transformed in 11 hours; (2) A second change giving rise to IX, ß and 'Y rays, for which A2 = 2·08 x 10-4, i.e., in which half the matter is transformed in 55 minutes. t

(1) A 'rayless' change for which Al

It is, at first sight, a somewhat unexpected result that the final rate of decay of the active deposit from thorium does not in reality give the rate of change of the last product itself, but of the preceding product, which does not give rise to rays at all. A similar peculiarity is observed in the decay of the excited activity of actinium, which is discussed in section 15. 14. For a long exposure in the presence of a constant supply of thorium emanation, the equation expressing the variation of activity with time is found from equation (8), section 11.

!!... _ .Q _ 10

A2

e-')'ll _

At

e- A21

-

Qo - A2 - AI

=

A e- Atl 2 (1 _ 0.083 e-t.90xtO-4t). A2 - At

Al - A2

About 4 hours after removal, the second term in the brackets becomes very small, and the activity after that time will decay very nearly according to an exponentiallaw with the time, falling to half value in 11 hours. For any time of exposure T, the activity at time t after the removal (see equation 11) is given by It Q ae- Alt - be-Atl 10 = QT = a-b where I o is the initial value of the activity, immediately after rem oval, and

a=

1 - e- A2T

A2

b = 1 - e- A1T At

• 'Ann. d. Physik', November, 1903. t The 'rayless change' certainly does not give out oe rays, and special experiments showed that no appreciable amount of {J rays were present. On the other hand, the second change gives out all three types of rays.

n,e Successioll of Challges ill Radioactil'e Bodies

691

By variation of T the curves of variation of activity for any time of exposure can be accurately deduced from the equation, when the values of the two constants Alt A2 are substituted. Miss BROOKS has examined the decay curves of excited activity for thorium for different times of exposure and has observed a substantial agreement between experiment and theory. * 15. Changes in Actinium.-Dr. GIESEL kindly forwarded me a radioactive preparation from pitchblende and called by him the 'emanating substance', on account of the large amount of emanation it gives out. This had an activity, measured in the usual way, of about 250 times that of uranium. The emanation and excited activity produced by it were kindly examined for me in detail by Miss BROOKS. The emanation was found to have a very rapid rate of decay, its activity falling to half value in a few seconds. The excited activity for a long exposure fell to halfvalue in 41 minutes. DEBIERNEt has shown that actinium gives off an emanation which loses half its activity in 3· 7 seconds and produces excited activity which falls to half value in 41 minutes. There can be no doubt that the 'emanating substance' of GIESEL and the actinium of DEBIERNE contain the same radioactive constituent. The name actinium will thus be used in this paper to denote the 'emanating substance' of GIESEL. Miss BROOKS investigated the rate of decay of the excited activity of actinium for different times of exposure; but, for the pur pose of elucidation of the changes occurring, we need only consider the curves of decay of excited activity for a short and for a long exposure. For a long exposure the activity decays very nearly according to an exponential law, falling to half value in 41 minutes. The value of the change-constant ,.\ is 2·80 X 10- 4 (sec)-l. The activity for a short exposure at first increases, rapidly passes through a maximum, and after some time decays according to an exponentiallaw, falling to half value in 41 minutes. The curve of decay (measured by the 01: rays) for an exposure of 1· 5 minutes in the presence of the actinium emanation is shown in fig. 10. The maximum is reached about 7· 5 minutes, reckoning from the moment the body is exposed to the emanation. The curve is very similar in general shape to the corresponding curve of thorium, and can be analysed in a similar way; the activity at any time t is proportional to e- i.2t - e-i..(t. These results show that the first change occurring in the active deposit from actinium is a rayless change. Since the activity finally decays aeeording to an exponentiallaw (half value in 41 minutes), one of the eonstants of the change has a value 2·80 X 10- 4 • By substitution in the eurve, the value of the other eonstant is found to be 7·7 X 10- 3 (half transformed in 1· 5 minutes). As in the ease of thorium, a diffieulty arises as to which value of A applies to the rayless change, but the question in the ease of actinium can at onee be settled by means of electrolysis. [* OPhit. Mag.', September, 1904.] t 'Comptes Rendus', 138, p. 411, 1904.

692

The Collected Papers

0/ Lord Ruthelford

Miss BROOKS performed the following experiment:-A platin um plate was made active in the presence of actinium and the active matter was dissolved off by hydrochloric acid and then electrolysed. The activity of the anode, after removal, fell very rapidly according to an exponential law, reaching half value in 1· 5 minutes. The corresponding value of ,\ is 7·7 X 10- 3• There is thus no doubt that the second change is the most rapid 100

80 s: o

~

~ 60

Deca.y

Q::

of exci ed . activi ty of Acl inium.

Exp psure

b

l ~inute .

1

~40

'dj

s: Q)

,l.)

s:

20

o

5

10

15

20

Time in Minutes.

25

30

35

Fig. 10 of the two. We may thus conclude that the active deposit from actinium undergoes two distinct successive transformations: (1) A rayless change, in which halfthe matter is transformed in 41 minutes;

(2) A change giving rise to cx rays, in which half the matter is transformed in 1· 5 minutes. *

It can readily be shown that, for a very short exposure of a body in the presence of the actinium emanation, the activity I t at any time t is given by

ItlIT

=

1·14 (e- A1t

-

e- A2t),

where IT is the maximum value of I which occurs at time T

=

7· 5 minutes.

* The radiations from the products have not yet been examined to see whether ß and

I' rays are present.

The Succession of Changes in Radioactil'e Bodies

693

For a very long exposure 1//10 = 1·038 e-i.,t - 0·038 e- i .zt , where 10 is the initial value after removal, and A. = 2·80

X

10-4 ,

A2 = 7·7

X

10- 3 •

For the first 10 minutes after removal, the activity in consequence decays more slowly than is to be expected on a simple exponentiallaw. This result has been observed experimentally. The variation of the activity for any time of exposure to the emanation is expressed by an equation of the same form as for thorium, equation (12), with the values A.. A2 found above. There is some evidence that there is a product actinium X in actinium, corresponding to thorium X in thorium. This point is at present under investigation, and the results will be given in a later paper. If this is the case, actinium and thorium are very closely allied in the number and nature of their products. Both give rise to an emanation, and this is transformed into an active deposit which undergoes two further transformations, the first change being a 'rayless' one. 16. Changes in the Active Deposit from Radium.-In the ca se of the active deposit from radium, we are dealing with matter that undergoes at least four successive changes. F or convenience, the matter initially produced from the emanation will be called the matter A and the succeeding products B, C, 0, E respectively. The equations expressing the quantities of A, B, C, 0 present at any time are complicated, but the comparison of theory with experiment may be simplified by temporarily disregarding some unimportant terms. For example, the activity of the matter 0 is generally negligible compared with that of A or C, being as a rule less than 1/100,000 of the initial activity observed for the matter A or C. A still further simplification can be made by disregarding the first 3-minute change. In the course of 6 minutes after removal, three-quarters of the matter A has been transformed into B, and 20 minutes after removal all but about 1 per cent. has been transformed. The variation of the amount of matter B or C present at any time agrees more c10sely with the theory, if the first change is disregarded altogether. A discussion of this important point is given later in section 21. 17. ß-Ray Curves.-The explanation of the ß-ray curves (see figs. 5 and 6), obtained for different times of exposure, will be first considered. For a very short exposure, the activity measured by the ß rays is small at first, passes through a maximum about 36 minutes later, and then decays steadily with the time. The curve shown in fig. 6 is very similar in general shape to the corresponding thorium and actinium curves. It is thus necessary to suppose that the change ofthe matter B into C does not give rise to ßrays, while the change of C into D does. In such a case the activity (measured by the ß rays) is proportional to thc amount of C present. Disregarding the first rapid change,

694

The Collected Papers of Lord Rutherfol'd

the activity I t at any time t should be given by an equation of the same form (see equation 12, section 13) as for thorium and actinium, viz., It e-A3t - e- A2t I T = e A3T - e AzT'

where I T is the maximum activity observed, which is reached after an interval T. Since the activity finally decays according to an exponentiallaw (half value in 28 minutes), one of the values of Ais equal to 4·13 x 10- 4 • As in the case of thorium and actinium, the experimental curves do not allow us to settle whether this value of Ais to be given to A2 or A3. From other data (see section 20) it will be shown later that it must refer to A3. Thus A3 = 4·13 X 10-4 (sec)-l. The experimental curve agrees very closely with theory if A2 = 3 ·10 X 10- 4 (sec)-l. The agreement between theory and experiment is shown by the table given below. The maximum value I T (which is taken as 100) is reached at a time T = 36 minutes. Time in minutes

0 10 20 30 36 40 50 60 80 100 120

Theoretical value of

activity

0 58·1 88·6 97·3 100 99·8 93·4 83·4 63·7 44·8 30·8

Observe? yalue of

actlVlty

i

0 55 86 97 100 99·5 92 82 61·5 42·5 29 !

In order to obtain the ß-ray curve, the following procedure was adopted. A layer of thin aluminium was placed inside a glass tube, which was then exhausted. A large quantity of radium emanation was then suddenly introduced by opening a stopcock communicating with the emanation vessel, which was at atmospheric pressure. The emanation was left in the tube for 1· 5 minutes and then was rapidly swept out by a current of air. The aluminium was then removed and was placed under an electroscope, such as is shown in fig. 3. The cx rays from the aluminium were cut off by an interposed screen of aluminium 0·1 millim. thick. The time was reckoned from aperiod of 45 seconds after the introduction of the emanation.

Ihe Succession

0/ Changes il/

RadioaClil'l! Bodies

695

There is thus a fairly good agreement between the ca1culated and observed values of the activity measured by the ß rays. The results are thus satisfactorily explained if it is supposed:(1) That the change B into C (half transformed in 21 minutes) does not

give rise to

ß rays;

(2) That the change C into D (half transformed in 28 minutes) gives rise

to

ß rays.

These conclusions are very strongly supported by observations of the decay measured by the ß rays for a long exposurc. The curve of decay is shown in fig. 6 and fig. 11, curve I. °Sol

°Sol

5

.~

~~5

.~ 601 ~

't :1'4°1 ~ ·in s:

Q)

.u C

20

o

15

30

45 Time

in

60 Minuties

75

:JO

105

Fig. 11 P. CURIE and DANNE made the important observation that the curve of decay C, shown in fig. 7, for a long exposure, could be accurately expressed by an empirical equation of the form It/l o = ae- A3t

-

(a - 1) e- A1t,

where A2 = 3·10 X 10- 4 (sec)-l and A3 = 4·13 X 10-4 (sec)-I, and a = 4·20 is a numerical constant. I have found that within the limit of experimental error this equation represents the decay of excited activity of radium for a long exposure,

696

The Collected Papers ofLord Rutherford

measured by the ß rays. The equation expressing the decay of activity, measured by the oe rays, differs considerably from this, especially in the early part of the curve. Several hours after removal the activity decays according to an exponentiallaw with the time, decreasing to half value in 28 minutes. This fixes the value of AJ. The constant a and the value of A2 are deduced from the experimental curve by trial. Now we have already shown (section 14) that in the case of the active deposit from thorium, where there are two changes of constants A2 and AJ, in which only the second change gives rise to a radiation, the value of

1 10

_1 _

-

A2 e- A3t A2 - A3

_

A e-A1t -:--_J-.,.-

A2 - AJ

for a long time of exposure (see equation 8). This is an equation of the same form as that found experimentally by CURIE and DANNE. On substituting the values A2' AJ found by them,

A

. 2 - 2 1\3 \ =

4·3, and \

Al

1\1 -

\

1\3

=

3·3.

Thus not only does the theoretical equation agree in form, but also closely in the values of the numerical constants. If the first as well as the second change gave rise to aradiation, the equation would be of the same general form, but the value of the numerical constants would be different, the values depending upon the ratio of the ionization in the first and second changes. If, for example, it is supposed that both changes give out ß rays in equal amounts, it can readily be calculated that the equation of decay would be

I ~

10

=

0·5A 2 e-A3t A2 - AJ

_

0.5

(A A )1 - A3 2

3

_

e-A2t.

Taking the values of A2 and AJ found by CURIE, the numerical factor becomes 2· 15 instead of 4· 3 and 1· 15 instead of 3· 3. The theoretica1 curve of decay in this case would be readily distinguishable from the observed curve of decay. The fact that the equation of decay found by CURIE and DANNE involves the necessity of an initial rayless change can be simply shown as follows:-

e- A3t

Curve I (fig. 11) shows the experimental curve. At the moment of removal of the body from the emanation (disregarding the initial rapid change), the matter must consist of both Band C. Consider the matter which existed in the form C at the moment of removal. It will be transformed according to an exponentiallaw, the activity falling to half in 28 minutes. This is shown in curve H. Curve IH represents the difference between the ordinates of curves I and H. It will be seen that it is identical in shape with the curve (fig. 5) showing the variation of the activity for a short exposure, measured by the ß rays. It passes through a maximum at the same time (about

The Succession

0/ Changes in

Radioacth'e Bodies

697

36 minutes). The explanation of such a curve is only possible on the assumption that the first change is a rayless one. The ordinates of curve III express the activity added in consequence of the change of the matter B, present after removal, into the matter C. The matter B present gradua1ly changes into C, and this, in its change to D, gives rise to the radiation observed. Since the matter B alone is considered, the variation of activity with time due to its further changes, shown by curve III, should agree with the curve obtained for a short exposure (see fig. 5), and this, as we have seen, is the case. The agreement between theory and experiment is shown in the following tables. The first column gives the theoretical curve of decay for a long exposure deduced from the equation The agreement between theory and experiment is shown in the following tables. The first column gives the theoretical curve of decay for a long exposure deduced from the equation

1 . \2 .\3 -I e-ÄJI e-Ä21 10 - .\2 - .\J .\2 - .\3 ' taking the value of .\2 = 3· 10 X 10-4 and .\J = 4 ·13 X 10-4 • The second column gives the observed activity (measured by means of an electroscope) for a long exposure of24 hours in the presence ofthe emanation. Time in minutes

Calculated values

o

100 96·8 89·4 78·6 69·2 59·9

10

20 30 40 50 60 80

100

120

49·2 34'2

22·7

14·9

Observed values

100

97·0 88·5 77·5 67·5 57·0 48·2 33·5

22'5 14·5

In cases where a steady current of air is drawn over the active body, the observed values are slightly lower than the theoretical. This is probably due to a slight volatility of the product radium B at ordinary temperatures. 18. oe-Ray Curves.-The analysis ofthe decay curves ofthe excited activity of radium, measured by the oe rays, will now be discussed. The following table shows the variation of the intensity of the radiation after a long exposure in the presence of the radium emanation. A platinum plate was made active by exposure of several days in a gJass tube containing a large

The Collected Papers 0/ Lord

698

Ruthe~rord

quantity of emanation. The active platinum after removal was placed on the lower of two parallel insulated lead plates, and a saturating electromotive force of 600 volts was applied. The ionization current was sufficiently large to be measured by me ans of a sensitive high-resistance galvanometer, and readings were taken as quickly as possible after removal of the platinum 100

Intensity of Ra.dia.tion.

80

20

20

20

B C

o

10

20

30

Time

40

in Minutes.

50

60

20

Fig. 12 from the emanation vessel. The initial value of the current (taken as 1(0) was deduced by continuing the curves backwards to meet the vertical axis (see fig. 12), and was found to be 3 X 10- 8 ampere. Time in minutes

0 2 4 6 8 10 15 20

Current

100

80 69·5 62·4 57·6 52·0 48·4 45·4

Time in minutes

Current

30 40 50 60 80 100 120

40·4 35·6 30·4 25·4 17·4 11·6 7·6

11,(, ."uc('essiulI of Changes in Raclioactil'c Bodics

699

These rcsults are shown graphically in the upper curve of fig. 12. The initial rapid decrease is due to the decay of the activity of the matter A. If the slope of the curve is produced backwards from a time 20 minutes after removal, it cuts the vertical axis at about 50. The difference between the ordinates of the curves A + B + C and LL at any time is shown in the curve AA. The curve AA represents the activity at any time supplied by the change in radium A. The curve LL starting from the vertical axis is identical with thc curve already considered, representing the decay of activity measured by thc ß rays for a long exposure (see fig. 6). This is shown by the agreement of thc numbers in the following tables. The first column in the table below gives the theoretical values of the activity deduced from the cquation

~

Jo

=

~__ e- I' 3t

A2 - A3

_

A3 e- i.2t A2 - A3

for the values of A2. A3 previously employed. The second column gives the observed values of the activity deduced from the decay curve LL. Time in minutes

0

10

20 30

40

50 60 80 100 120

Calculated value of activity

Observed value of activity

100 96·8 89·4 78·6 69·2 59·9 49·2 34·2 22·7 14·9

100 97·0 89·2 80·8 71·2 60·8 50·1 34·8 23·2 15·2

The c10se agreement of the curve LL with the theoretical curve deduced on the assumption that there are two changes, the first of which does not emit rays, shows that the change of radium B into C does not emit IX rays. In a similar way, as in the curve I, fig. 11, the curve LL may be analysed into its two components represented by the two curves ce and BB. The curve ce represents the activity supplied by the matter C present at the moment of removal. The curve BB represents the activity resulting from the change B into C and is identical with the corresponding curve in fig. 11. Using the same line of reasoning as before, we may thus conclude that the change of B into C is not accompanied by IX rays. It has already been shown that it does not give rise to ß rays, and the identity of the ß and y-ray curves show

700

The Collected Papers of Lord Rutherford

that it does not give rise to y rays. The change B into C is thus a 'rayless' change, while the change C into D gives rise to all three kinds of rays. An analysis of the decay of the excited activity of radium thus shows that three distinct rapid changes occur in the deposited matter, viz.:(1) The matter A, derived from the change in the emanation, is half transformed in 3 minutes and is accompanied by Ot rays alone; (2) The matter B is half transformed in 21 minutes and gives rise to no ionizing rays; (3) The matter C is half transformed in 28 minutes and is accompanied by Ot, ß and y rays; (4) A fourth very slow change, which will be discussed later (section 23). 19. Equations Representing the Activity Curves.-The equations repre-

senting the variation of activity with time are for convenience collected below, where Al = 3·8 X 10- 3, A2 = 3·10 X 10- 4, A3 = 4·13 X 10-4(1) Short exposure: activity measured by

It/IT

=

10·3 (e- A3t -

ß rays, e-~'2'),

where I T is the maximum value of the activity; (2) Long exposure: activity measured by ß rays, 1,/10 = 4·3 e- A3' - 3· 3 e-~'2t, where 10 is the initial value; (3) Any time of exposure T: activity measured by the I, ae- A3t - be- A2t 10 = a- b where a = 1 - e- A3T b = 1 - e- AzT

A3

(4) Activity measured by

1

Ot

I~ = te-AI'

ß rays,

A2

rays: long time of exposure,

+ 1- (4· 3 e- A3t -

3·3 e- AZ').

The equations for the Ot rays for any time of exposure can be readily deduced, but the expressions are somewhat complicated. 20. Equations of Rise of Excited Activity.-The curves expressing the gradual increase to a maximum of the excited activity produced on a body exposed in the presence of a constant amount of emanation are complementary to the curves of decay for a long exposure. The sum of the ordinates of the rise and decay curves is at any time a constant. This necessarily follows from the theory and can also be simply deduced from apriori considerations. ('Radioactivity', p. 267.)

The Successiol/

0/ Changes in

Radioactil'e Bodies

701

The c.urvcs of rise and decay of the cxcited activity are shown graphically in fig. ] 3 for both the (X and ß rays. The thick line curves are for the (X rays. The differenee between the shapes of the deeay eurves when measured by the (X or ß rays is clearly brought out in the figure. The equations representing the rise of activity to a maximum are given below. For the ß and y rays,

T,/I max = 1 - (4,3 e- Alt For the

(X

-

3·3 e-~·2').

rays,

T,/I max = I - 1- e- A1t

-

1- (4,3 e- Alt

-

3· 3 e- A2t ).

20

20

5

i

~ a::

601

'5 :s\

.u e;n

Anl --r-

t:

GI

u

c 20

o

20

20

30 40 Time in Minutes.

so

60

70

Fig. 13

21. Effect 0/ Temperature.-We have so far not considered the evidence on which the 28-minute rather than 21-minute change is supposed to take place in the matter C. This evidenee has been supplied by some reeent important experiments of P. CURIE and DANNE* on the volatilization of the aetive matter deposited by the emanation. Miss GATESt showed that this aetive matter was volatilized from a platinum wire above a red heat and deposited on the surfaee of a cold cylinder surrounding the wire. CURIE and DANNE extended these results by subjecting an active platinum wire tor a * 'Comptes Rendus', 138, p. 748, 1904. t 'Phys. Rev.', p. 300, 1903.

702

The Collected Papers o[ Lord Rutllelford

short time to the action of temperatures varying between 150 C. and 1350"C., and then examining at room temperatures the decay curves not only for the active matter remaining on the wire, but also for the volatilized part. They found that the activity of the distilied part always increased after removal, passed through a maximum, and finally decayed according to an exponential law (half value in 28 minutes). At a temperature of about 630 0 C. the active matter left behind on the wire decayed at once according to an exponential law, falling to half value in 28 minutes. P. CURIE and DANNE showed that the matter B is much more volatile than C. The former is completely volatilized at about 600 0 c., while the latter is not completely volatilized even at a temperature of 1300 C. The fact that the matter C, left behind when B is completely volatilized, decays at once to half value in 28 minutes shows that the matter C itself and not B is half transformed in 28 minutes. CURIE and DANNE also found that the rate of decay of the active matter varied with the temperature to which the platinum wire had been subjected. At 630 0 C. the rate of decay was normal, at 11000 C. the activity fell to half value in about 20 minutes, while at 13000 C. it fell to about half value in about 25 minutes. I have repeated the experiments of CURIE and DANNE and obtained very similar results. It was thought possible that the measured rate of decay observed after heating might be due to a permanent increase in the rate of volatilization of C at ordinary temperatures. This explanation, however, is not tenable, for it was found that the activity decreased at the same rate whether the activity of the wire was tested in a closed tube or in the open with a current of air passed over it. These results are of great importance, for they indicate that the rate of change of the product C is not a constant, but is affected by differences of temperature. This is the first case where temperature has been shown to exert an appreciable influence on the rate of change of any radioactive product. 22. Effect of the First Rapid Change.-We have seen that the law of decay of activity, measured by the ß or y rays, can be very satisfactorily explained if the first 3-minute change is disregarded. The full theoretical examination of the question given in sections 10 and 11 and the curves of figs. 8 and 9 shows, however, that the presence of the first change should exercise an effect of sufficient magnitude to be detected in measurements of the activity due to the succeeding changes. The question is of great interest, for it involves the important theoretical point whether the substances A and B are produced independently of one another, or whether A is the parent of B. In the latter case, the matter A which is present changes into B, and, in consequence, the amount of B present after A is transformed should be somewhat greater than if B were produced independently. Since the change of A is fairly rapid, the effect should be most marked in the early part of the curve. In order to examine this point experimentally, the curve of rise of activity, 0

"flic SUC'l'l'ssicJIl of Change,\' in RadioClctil'c Boc!ies

703

measured by the ß rays, was determined immediately after the introduction of a large quantity of the radium emanation into a closed vessel. The curve of decay of activity on a body after removal of the emanation, and the rise of activity after the introduction of the emanation, are in all cases complernentary to one another. While, however, it is difficult to measure with certainty whether the activity has fallen in a given time, for example, from 100 to 99 or 98· 5, it is easy to be sure whether the corresponding rise of activity in the converse experiment is 1 or 1·5 per cent. of the final amount. Fig. 14, curve I, shows the rise of activity (measured by the ßrays) obtained 201

201

.

:::Jo

Cl

'>j3

~I51

" j3

:n ~

~IO

j3

~

o-e SI

o

5

IO

15

20.

Time in Minutes.

25

3°

Fig. 14 for an interval of 20 minutes after the introduction of the emanation. The ordinates represent the percentage amount of the final activity regained at any time. Curve III shows the theoretical curve obtained on the assumption that A is a parent of B. This curve is calculated from equation (9) discussed in section 11, and '\1> '\2' '\3 are the values previously found. Curve II gives the theoretical activity at any time on the assumption that the substances A and B arise independently. This is calculated from an equation of the same form as (8). lt is seen that the experimental results agree best with the view that A and B arise independently. Such a conclusion, however, is of too great

704

The Collected Papers ofLord Ruthelj'ord

importance to be accepted before examining closely whether the theoretical conditions are fulfilled in the experiments. In the first place, it is assumed that the carriers which give rise to excited activity are deposited on the surface of the body, to be made active immediately after their formation. There is some evidence, however, that some of these carriers exist for a considerable interval in the gas before their deposit on the body. For example, it is found that if a body is introduced for a short interval, about 1 minute, into avessei containing the radium emanation, which has remained undisturbed for several hours, the activity after the first rapid decay (see fig. 4, curve B) is in much greater proportion than if an electric field had been acting for some time previously. This result indicates that the carriers of Band C both collect in the gas and are swept to the electrode when an electric field is applied. This effect may in part be due to a slight volatility of the matter B at ordinary temperatures. * If the matter B exists to some extent in the gas, the difference between the theoretical curves for three successive changes would be explained; for, in transferring the emanation to another vessel, the matter B mixed with it would commence at once to change into C and give rise to a part of the radiation observed. The equal division of the activity between the products A and C (see fig. 12) supports the view that C is a product of A, for when radioactive equilibrium is reached, the number of partic1es of Achanging per second is equal to the number of B or C changing per second. If each atom of A and C expels an oe partic1e of the same mass and with the same average velocity, the activity due to the matter A should be equal to that due to the matter C; and this, as we have seen, is the case. Further investigations are in progress, which it is hoped will throw more light on this difficult question. 23. Very Slow Change in the Active Deposit from Radium.-M. and Mme. CURIEt have observed that bodies which have been exposed for a long interval in the presence of the radium emanation do not lose all of their activity. The excited activity at first decays according to the equations already considered, but a residual activity always remains, of the order of 1/20,000 of the initial activity. This residual activity seemed fairly permanent, for it did not decay during an interval of six months. GIESEL observed that a platinum wire which has been exposed to the radium emanation shows residual activity, and he states that the radiation consists entirely of oe rays. I have examined this residual activity in the following way. The emanation from 30 milligs. of pure radium bromide was condensed in a small glass tube and the ends of the tube sealed. After standing for a month the tube was

* This result is supported by some recent experiments of Miss BRoOIes ('Nature', July 21, 1904). It was shown that the matter B is volatile at ordinary temperatures, and a small part escapes from the active body and is deposited in the neighbourhood. It was also observed that the vo]atility ofthe matter B was far more marked during the first 10 minutes after removal. i.e., during the time the first change is in progress. If Ais the parent of B, the expulsion of a charged 0( particle must set B in motion, and in consequence some of the atoms of B may acquire sufficient velocity to escape from the active body. t 'These presentee a la Faculte des Sciences', Paris, 1903, p. 116.

71/(:

Suc('e~sioll

(d' Changes in Radioactil'c Bodies

705

opened, and left to stand for several days, in order to allow all the remaining emanation to escape. The inside of the glass tube was found to show considerable activity. The tube was then filled with dilute sulphuric acid, and, after standing several hours, the acid was removed and evaporated down to dryness in a flat gl ass dish. Most of the active matter in the glass tube was dissolved out by the acid, and, after evaporation, a strongly active residue was obtained in the glass dish. The active matter was found to give out both ex and ß rays, and the ß rays were present in a very unusual proportion; thus, compared with the intensity of the ex rays, the ß rays were present in at least 10 times greater proportion than for a thin layer of uranium, thorium, or radium, in astate of radioactive equilibrium. The intensity of the ex radiation was first tested when the active deposit was about 2 months old. The activity, at that time, did not vary apparently over a week's interval. Owing to a numerical error it was at first thought that the activity did not change during a further interval of 3 months; but the corrected result showed that the activity had more than doubled itself du ring that interval. Still later observations show that the ex-ray activity is steadily . . mcreasmg. The increase of the ex-ray activity with time has been confirmed by observations of the residual activity left behind on a platinum plate exposed to the emanation. The results then showed that the ex-ray activity during the first month, after removal, increases considerably. The relative proportion of ß to ex rays steadily diminishes with the time. This is not due to a diminution ofthe ß-ray activity, but to an increase ofthe ex-ray activity. Further observations are in progress to ex amine the variations of the activity over long periods of time. The results, so far as they have gone, show that the residual activity produced on bodies by exposure to the radium emanation is very complicated. The results discussed in the next section show that the large ß-ray activity is due to matter of different chemical properties from that which gives rise to the ex rays. The increase of the ex-ray activity with time indicates that the deposited matter undergoes a slow 'rayless' change. The evidence at present obtained points to the conc1usion that the deposited matter is initially complex. A small portion of the total amount undergoes a change, accompanied by the emission of ß rays alone. The main portion of the deposited matter undergoes a very slow 'rayless' transformation, and the resulting product or products give rise to ex rays. Observations extending over a long period of time will be required to determine the period of these changes. It seems probable that radium C breaks up into two distinct products. The major part of the product then undergoes a further change or succession of changes. *

* [October 10, 1904.-Further experimental work on this subject has led to a modification of the above conclusions. It has been found that the results are best explained on the supposition that radium C gives rise to radium D, which in breaking up emits only P(and Z

706

The Collected Papers 0/ Lord Ruthelford

24. Connection 0/ Slowly Decaying Product with Radio-Tellurium.-Some evidence will now be considered which points to the strong probability that one of these slowly changing products of radium is the same as the active constituent present in radio-tellurium, separated from pitchblende by MARCKWALD. It will be recalled that the polonium of Mme. CURIE is always obtained with bismuth, but can be partially separated from it by several distinct methods. GIESEL early observed that a bismuth plate dipped into a radium solution became permanently active and gave out only oe rays, and in this respect resembled the polonium of Mme. CURIE. GIESEL has throughout insisted that polonium was nothing more than 'induced bismuth'-apparently considering that the bismuth acquired the radioactive property by mixture with a radium solution. Taking the point of view that radioactivity is always the result of changes occurring in special kinds of matter, the experiments of GIESEL indicate that a radioactive product is deposited from the radium solution on to the bismuth plate, and that the activity of the plate is not due to the bismuth at all, but to radioactive matter deposited on its surface. MARCKWALD found that a bismuth plate dipped into a solution of pitchblende was coated with a radioactive deposit. He at first thought this activity was due to the tellurium which was deposited on the plate, and consequentlY gave it the name radio-tellurium. Later results have, however, shown that the activity has nothing whatever to do with the tellurium, and by suitable chemical methods he has separated an extreme1y active substance. MARCKWALD states that the radio-tellurium gives out only oe rays, and has not appreciably changed during six months after separation. Mme. CURIE found that the active constituent present in bismuth, which had been made active by placing it in a radium solution, could be fractionated in the same way as polonium, and in that way was able to obtain bismuth 2000 times as active as uranium, but this activity decreased with the time. An account will now be given of some experiments to ascertain if there is any connection between this slowly decaying product of radium and radiotellurium or polonium. The active matter deposited on the glass tube, in which a large amount of radium emanation had been stored for a month, was dissolved in sulphuric acid. A bismuth plate was introduced into the solution and left for several hours. The bismuth plate was found to be strongly active and to give out only oe rays. On adding a second bismuth plate, very little active matter was deposited upon it. The remaining solution was then evaporated, and the probably y) rays. Radium D produces radium E, which breaks up with the emission of only oe rays. It has been deduced that D is probably half transformed in 40 years and E in about 1 year. This modification has been introduced into the subsequent schedules in the text. A full account of these experiments was communicated to the Electrical Congress at St. Louis (September 16, 1904) in a paper entitled 'Further Transformation Products of Radium'.]

The SlI('('('.vs;on of ('hanges ;n Rad;oartil'e Bodies

707

residue was found to be about as active as the bismuth plate, and to give out both IX and ß rays. There is thus no doubt that the matter dissolved off the glass is complex, and one part is deposited on bismuth and the other not. The activity of the bismuth plate did not appreciably change during a month's interval. * Part of the active matter obtained from the sulphuric acid solution then behaves in a similar way to the radio-tellurium of MARC'KWALD, inasmuch as it gives out only IX rays, and is readily deposited on bismuth. In order to test the apparent similarity still further, an accurate comparison was made between the penetrating power of the rays from the active bismuth and a bismuth plate of radio-tellurium obtained from Dr. STHAMER, of Hamburg. The rays were found initially to be about half absorbed in a thickness of O· 00034 centim. of aluminium, and exhibited the characteristic property of the rays of rapidly increasing absorption with the thickness of matter traversed. No appreciable difference in the penetrating power of the IX rays from the two substances was observed, although the intensity of the radiations was reduced to less than 1 per cent. of the initial value. It has been found experimentally that the rays from the different radioactive products differ in penetrating power, and the curve of absorption for different thicknesses of absorbing material is, in general, a characteristic of each product. The identity of the curves of absorption of the IX rays from the active bismuth and the radio-tellurium, coupled with the similarity of the radiations and chemical properties, is very strong evidence that the active product is in each ca se the same. I think there can be Httle doubt that the active constituent of radio-tellurium of MARCKWALD is a disintegration product of radium. The polonium (radioactive bismuth) obtained by me from Paris loses its activity fairly rapidly, and the IX rays from it are more readily absorbed than the IX rays of radio-tellurium. This greater absorption may, however, be due in part to the fact that the radiations from the polonium co me from the mass of the bismuth, and in consequence are made up of rays of widely different penetrating power, while the rays from the radio-tellurium arise from a thin film of matter deposited on the bismuth plate, in which the absorption of the issuing rays would be small. The identity or otherwise of the constituent present in the polonium of Mme. CURIE and the radiotellurium of MARCKW ALD has been a much vexed question. Adefinite answer cannot be given until accurate observations have been made of the change of activity with time of these products. 25. Radioactivity observed in Ordinary Matter.-A large number of experimenters have observed that ordinary matter possesses the property of radioactivity to a feeble degree. R. J. STRUTT found that different sampies of the same metal showed wide differences in radioactive power. It is a matter • This bismuth plate was unfortunately mislaid during my visit to England in May. I am, in consequence, unable to give a more definite statement in regard to the change of the activity with time.

708

The Collecfed Papers of Lord Rutheljord

of great importance to settle whether the weak radioactivity observed is a property of the substance under examination or is due to aminute radioactive impurity. I think there is !ittle doubt that the radioactivity observed in some substances is due in part to a deposit of radioactive matter on its surface from the atmosphere. It has now been conclusively shown that the radium emanation is present in the atmosphere. This, in the course of its disintegration, gives rise to the slowly decaying product, which will be deposited from the air on the surface of all bodies exposed in the open. Such bodies will thus

Radium

ThoriUM

Emanation

Radium A

Rad.B

Rad.C

Thorium X Emanation Thor;um A Thor.B

Uranium

Ur'4nium){ Final Produet

Actinium

Actinium x Emanal:ion Actin.A

?

Actin. 8

Rad.D

Rad.E

Thor.C

Actin.C

Fig. 15 be covered with an invisible film of radioactive matter of very slow rate of change. The results of COOKE* are a strong confirmation of this point of view. By carefully polishing the surface of a brass electroscope, he was able to reduce its rate of discharge to almost one-third of the normal amount. In such a case the radioactive matter was removed from the surface by the process of polishing. The strong radioactivity observed in a room in which radium preparations have once been used, is probably due to the deposit on the walls of the room of this slowly decaying matter from the emanation. 26. Comparison 0/ the Changes in the Radio-Elements.-The changes occurring in the three radio-elements and the radiations which accompany them are shown graphically in fig. 15. The radiations from actinium have not yet been sufficiently examined to be certain where the ß or y rays appear in the last change as in the case of thorium. It will be seen that there are, at

*

'Phil. Mag.', August, 1903.

fllc Suc('cssion of Changes in Radio(letil'e Bodics

709

least, six successive changes in radium, five in thorium, and two in uranium. The first five changes in radium are analogous, in many respects, to the corresponding changes in thorium and actinium. Each of these elements gives rise to a gaseous product, the emanation, and this in turn is transformed into a type of matter which is deposited on the surface of bodies. In both thorium and radium, the fourth change is followed by a change in which all three types of rays appear together, while the third product in all three elements does not emit rays at all. The remarkable similarity in the nature of the changes occurring in radium, thorium, and actinium, indicates that the constitution of the atoms of these bodies is very similar. The time T taken for each product to be half transformed, and the value of the related constant A, the fractional amount of the product changed per sccond, are shown in the following table, together with some of the physical and chemical properties of the products. The value of T may be taken as a comparative measure of the stability of the atoms of each product. The atoms of the radium emanation and thorium X have about the same stability. The apparent agreement of the rate of change in the two cases must be considered as a coincidence, and does not, in any way, indicate that the atoms of the two products are the same, for ThX and the radium emanation differ both in physical and chemical properties. If the atoms of the two products were identical, it would be expected that the subsequent changes would take place in the same way and at the same rate; but such is not the case. The stability of the atoms of the products varies over a very wide range. The stability of the atom of ThX, for example, is about 100,000 times greater than that of the actinium emanation. There is every reason to believe that the radio-elements themselves must be regarded as radioactive products of very slow rate of change. In a case like uranium, which is probably half transformed in about 1,000,000,000 years, the atoms must be considered as very stable compared with products like the emanations of thorium or radium. 27. Rayless Changes.-The existence of a well-marked change in radium, thorium, and actinium, which is not accompanied by the expulsion of oe or ß particles, is of great interest and importance. Since the rayless changes are not accompanied by any appreciable ionization of the gas, their presence cannot be detected by direct means. The rate of change of the substance can, however, be indirectly determined, as we have seen, by determination of the variation with time of the activity of the succeeding product. The law of change has been found to be the same as for the changes which give rise to oe rays. The rayless changes are thus analogous, in some respects, to the monomolecular changes observed in chemistry, with the difference that the changes are in the atoms themselves, and are not due to a molecular combination of the atoms with another substance. It must be supposed that a rayless change is not of so violent a character as one which gives rise to the expulsion of oe or ß particles. The change may

710

The Collected Papers

0/ Lord Ruther/ord Some physical and chemical properties

Product

T

URANIUM

1()11 years

2·2xl0- 17

Soluble in excess of ammonium carbonate.

UrantumX

22 days

3·6xl0- 7

Insoluble in excess of ammonium carbonate.

THORIUM

3 X 109

7 X 10- 18

Thorium X

4 days

2·00 X 10- 6

Soluble in ammonia.

Thorium emanation

1 minute

1·15 x 10- 2

Chemically inert gas; condenses about -1200 C.

Thorium A

11 hours

1·75 x 10- 5

Thorium B

55 minutes

2·1 x 10- 4

Behaves as solid; insoluble in ammonia; volatilized at a white heat; soluble in strong acids; Thorium A can be separated from B by electrolysis.

Actinium emanation

3·7 seconds

1 ·87 X 10-1

Behaves as agas.

ActiniumA

41minutes

2·80xl0- 4

Actinium B

1· 5 minutes

7·7 x 10- 3

Behaves as solid; soluble in strong acids; A can be partially separated from B by electrolysis.

RADIUM

800 years

2·8x 10- 11

Radium emanation

4 days

2·00xl0- 6

Chemically inert gas; condenses about -1500 C. Definite spectrum; volume diminishes with time.

Radium A

3 minutes

3·8x1O- 3

RadiUm B

21 minutes

5·38 X 10- 4

Radium C

28 minutes

4·13 X 10- 4

Behaves as solid; soluble in strong acids; volatilized at a white heat; Bismore volatile than A or C and can thus be temporarily separated from them.

RadiumD

About 40 years

Gives out onlr P rays. Soluble in strong aClds.

RadiUmE

About 1 year

Probably active constituent of radio-tellurium; soluble in strong acids; volatilized at a red heat; deposited on bismuth in solution.

" (sec)-1

Final product

Insoluble in ammonia.

ThoriilmC (final product) ACTINIUM Actinium X?

Actinium C (final product)

The Sl/('('eSSiOI1 of Changcs in Radioarti\'c Bodics

711

be accounted for either by supposing that there is are-arrangement of the components of the atom, or that the atom breaks up without the expulsion of its parts with sufficient velocity to produce ionization by collision with the gas. The latter point of view, if correct, at once indicates the possibility that changes of a similar character may be taking place slowly in the nonradioactive elements; or, in other words, that all matter may be undergoing a slow process of change. The changes taking place in the radio-elements have been detected only in consequence of the expulsion with great velocity of the parts of the disintegrated atom. lf the oe particles had been expelled with a velocity less than 108 centims. per second, it is improbable that any ionization would have been produced, and the changes, in consequence, could not have been followed by the electric method. 28. Radiations from the Products.-The radiations from the successive products of the dis integration of radium have been very closely investigated, and it has been found that, with the exception of the rayless change, all the changes are accompanied by the emission of oe partic1es with great velocity. The ß and y rays appear only in the fifth change. In the case of thorium, it has not been found possible to completely free the product ThX from ßrays, on account of the difficulty of entirely removing from it the products of the subsequent change. The proportion of ß rays is, however, greatly reduced if the emanation produced by the ThX is removed by passing a rapid current of air through a solution of ThX. The emanation itself gives out only oe rays, but the second product, thorium B, arising from it gives out all three types of rays. A rem oval of part of the emanation thus decreases the amount of ßrays from the ThX. I think there is little doubt that, if the emanation could be removed from the ThX as fast as it was formed, it would be found that the ThX itself gives out only oe rays, and that the ß and y rays, as in the case of radium, appear only in the fifth change. It is remarkable that the ß and y rays of uranium, thorium, and radium appear only in the last of the rapid succession of changes occurring in those bodies. It has al ready been pointed out that the ß and y rays always appear together and in the same proportion. There is now Httle doubt that the y rays are electromagnetic pulses, similar to X rays, generated at the moment of the sudden expulsion of the ß partic1e from the radio-atom. In the three radio-elements, the expulsion of the ß partic1e results in the appearance of a product either permanently stable, or, in the case of radium, of a product far more stable than the preceding one. It would appear that the initial changes are accompanied only by the expulsion of an oe partic1e, and that once the ß partic1e is expelled, the components of the residual atom fall into an arrangement of fairly stable equilibrium, when the rate of disintegration is very slow. I think that it is more than a coincidence that the ß and y rays appear only in the last of the rapid changes in the three radio-elements. It appears probable that the ß partic1e, which is finally expelled, may be regarded as the active agent in promoting the disintegration of the radio-atom

712

The Collected Papers of Lord Rlithel'ford

in successive stages. According to the modern point of view of regarding atomic structure, the atoms of the radio-element may be supposed to be made up of electrons (ß particles) and groups of electrons (0: particles) in rapid motion, and held in equilibrium by their mutual forces. If the atom is to remain permanently stable, it is necessary that there should be no loss of energy as a whole from the moving charged parts of which the atom is buHt up. LARMOR has shown that this condition is fulfilled if the vector sum of the accelerations of the moving particles is permanently null. If this is not the case, there must be a continuous drain of energy from the atom in the form of electromagnetic radiation. This, in the course of time, must disturb the equilibrium of the atom, and result either in are-arrangement of its component parts or to its final disintegration. It may, perhaps, be supposed that occasionally one of the outlying revolving electrons, comprising the radio-atom, 1apses into a position which results in a slow 10ss of energy from the atom in the form of radiation. In consequence of this loss of energy, the atom becomes unstable, and ultimately an cx particle flies off with its great orbital velocity, but the atom still retains the disturbing cause. The residue, in consequence, again becomes unstable and ejects another cx particle, and the process goes on from stage to stage, until finally the ß particle is violently ejected from the system. Following the general point of view suggested by Sir OLIVER LODGE in a recent letter to 'Nature', it may be possible that, as a result of continuous radiation, the velocity of the ß particle in its orbit has steadily increased, until finally in the last stage a sudden lapse into a new state of the atom occurs, in which not only does an cx particle escape, but also the ß particle. When the ß particle is removed from it, the residual atom adjusts itself again into a position of more permanent equilibrium. The experimental evidence as a whole points strongly to the conclusion that the change in which the ßrays appear is far more disruptive in character than any of the preceding ones; for not only is the ß partic1e thrown off with nearly the velocity of light, but the cx particle, ejected at the same time, has greater penetrating power and probably greater kinetic energy than in any of the other changes. In addition, there is at present some evidence that this final change is of such a violent character that the atom is in some cases disrupted into several fragments, and that, in addition to the cx and ß particles, two or more atoms are produced, each of which has some distinctive physical and chemical properties, and also a distinctive rate of decay. If the greater proportion of the matter resulting from the disintegration is of one kind, it would be difficult to detect the presence of a small quantity of rapidly changing matter from observations of the curves of decay; but if the products have distinctive electrochemical behaviour, a partial separation, in some cases, should be effected by electrolysis. The electrolytic method is a very powerful means of separating active products which may be present in small quantity compared with the other radioactive products. I t has already been mentioned (section 13)

Fhe Successioll of Changes in Radioactil'e Bodies

713

that the results of PEGRAM and VON LERCH, obtained by electrolysis of thorium solutions, may be in part explained on the supposition that thorium A and thorium B have distinctive electrochemical behaviour. PEGRAM, however, in addition, observed the presence of a product which lost half of its activity in about 6 minutes. This active product was obtained by electrolysing a solution of pure thorium salt, to which a small quantity of copper nitrate was added. The cop per deposit was found to be radioactive to a slight extent, and the activity decayed to half value in about 6 minutes. The presence of such radioactive products, which do not come under the main scheme of changes, indicates that at some stage of the disintegration more than one radioactive substance results. In the violent disintegration which occurs in radium C and thorium B, such a result is to be expected; for it is not improbable that there are several arrangements ofthe constituents of the atoms to form a system of some slight stability. The two radioactive products resulting from the dis integration of a single atom would probably be present in unequal proportions. A closer investigation of the radioactive bodies will very probably lead to the separation of a number of radioactive products, present in a small proportion among the main products. 29. Difference between Radioactive and Chemical Change.-The successive changes occurring in the radio-elements are distinguished in certain important particulars from ordinary chemical change. We have seen that each active product, left to itself, is transformed according to adefinite law and at a definite rate. The law of change is the same as for a monomolecular reaction in chemistry, and shows that only one changing system is involved. The constant of change ,\ is independent of the degree of concentration of the product, and of the nature and presence of the surrounding gas, and, in most cases, is not much affected by wide differences of temperature. The work of CURIE and DANNE, however, shows that the constant ,\ of the product radium C (see section 21) is certainly altered by temperature. After this substance has been subjected for a few minutes to a temperature of about 1100 c c., its value of'\ (when cooled to atmospheric temperature) is permanentl)' altered. The value of'\ at 1100 C. is about 1·4 times the normal value. Above 1100" C. the value of ,\ decreases again, and at 1300 C. it is about 1 . 1 times the normal. These results show that increase of temperature to a certain point increases the rate of disintegration of radium C, but that on exposure to a still higher temperature the rate of disintegration decreases again and becomes nearly equal to the normal value. The two features which differentiate the radioactive changes from ordinary chemical change are:0

0

(1) The expulsion of charged particles with great velocity; (2) The emission of an enormous amount of energy compared with the amount of matter involved. Except in the case of the radio-elements, no chemical change is known which is accompanied by an expulsion with great velocity of a product of

z*

714

The Collected Papers

0/ Lord Rutile/lord

the change. In each change that is accompanied by the expulsion of IX rays, the amount of energy liberated, weight for weight, is over 100,000 times greater than has previously been observed in any chemical reaction. Dr. BARNES and the writer* showed that 75 per cent. of the heat-emission of radium was due to the emanation and its further products. The emanation from agramme of radium gives out heat at the rate of 75 gramme-calories per hour. The total amount of heat liberated during its life is 10,000 grammecalories approximately. Now from the work of RAMSAY and SODDY it is known that the volume of the emanation extracted from 1 gramme of radium is not greater than 1 cub. millim. The energy emitted from 1 cub. centim. of the radium emanation is therefore equal to 107 gramme-calories. Now the heat emitted during the combination of 1 cub. centim. of hydrogen and oxygen to form water is 2 gramme-calories. Thus the emanation gives out during its changes 5 x 106 times as much energy as the combination of an equal volume of hydrogen and oxygen to form water. The energy emitted from avessei containing the radium emanation is almost equally divided between the emanation and the products radium A and radium C. Each of these products gives out IX rays. It is probable that the 'rayless' product radium B gives out far less heat than the other products. There seems to be little doubt that the energy emitted from radium is about equally divided between the products which break up with the expulsion of IX partic1es, i.e., 25 per cent. of the total heat emission is supplied in each case by the breaking up of radium, the emanation, radium A and radium C. The energy radiated is, in all probability, mainly derived from the kinetic energy of the expelled IX partic1es. Since the IX partic1es expelled from the products of uranium, thorium, and actinium are projected with ab out the same velocity as from radium, it necessarily follows that each atom of the radioactive products which breaks up with the expulsion of IX partic1e gives out about an equal quantity of energy. This amount of energy is about 6 X 10- 6 erg for each atom at each stage of its disintegration. Since there is the same number of changes in thorium as in radium, the heating effect of thorium will be proportional to its activity, i.e., will be only about 5 X 10- 7 of that from an equal quantity of radium. Since the discovery of the actual production of helium from the radium emanation by RAMSAY and SODDY, there has been a tendency to assume that helium is the final transformation product of radium. There is no evidence in support of such a conc1usion, for, as we have already seen, the radium atom goes on through a further series of slow changes after the first rapid changes have taken place during which the helium makes its appearance. In addition, the evidence supports the view that one IX partic1e is expelled from each atom at each stage of its disintegration, excepting possibly the rayless change. The expulsion of four IX partic1es, of mass about that of the helium atom, still leaves a heavy atom behind. I have previously pointed out that the IX partic1es, in all probability, consist of helium atoms expelled at the • 'Phil. Mag.', February, 1904.

Thc Succession

0/ CIlanges in Radioaetil'e Bodies

715

succösive stages of the disintegration. This conclusion is supported by measurements of the mass of the !X partic1e, and by the observations of the rate of production of helium by the radium emanation made by RAMSAY and SODDY. The similiarity of the oe partic1es from the different radio-elements indicates that they consist of expelled particles of the same kind. On this view, helium should be produced by each of the radio-elements. The presence of helium in minerals such as thorium, for example, in monazite sand, and the Ceylon mineral described by RAMSA Y, suggests that helium is a product of thorium as weIl as of radium. Taking the view that the oe partic1es are projected helium atoms, we must regard the atoms of the radio-elements as compounds of some known or unknown substance with helium. These compounds break up spontaneously and at a very slow rate even in the case of radium. The disintegration takes place in successive stages, and at most of the stages a helium atom is projected with great velocity. This disintegration is accompanied by an enormous emission of energy. The liberation of such a large amount of energy in the radioactive changes at once explains the independence of the rate of change on the physical and chemical agencies at our command. On this view, uranium, thorium and radium are in reality compounds of helium. The helium, however, is held in such strong combination that the compound cannot be broken up by chemical or physical forces and, in consequence, these bodies behave as chemical elements in the ordinary accepted chemical sense. lt appears not unlikely that many of the so-called chemical elements may prove to be compounds of helium, or, in other words, that the helium atom is probably one of the secondary units with which the heavier atoms are built up. 30. The Charge Carried by the oe Rays.-It is of great importance to determine as direct1y as possible the total number of oe particles expelled from a known weight of radium in order to deduce the number of atoms which break up per second. The most direct method of determining this number is to measure the positive charge carried off by the oe rays. Assuming that the charge of the oe partic1e is equal in magnitude to that carried by the ions in gases, the number of oe partieles expelled per second can at once be determined. A thin film of radium was obtained on a plate by evaporation of a radium bromide solution of known strength. Some hours after evaporation, the activity of radium measured by the oe rays is about 25 per cent. of its maximum value, and the ß rays are almost complete1y absent. The arrangement of the experiment is shown in fig. 16. The active plate A was insulated in a metal vessel, C, and was connected to one pole of the battery, the other pole being earthed. The upper plate B was insulated and connected to a Dolezalek electrometer. The outside vessel C could be connected to either A or B, or to earth. By me ans of a

716

The Collec1ed Papers 0/ Lord RU1her/ord

mercury pump, the vessel C was exhausted to a very low pressure. If the