History of Technology Volume 8: Volume 8, 1983 9781350018204, 9781350018228, 9781350018198

Citation preview

H i s t o r y

f

i

o f

T e c h n o l o g y

History of Technology Volume 8, 1983

Edited by Norman Smith

Bloomsbury Academic An imprint of Bloomsbury Publishing Plc LON DON • OX F O R D • N E W YO R K • N E W D E L H I • SY DN EY

Bloomsbury Academic An imprint of Bloomsbury Publishing Plc 50 Bedford Square London WC1B 3DP UK

1385 Broadway New York NY 10018 USA

www.bloomsbury.com BLOOMSBURY, T&T CLARK and the Diana logo are trademarks of Bloomsbury Publishing Plc First published 1984 by Mansell Publishing Ltd Copyright © Norman Smith and Contributors, 1984 The electronic edition published 2016 Norman Smith and Contributors have asserted their right under the Copyright, Designs and Patents Act, 1988, to be identified as the Authors of this work. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. No responsibility for loss caused to any individual or organization acting on or refraining from action as a result of the material in this publication can be accepted by Bloomsbury or the authors. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. History of technology. 8th annual volume: 1983 1.Technology – History – Periodicals 609  T15 ISBN: HB: 978-1-3500-1820-4 ePDF: 978-1-3500-1819-8 ePub: 978-1-3500-1821-1 Series: History of Technology, volume 8

C o n t e n t s Preface W. ADDIS A New Approach to the History of Structural Engineering

1

HANS-JOACHIM BRAUN The National Association of German-American Technologists and Technology Transfer between Germany and the United States, 1884-1930

15

W. BERNARD CARLSON Edison in the Mountains: the Magnetic Ore Separation Venture, 1879-1900

37

THOMAS DAY Samuel Brown: His Influence on the Design of Suspension Bridges

61

ROBERT H J . SELLIN The Large Roman Water Mill at Barbegal (France)

91

G. HOLLISTER-SHORT The Use of Gunpowder in Mining: A Document of 1627

111

MIKULAS TEICH Fermentation Theory and Practice: the Beginnings of Pure Yeast Cultivation and English Brewing, 1883-1913

117

GEORGE TIMMONS Education and Technology in the Industrial Revolution

135

The Contributors

150

Contents of Former Volumes

151

P r e f a c e This year, for the first time, History of Technology does not bear the signature of one of its original editors. Professor A. Rupert Hall has decided to retire from the annual round of reading and correcting manuscripts and proofs. At least for the time being I shall carry on as sole editor mindful of the high standards Rupert Hall set and thankful that I was able to learn at first hand how it should be done. The idea for this yearly presentation of papers was Professor Hall's and it was his energy and persuasion which got Volume 1 off the ground in 1976. His belief then was that there was a need for a periodical that was willing to publish long articles, that would provide space for subjects not traditionally in the repertoire of established publications and, by no means least, could do something to disturb the belief all too commonly encountered that technology's history begins in the Industrial Revolution. Personally I believe that these objectives are as valid now as they were eight years ago: and for nearly a decade our contributors have not disagreed. This year it is especially pleasing to be able to include an up-to-date assessment of one of the most impressive and enigmatic specimens of Roman engineering, the great water-mill at Barbegal. From time to time I shall continue to include original documents and Dr Hollister-Short has provided a particularly interesting one this year. I hope my good friend Bill Addis's paper will not be the last on his theme. Engineering is, and always has been, an intellectual and creative activity in its own right and the history of design is perhaps the most neglected of all aspects of the subject. NORMAN A. F. SMITH

A

N e w

A p p r o a c h

H i s t o r y

o f

t o

t h e

S t r u c t u r a l

E n g i n e e r i n g W. ADDIS

Introduction Anyone with a knowledge of modern structural engineering design methods is likely to meet with two major areas of dissatisfaction when delving into the history of the subject. The first is that the standard histories of the subject1-7 do not tackle the question of how engineers actually designed structures such as Gothic cathedrals, bridges and large span roofs. They deal with the development of various types of structure and with noteworthy exemplars, with the materials used and also, to some extent, the techniques of construction. But concerning design methods, in modern times it is implicitly assumed that design is the 'putting of theory into practice', and that this process is adequately dealt with by highlighting the important milestones in the development of the modern theory of structures and strength of materials. In the days 'before theory', design is deemed to have been a purely architectural matter in which the overall scheme (plan and elevations) and the architectural detailing were arrived at by means of simple geometric constructions sketched on a floor of the building: any structural engineering input to the process lay in the craft skill of the masons et ai, was expressible only as practical rules of thumb and progressed by the primitive means of'trial and error'. Yet such a supposition is as misleading as to suggest that modern structural design* is simply a matter of feeding numbers into a few equations and handing the results to the site engineer. The second area of dissatisfaction with the standard historical approach is that it suggests that there have been two quite distinct periods of structural history with virtually nothing in common between them, one coming before the 'scientific design' of structures using modern theory, and one after. The point of demarcation is presented as being almost as distinct and decisive as the date of the Wright brothers' first flight was in the history of aviation. Two authors have even specified the date of the transition, though it is significant that they chose different dates—17425 and 1815.6 The implication is that the whole approach to modern structural design is totally different from that of, say, Gothic engineers to their design problems. It is as if one was suggesting that after the invention of writing men began to discuss issues different from those they had discussed before. The aim of this paper is to suggest how the present unsatisfactory

2

A New Approach to the History of Structural Engineering

approach to the history of structural engineering has arisen, and to offer an alternative approach which allows the central question of how engineers used to design structures to be approached directly, and which does not lead to the rather implausible 'two period' view of history. A further consequence of the proposed approach is a model for the development of engineering design which parallels the widely accepted model for the development of science put forward by Thomas Kuhn. 8 T h e Conventional Approach to H i s t o r y — T h e o r y a n d Practice It is prevalent to talk of structural design as the process of'putting theory into practice' or of'applying theory'. Such a view is accentuated, especially during an engineer's formative student years, by the almost exclusive concentration upon the mathematics of structural analysis, an emphasis which is nowadays very common. While acknowledging that some academic institutions would not claim to be teaching structural design, the impression gained from nearly all structural engineering textbooks as well as many lecturers is that design is largely a matter of finding the unique solutions to a number of equations. Fortunately many experienced engineers think otherwise and a few have written their views down. 9-11 The argument against the notion that design is a matter of'putting theory into practice' is too lengthy to present here and in any case goes beyond a discussion of the present subject.12 The convention of dividing structural engineering into just two areas, theory and practice, does, however, introduce a number of problems which are relevant also to the history of the subject.13 The most serious of these is that an important aspect of the subject, namely the activity of structural design itself, is almost completely overlooked. Under the history of 'theory' one finds a mixture of the history of statics (a branch of mathematics) and of engineering science (more closely related to physics than to engineering design). In turn, the treatment of statics ranges from the abstruse origins of the notions of force as a vector quantity and of stress as a tensor, through to the many theorems of statics, such as the polygon of forces and Castigliano's theorems, whose status is equivalent to, and as mathematical as, the Euclidean theorems of geometry. (And notice therefore how misleading it is to suggest to students, when they measure the forces in a simple truss, that they are in any way testing the mathematical theory of trusses!) In using 'theory' to mean engineering science, the term is being employed somewhat more accurately because now it covers hypotheses concerning the behaviour of the physical world such as the bending of beams, continuum mechanics and the strength of materials. The resulting equations of statics are the mathematical formulation of the scientific hypotheses, and the quantitative predictions to which they may lead must be subjected to experimental verification (or rather, non-falsification) just as experiments had to be devised and performed to test the wave theory of light in physics. These tests are, however, carried out in engineering science laboratories

W. Addis

3

using idealised components—beams, pin-joints, materials and loading conditions—which cannot be realised in normal structural engineering practice where conditions are not only far from 'ideal', they are very imprecisely known. The other half of the conventional history of structural engineering comprises the history of'practice'—the combined arts of manufacturing, production and construction techniques, or in human terms, the achievements of practical men. Its realm is covered at its most general by the all-embracing history of technology, and more particularly by branches of industrial archaeology and the detailed study of individual structures. This approach, however, places emphasis on how a structure was built rather than on how it, and those like it, were designed. Reference may be made, in passing, to the craftsman's rules of thumb, and at its most philosophical, this approach can touch upon design; for instance, in the way that the use of a certain material, say, masonry, or a certain construction technique, such as movable formwork, can lead to a particular structural form—the inverted funicular arch and the barrel vault, respectively.5 More usually, however, this approach to history leads to an inventory of remarkable structures, although the precise reasons why they are remarkable is nearly always left unsaid, and there arises the dangerous practice of judging past structural achievements only according to present day notions of good design. (The activity of structural criticism, a discipline analagous to musical or literary criticism, sadly has very few followers.14) Thus the conventional approach to the history of structural engineering, through the unconscious application of a historiographical version of the logician's Law of the Excluded Middle, overlooks much of what engineers believe to be central to their activity, namely the actual process of design; the way in which an engineer takes account of, and makes use of, the theoretical aspects of the subject and generates a scheme to guide precisely the efforts of a builder or constructor. In particular, this approach diverts attention away from important and influential matters such as design rules and rules of thumb; empirical constants which frequently modify the bald equations of statics; factors of safety; codes of practice; the results of inquiries into structural failures; tests on completed structures to check their safety or to test the method by which they were designed; the testing of a mathematical model of a structure prior to its being accepted as a valid design tool; laboratory or field experimentation, the aim of which is to devise, develop or check design rules rather than to test theories of engineering science; (it is interesting to see how often this type of empirical work is dressed up as 'respectable' scientific enquiry of the hypothesis-testing kind which Popper has made so famous15); the valuable function of 'incorrect' theory in design, such as in the

4

A New Approach to the History of Structural Engineering

approximation of complex structures to simple models like the pin-jointed truss and the simple beam; curious pseudo-theoretical notions such as the end fixity of struts in compression and Saint Venant's constant (m) which allows the stress distribution across a bent beam to be adjusted to fit experimental results (see Timoshenko3 page 138); All of these matters are of great significance to the designer, yet fall between the two stools of 'theory' and 'practice' (in the present as much as in the past). It was probably Rankine who first drew attention to the middle ground between theory and practice in 1856—'the kind of knowledge intermediate between purely scientific and purely practical'.16 He more than any other academic or engineer, showed how to make use of the apparently unrealistic and hence useless notions of statics by deliberately and systematically simplifying and idealizing real structural elements to make them amenable to the powerful tools of mathematics.17 With the aid of a 'safety factor' to allow for the inherent approximations of the mathematical model, statical calculations could be used for the purposes of designing new structures. Many men previously had developed the mathematics of ideal beams, struts and so on, but in the absence of the careful investigation of the relationship between these mathematical models of structures and the real things, and with the crude notion of the safety factor as the ratio of breaking load to working load, engineers were, before the 1850s, rightly sceptical of the mathematicians' claims for their highfalutin equations. In essence, Rankine showed engineers how they could use statics to help them design real and complex structures. Since Rankine's day few engineers have granted this middle ground the importance due to it and the division between theory and practice, although perhaps no wider, has been accentuated. One effect of this has been to diminish the attention paid to design by academic institutions.18 Another has been to overlook the history of engineering design and thus to generate a distorted view of the history of structural engineering. For example, with the implicit notion that design is simply the putting of theory into practice, it has been assumed that the history was complete if it comprised only the histories of theory and practice. Furthermore, with the tendency to equate theory with modern engineering science and its associated mathematics, the idea has grown up that before this body of knowledge had begun to develop then by definition there could be no such activity as design. Also, the idea is perpetrated that design nowadays is statics or mechanics and nothing but statics or mechanics. Yet such beliefs conveniently ignore both the considerable role played by 'theory' for two millenia before statics was invented, and the considerable role still played by empirical elements such as factors of safety and the various 'coefficients' which are embodied in design rules and formulae. The continuing use of such empirical elements challenges the tacit and widespread belief that modern design is purely scientific and logical. At a more philosophical level, the assumed dependence of design upon

W. Addis

5

logical theory implies that the historical development of design is also entirely logical. By analogy with the Popperian notion of the Logic of Scientific Discovery,15 engineering design is also deemed to develop in an entirely rational way, dependent upon a rigorous programme of experimentation. Yet, in this light, it is difficult to account for utterly unpredicted changes in the course of design methods, such as that which affected suspension bridge design after the collapse of the Tacoma Narrows bridge (in this case a static problem instantly became one of statics and dynamics). It is also difficult to account for the fact that there are today two logically incompatible, indeed uncomparable, methods by which many steel structures can be designed (elastic and plastic). Similarly, if engineering design and analysis be such logical affairs, how can it be that there is often irreconcilable disagreement between experts, for instance about the way in which a Gothic cathedral 19,20 or a wooden roof truss21 work as structures? These philosophical issues have particular relevance to the process of the historical development of design methods and will be raised later in the paper. The immediate step is to suggest a way of avoiding the problems and inadequacies of the conventional approach to the history of structural engineering which have been noted above. It seems to the writer that it is entirely unreasonable that the history of structural engineering should be approached only from the viewpoints of mathematicians, theoretical or experimental engineering scientists, technologists or industrial archeologists. There should be at least an attempt to approach it from the point of view of the engineer, from that middle ground between theory and practice which Rankine identified as the particular domain of the engineer. With this aim in view the suggestion now is that the historical parameter should be the engineer's 'design procedure'. One advantage of this approach is that the idea of a design procedure is not new: indeed it is used naturally by engineers when talking of their work. 9,10 What is new is its use as a historical tool and to this end it requires special elucidation. T h e N e w Approach t o H i s t o r y — t h e D e s i g n P r o c e d u r e Reduced to its most basic elements, the task of a structural engineering designer is to specify the material and size of each constituent of a complete structure and to indicate the relative dispositions of these constituent parts. This is a necessary element, but it is not sufficient, for the specification must also be justified (or justifiable). The process by which a proposed structure is specified and justified is the design procedure. Expressed in this general way there are clearly no restrictions as to how a design is specified and justified. Indeed in the specification there is a considerable freedom of choice—exact shapes and sizes, even today, frequently appear inexplicably from an engineer's head, tempered by the requirements of Codes of Practice and the availability of most materials in certain standard forms. Many structural aspects of a design are chosen for such 'illogical' reasons as the aesthetic impact in terms of visual appearance or structural cleverness, or of

6

A New Approach to the History of Structural Engineering

requirements imposed by the use of a structure, or convenience of construction, or cost, or even current fashion. A separate category of reasons is of a more rational kind. With half a mind on the need to justify a particular design, the engineer is likely to choose certain shapes, sizes and dispositions of structural elements which are particularly easy to justify, or which can be justified with more certainty than others. A nice example is provided by the Bank of England's building at Debden, near London, whose cross section is part of the curve known as the Lemniscate of Bernoulli—a choice made not for reasons of statical elegance, but because the curve provided the desired shape of the building (smoothly assymetric) and it has a single convenient mathematical equation which made the justification (the stress analysis of the thin shell) relatively simple. The justification of a proposed design has only one function—it must convince certain people that the structure will stand and perform as required by the client and, from the point of view of public safety, by society at large. The justification must, then, convince the engineer himself, the client, society (represented by building inspectors and the like) and, in the last resort, the committee of inquiry into the structure's collapse. In achieving this goal there is, as with the method of specification, a considerable freedom of choice. The precise methods of justification may vary considerably from structure to structure (compare a suspension bridge and an ordinary block of flats), from engineer to engineer and firm to firm (compare the use of elastic and plastic design procedures or the choice between different mathematical models used to represent a structure), from place to place (compare the building codes of different countries) and from time to time (compare the periods before and after the box girder bridge failures of the early 1970s, or the many ways in which nearly identical masonry arches have been justified over the last thousand years1,2> 19 ). There are, then, no precise criteria by which a method of justification can be supported. For a given structure the proposed justification must simply appear appropriate to the people who have to be convinced by it. The argument presented must be believed by the relevant people. Or, if they do not initially believe it, then the argument must be able to persuade them. (It is not unknown for borough engineers to be unfamiliar with the latest and most sophisticated techniques of analysis which may be used by large firms of structural engineers, especially if the structure concerned is of an unusual type.) The factors which influence the acceptance of a method of justification in a design procedure are clearly many and range well beyond the pages of textbooks of structural analysis. A list of the factors can only be vague—the experience of a whole engineering community, of a firm or a particular engineer; research work carried out by engineering and other scientists; the level of technical education in a country; the legal system and Codes of Practice; the skills of the construction work force; financial and political pressures, and so on. Pugsley22 (originally concerned with structural safety) has coined a useful phrase to summarise these many influences—the engineering climatology. This simple requirement, that a design

W. Addis

7

procedure be generally accepted by the community, is vindicated by the fact that inquiries into terrible structural accidents frequently exonerate the designers (even when errors of judgement are identified retrospectively) if they can show that they approached the design in a responsible manner and in accordance with the accepted design procedures of the day. This was the case with the disasters which befell the Tay Bridge (1879), the Tacoma Narrows Bridge (1940) and the Milford Haven box girder bridge (1970). Having introduced the design procedure in the above manner, its uses as a historical tool can now be considered. Most importantly, the historian's task is focussed directly upon the design engineer's central activity rather than upon only practical matters and two bodies of knowledge upon which modern engineers rely heavily (statics,and engineering science). It becomes just as important to consider the many other ,bodies of knowledge upon which he might draw, such as precedent, 'rules of thumb' and design rules not based solely upon modern engineering science, the empirical testing of model and real structures, personal experience (tacit knowledge) and the 'folk experience' which is passed on within the engineering community— indeed anything which he may feel is relevant in a particular instance (many engineers have openly copied structural ideas from nature!). Concerning the use which an engineer makes of'theory', it becomes equally significant how he makes use of it, which particular mathematical model or calculation technique he uses and how the results of such calculations are modified by safety factors and other empirical coefficients. A calculation which yields the answer 20 and is increased by a safety factor of 1.5 gives the same final dimension as a calculation yielding 6 modified by a safety factor of 5. By placing the various branches of 'theory' alongside other bodies of knowledge as subservient to a designer's needs, the historian is better able to follow the progress of structural engineering design, since the fundamental requirements of design (specification and justification) have not changed over many centuries. It begins to emerge that the problems facing the designer of a Greek temple, a Gothic cathedral or a nineteenth century railway station roof were substantially the same as those facing a modern engineer. The differences are, relatively speaking, ones of detail—which particular calculation techniques and branches of mathematics; which other branches of human knowledge and which theories must be called upon to specify and justify a design. This approach enables us to draw parallels between different periods, such as between the cathedral builders of twelfth century France and the railway station and bridge builders of the middle of the last century. This is an example which it is worth elaborating briefly. With geometry in the Middle Ages and graphical statics in the nineteenth century engineers of these two eras had developed theoretical tools of great power which enabled them to construct the greatest structures of their day. Both geometry and graphical statics had the power of simultaneously specifying a design and justifying it. Both had existed already but each benefited from a crucial development which enabled them to be used to create structures of unprecedented boldness. (The engineers were, of course, also driven by other aspects of the engineering climatology, not the least of which

8

A New Approach to the History of Structural Engineering

was a desire in the respective societies to create particularly bold structures.) Practical geometry was well known to builders throughout the 'dark' ages, but it was merely a trade skill; it lacked any explanatory powers. In Chartres in the mid-twelfth century there conjoined three separate elements to change the status of geometry. One was the practical skill itself. The second was a particular school of theology and cosmology based upon Augustine's doctrines, in which harmony and proportion were of great significance. And third there arrived from England one of the very first editions of Euclid's work on geometry translated from Arabic into Latin. 23,24,25 Together these three elements provided the means by which geometry could be used both to describe and to explain the world. The masons had a tool by which they could justify to themselves, to their clients and to society the bold designs for which they are now famous. Precisely how they used their geometry was, of course, their secret, a secret which is still slowly being unravelled. 26,27,28 By the middle of the nineteenth century the mathematical subject of statics had been very fully developed. Vector diagrams were well known. But this knowledge had been shown to be of relevance to the real world of structures in only the simplest of cases, simple beams and a few bars joined by frictionless hinges. These exemplars were clearly of little use to the designer of a large roof or bridge truss. However, in statics lay the potential to explain the world and its behaviour. What was still lacking was a clarification of the relationship between real structures and the simplified, idealised models of them. As mentioned above, it fell to Rankine to show how this could be done. By using the factor of safety in a new way, to mask any discrepancies between the behaviour of a real structure and the model used to represent it, the designer could possess a powerful tool. In a relatively short time the subject of graphical statics was developed to exploit the full potential of truss structures.29 Indeed it only waned in popularity with the increasing use of steel in bending, rather than in axial load, for which graphical statics was more cumbersome than algebraic and arithmetic methods. It will be well known to anyone who has perused the engineering periodicals of the latter half of the last century that the topics of theory and practice, and more particularly the uselessness of the one without the other, were favourite topics of debate. Thus it strengthens the present comparison to encounter the very same arguments being put forward some seven hundred years earlier.24 In this section the author has intended merely to introduce a new approach to the history of structural engineering. It would be the subject of a specialised paper, rather than a general one, to work fully through particular examples. Continuing the general approach, it remains now to show how the ideas developed above allow the possibility of creating a model for the historical development of structural engineering along similar lines to those employed by Thomas Kuhn in his books The Essential Tension30 and The Structure of Scientific Revolutions*

W. Addis

9

The H i s t o r i c a l D e v e l o p m e n t o f Engineering D e s i g n In summary, Kuhn's ideas sound deceptively obvious, yet they represent a radical departure from the conventional approaches to the history of science. These, he claims, portray the development of science as a fundamentally logical process; it is held that debate between rival ideas is entirely rational, that the process of development has a rigorous methodology and that science is the archetypal cumulative body of knowledge. Kuhn argues that none of these statements is true. He suggests that there are, in fact, two types of scientific development. One, normal science, is indeed largely logical and cumulative. He likens it to puzzle solving where the rules of the puzzle have been laid down and the game is played according to these rules. No radical departure from current beliefs is sought, and none is usually found. These relatively stable and peacefully uncontroversial periods of normal science are, Kuhn argues, punctuated by periods of another type of scientific development—revolutionary science. If normal science is playing according to the rules, then revolutionary science is changing those rules. It is a time of upheaval, of radical change, of revolution. Most particularly it is an utterly irrational period. During such a period, rather than the body of scientific knowledge growing cumulatively, the old body of knowledge is, to a greater or lesser extent, replaced by the new. Being logically incompatible the two cannot be fused together. There is, Kuhn suggests, a good reason why scientific development does indeed usually appear logical. It is that, after a scientific revolution, the history of science has to be rewritten to fit in with the new logical framework created by the revolution. History is thus written backwards, it always appearing that past scientific developments lead, almost deliberately, towards the present day. This process, however, knowingly ignores any past achievements which do not fit into the current logical framework. Thus, the caloric theory of heat is treated as no more than a curiosity which we now know to have been, simply, 'wrong'. The fact that the caloric theory, in its day, was highly successful, is usually ignored. Kuhn argues that history should be written forwards, not backwards. Only in this way can the historian hope to understand how scientists actually thought about their problems and how they perceived their attempted solutions. In the light of Kuhn's view of the conventional history of science, a deeper understanding can now be reached of the problems arising out of the conventional history of structural engineering which were outlined above. This conventional approach is also guilty of writing history backwards and of giving the impression that events have been leading almost inevitably to our present position. There is, thus, an exaggerated interest in examples from the past which show evidence of the use of statics and modern techniques of structural analysis. The result is that the history of structural design is usually portrayed as the history of design using statics. Straub is not alone in claiming that before the 'quantitative application of mathematical and physical laws . . . shapes and dimensions [of structures] were determined merely by what may be called "trained intuition"' (Straub2

10

A New Approach to the History of Structural Engineering

p. 7). At the other extreme, there are many authors who credit Leonardo with the real beginnings of statical design methods (despite a total lack of evidence), and one author goes as far as asserting that the Gothic engineers must have used statics, for how else could they have designed and built their cathedrals.31 A very recent example of the conventional historical position being rejected after a 'structural design revolution' is presented by the case of Coulomb (late eighteenth century). His work on arches and beams fell into total obscurity after elastic design methods took hold of the engineering community in the 1820s. Only recently has he been reinstated to his rightful position following the plastic design revolution of the 1950s, because only then could it be seen that he had made an early contribution to the development of limit state design methods.32 The plastic design revolution is a good illustration of the characteristics of engineering design revolution generally. It was born out of a crisis which arose in the 1930s in the use of design procedures based upon the elastic analysis of structures. In this approach, failure was defined by the load which caused the highest stress in the structure to reach a prescribed maximum value (a certain fraction of the yield stress). For many years the logical progress of elastic design procedures had followed the path of developing more accurate mathematical models, more powerful calculation techniques, load models which more accurately represented the real loads, and more accurate techniques for predicting the 'actual' strength of the materials (mainly wrought iron and steel). The crisis point was reached when the effort to predict stresses with more precision came into conflict with a number of common sense observations. Although the mathematical models of buildings resting on many supports suggested that the differential sinking of these supports would lead to high stresses and hence to failure, it was well known that many such buildings appeared utterly safe. It was also well known that, especially in riveted joints, steel could safely yield at highly stressed points without really failing, and that the high stress was effectively redistributed to areas of lower stress. Moreover it was found that most rolled-steel sections contained high locked-in stresses which made a mockery of the elastic design assumption that, at zero load, the material is unstressed. This assumption was also known to be violated by the stresses created in the construction process when inaccurately made parts had to be forced together. All these fears were finally corroborated when the development of strain gauges made it possible to measure the sometimes enormous differences between the predicted and the actual stresses found in completed structures. A few people saw that the elastic design procedures had reached the limit of their logical development and so they pursued a new concept which looked at the problem of design from a completely different point of view. To the 'plastic' designers, stress was of no interest at all. Their only parameter was the resistance of a section to further bending after a fully plastic hinge had formed—the fully plastic moment. Such a state was well beyond the elastic designer's failure criterion—and thus the two schools of thought

W. Addis

11

were bound to adopt quite different definitions of failure. As to the calculations used in the design procedure, these too were utterly different. The advocates of each type of design procedure might as well have been speaking different languages as indeed in a sense they were. There was no conceivable way in which the plastic designer could engage in rational debate with the other, nor convince him by logical argument to change his views. The gap could not be bridged by logic any more than could the gap between believers in the geocentric universe and the heliocentric universe be rationally bridged in Copernicus' time.8 Such gaps can only be bridged by a leap of faith. Nevertheless 'elastic men' do try to argue on logical ground that their 'plastic' colleagues are mistaken; but there is never any agreement. 21,33 Conclusion By approaching the history of structural engineering in terms of the designer's method of tackling his problem, what we call the design procedure, two major inadequacies of the conventional approach to the subject can be overcome. The middle ground between theory and practice can be given due emphasis and the tendency to write history backwards and to write it from a present day view of structural design, can be avoided. If this practice is followed, then it can be expected that the central activity of structural engineers of the first, twelfth, nineteenth or twentieth century, namely design, will be more accurately representative of their respective points of view than has been usual hitherto. While the present paper has concentrated upon structural engineering, the idea of the design procedure is of relevance to other branches of engineering too. Although still some way off, it would appear worthwhile to broaden the application of the approach advocated here, to take note of similar work in other branches of engineering history,34,35 and to formulate a generalised theory of development in the history of engineering. In so doing, the intellectual side of engineering could be accorded its rightful place alongside the more practically orientated discipline of the History of Technology.

Notes 1. W.B. Parsons, Engineers and Engineering in the Rennaissance, Williams & Wilkins, Baltimore, 1939; reprinted MIT Press, Cambridge, Mass., 1968. 2. H. Straub, A History of Civil Engineering, Leonard Hill, London, 1952. 3. S.P. Timoshenko, History of Strength of Materials, McGraw Hill, New York, 1953. 4. J. P.M. Pannell, An Illustrated History of Civil Engineering, Thames and Hudson, London, 1964. 5. R.J. Mainstone, Developments in Structural Form, MIT Press, Cambridge, Mass., 1975. 6. H.J. Cowan, The Masterbuilders, Wiley, New York, 1977; Science and Building, Wiley, New York, 1977.

12

A New Approach to the History of Structural Engineering

I. D.I. Blockley, The Nature of Structural Design and Safety, Ellis Horwood, Chichester, 1980. 8. T.S. Kuhn, The Structure of Scientific Revolutions, University of Chicago Press, Chicago, 2nd edition, 1970. 9. P.L. Nervi, Structures, F.W. Dodge Corp., New York, 1956. 10. E. Torroja, Philosophy of Structures, University of California Press, 1967. II. O. Arup, 'Furture Problems Facing the Designer', in Arup Journal, Vol. 7 No. 1, March 1972, pp. 2-4. 12. W. Addis, Theory and Structural Engineering, Thesis to be submitted at Reading University, 1983. 13. W. Addis, 'Some Problems of Looking at the History of Structures in Terms of 'Theory' and 'Practice', and a Proposed Solution,' Presented at International Association for Bridge and Structural Engineering Colloquium on History of Structures, July 1982. Proceedings to be published late 1983. 14. P.L. Nervi, 'Critica delle Strutture,' Casabella, No. 223, January 1959, pp. 55-6; No. 224, February, 1959, p. 54; No. 225, March 1959, p. 50. 15. K.R. Popper, The Logic of Scientific Discovery, Hutchinson, London, 1977. 16. W.J.M. Rankine, Introductory Lecture on the Harmony of Theory and Practice in Mechanics, Richard Griffin & Co., Glasgow, 1856. 17. D.F. Channell, A Unitary Technology: The Engineering Science of W.J.M. Rankine, PhD Thesis, Case Western Reserve University, 1975. 18. A. Pugsley, 'The Teaching of Theory of Structures', The Structural Engineer, Vol 58A, No. 2, February 1980, pp. 49-51. 19. J. Heyman, 'The Stone Skeleton,' Int. J. Solids Structures, Vol. 2, 1966, pp. 249-79. 20. R. Mark, 'Wind Loading on Gothic Structure', Journal of the Society of Architectural Historians, Vol. 29, 1970, pp. 222-30. 21. J. Heyman, 'Westminster Hall Roof, Proc. Instn, Civ. Engrs., v. 37, May 1967, pp. 137-62; discussion, v. 38, December 1967, pp. 785-96. 22. A. Pugsley, 'The Engineering Climatology of Structural Accidents,' International Conference on Structural Safety and Reliability, Washington, 1969, pp. 335-40. 23. O.G. Von Simson, 'The Gothic Cathedral: Design and Meaning,' Journal of the Society of Architectural Historians, Vol. 11, 1952, pp. 6-16. 24. S.K. Victor, 'Practical Geometry in the High Middle Ages', The American Philosophical Society, Philadelphia, 1979. 25. F. Bucher, 'Design in Gothic Architecture', Journal of the Society of Architectural Historians, Vol. 27, 1968, pp. 49-71. 26. G. Kubler, 'A Late Gothic Computation of Rib Vault Thrusts,' Gazette des Beaux Arts, Vol. 26, 1944, pp. 135-48. 27. L.R. Shelby, Gothic Design Techniques, Southern Illinois University Press, Carbondale, 111., 1977. 28. L.R. Shelby, and R. Mark, 'Late Gothic Structural Design in the "Instructions" of Lorenz Lechler', Architectura {Munich) Vol. 9, 1979, pp. 113-31. 29. K. Culmann, Die Graphische Statik, Zurich, 1866. 30. T.S. Kuhn, The Essential Tension, University of Chicago Press, Chicago, 1977. 31. A. Hertwig, 'Aus der Geschichte der Gewolbe—ein Beitrag zur Kulturgeschichte,' Technikgeschichte, Vol. 23, 1934, pp. 86-93. 32. J. Heyman, Coulomb's Memoir on Statics, Cambridge University Press, Cambridge, 1972. 33. J. Heyman, 'Plastic Design and Limit State Design,' The Structural Engineer,

W. Addis

13

Vol. 51, No. 4, April 1973, pp. 127-31; correspondence, Vol. 51, No. 11, November 1973, pp. 434-5. 34. M. C. Duffy, 'The Evolution of Engineering Design Technique,' Engineering Designer, Vol. 5, 1979, Jan/Feb pp. 19-22, Mar/Apr pp. 19-22, May/Jun pp. 21-6, Jul/Aug pp. 31-5, Sep/Oct pp. 19-23, Nov/Dec pp. 21-6. 35. E.W. Gonstant, The Origins of the Turbojet Revolution, Johns Hopkins University Press, Baltimore, 1980.

T h e

N a t i o n a l

A s s o c i a t i o n

o f

G e r m a n - A m e r i c a n T e c h n o l o g i s t s T r a n s f e r

a n d

b e t w e e n

T e c h n o l o g y G e r m a n y

a n d t h e

U n i t e d

S t a t e s ,

HANS-JOACHIM

1 8 8 4 - 1 9 3 0 BRAUN

I Most of the research in the last two decades on the general topic of technology transfer has concerned developing countries.1 There is little information, however, on technology transfer, especially in a historical perspective, between countries of a roughly equal technological level.2 This is true, as I shall show later, of the United States and Germany in the period under review. The topic of (multinational) corporations and technology transfer has also been quite fashionable during the last decade. The historical work, which might be of interest in my context, deals, however, almost exclusively with American multinational firms in Europe and not the other way round. 3 So far as the literature on technology transfer is concerned, classificatory models have often been put forward.4 The mechanisms, however, which bring technology transfer into being are generally explained only unsatisfactorily. Anthropological models deal with technology in a static way and often ignore the process of adaptation and modification.5 Economists and economic historians normally stress only economic incentives and disregard non-economic factors.6 This paper centres on technology transfer in mechanical engineering, although a case of chemical engineering will also be treated. The cases reviewed deal with refrigeration, steam pumps, steam turbines, injectors, elevators, uni-flow steam engines, coke ovens and gasholders. For lack of space only the latter three cases will be treated in some detail, whereas the cases mentioned earlier will be considered very briefly. Common to all the cases listed above is the fact that in their transfer to the United States the National Association of German-American Technologists7 played a more or less important role. I shall briefly describe this

16

Technology Transfer between Germany and the US, 1884-1930

association and then put my cases of technology transfer in a general perspective of the problem of technology transfer. Before talking about my case studies, some hypotheses in connection with them will be mentioned. The National Association of German-American Technologists existed between 1884 and 1941 and its journal, The Technologist, was published in German in New York and later in Philadelphia from 1896 onwards. Its forerunner, Der Techniker, was published between 1878 and 1896. The Association had members in many cities, where the German element was strong and where large industry was to be found, especially in New York, Philadelphia, Chicago, Baltimore and Pittsburgh. It had about 600 members, organized exhibitions, yearly meetings and was instrumental in helping German engineers who wanted to work in America to find a position there. The Association was midway between a proper professional engineering association and a social club, and played quite an important role in the dissemination of engineering knowledge within the United States. Looking at the level of technology, especially in mechanical engineering, in the United States and Germany in the period 1870-1939, the following general remarks can be made. Although at the end of the nineteenth century the United States was more advanced than Germany in the field of machine tools and some types of power machinery (large machinery for mining and for the iron and steel industry), Germany was—generally speaking—superior in sophisticated power technology.8 This was mainly due to research in German institutes of technology, although the standard of similar American institutes should definitely not be underrated. In the manufacture of highly standardized products, e.g., boilers, hydraulic presses, elevators, the United States was more advanced than Germany,9 although there are many cases in which the relevant inventions had been made in Europe but were later transferred to the United States because of considerations of scale.10 Standardization, specialization and mass production were used earlier in America than in Europe. A notable difference lay in the structure of productive factors. As is well known, labour was, generally speaking, scarcer and therefore more expensive in the States than in Europe, whereas the raw material situation was more favourable in America.11 Productivity in America was commonly higher because of the different input of productive factors. Of the three productive factors—land, labour and capital—land was cheaper in the United States than in Germany, whereas labour, especially skilled labour, was scarcer and therefore more expensive. Capital seems to have been more readily available in America. There was a tendency in the United States to substitute labour by capital. Mechanization (the American system of manufacture), which was very capital-intensive, assured in the branches in which it was applied a higher productivity in America than in Germany around 1880. The motivations to transfer technology to another country, sometimes connected with the establishment of a subsidiary, can rest in a difficult economic situation in the mother country and a better situation in the country to which the technology is to be transferred. Also the exchange rate can play a role. Other incentives result from a more favourable distribution

Hans-Joachim Braun

17

of productive factors, high freight rates, a larger market, attempts to avoid protective import duties (therefore direct investments), more favourable corporate legislation, or weaker trade unions.12 In my cases, the larger market in the United States seems to play the dominant role. Technology can be transferred by several means, including licences and direct investments. As I will show, licensing played the most important role although, in later stages of the process, direct investments also took place. Licensing was possible and successful because, as already indicated above, the technological level of the two countries was similar. German firms took out patents in America, but this did not help them very much because American firms quite often ignored those patents and patent law suits were expensive and took a long time.13 It is usually necessary for a successful transfer and diffusion of a new technology for one to first adapt and modify the innovation to make it suitable for the new market.14 This process of adaptation is often neglected in technology transfer to developing countries, but will be shown below. II Looking at the membership lists of the National Association of GermanAmerican Technologists in New York City it is striking that several of the members held important positions in the De La Vergne Machinery Company, which was especially concerned with building refrigerators.15 Quite often the topic of refrigeration and ice-making was discussed at the meetings of the Association, which means that it must have been of great interest to its members. Although a lot of research had been done, especially in the United States, on the problem of refrigeration, it was Carl von Linde, a professor in Munich, who built the first successful ammonia compressor and made a complete theoretical analysis of the compression cycle of refrigeration.16 He built his first ammonia refrigeration machine in 1880. Fred W. Wolf, a member of the National Association of German-American Technologists in Chicago, knew of Linde's work. Wolf had studied mechanical engineering at the Technical Institute in Karlsruhe and later became an assistant to the famous mechanical engineer Ferdinand Redtenbacher, who is often regarded as the father of scientific mechanical engineering in Germany. In 1867, Wolf settled in Chicago as a consulting engineer and in 1875 established a small mechanical engineering firm. He realized the importance of ice-making, especially for breweries.17 In 1879, Wolf visited Carl von Linde in Munich and acquired the rights to sell Linde machines in the United States. At first Wolf obtained compressors and steam engines from Linde, but he soon built them himself in his newly erected works in Chicago according to Linde's specifications.18 By 1912 Wolf had built about 1,200 commercial and industrial refrigerators.19 As is well known, competition in the field of refrigeration was very strong. One of Wolf's competitors was John C. de la Vergne who, in collaboration with William H. Mixer, built a vertical twin compressor that was oil-sealed and

18

Technology Transfer between Germany and the US, 1884-1930

directly connected to a horizontal steam engine.20 De la Vergne and some of the leading engineers in his factory, in which he later also produced general machinery parts, were members of the Association in New York.21 Although nothing is known of direct contacts with Carl von Linde, it is highly likely that he knew Linde's work well. He recruited many of his engineers from Germany via the Association. So, in the case of F. W. Wolf and refrigeration, contacts with a German firm were started by a GermanAmerican. There was no direct investment by the German firm, only a licence was given. Another field in which German engineers in the United States and members of the Association played an important role was the construction of steam pumps. There were several German-American engineers with the Henry R. Worthington Hydraulic Works—they were members of the New York branch of the Association22—and with the Fraser & Chalmers Machinery Works in Chicago.23 In the second half of the nineteenth century, Worthington pumps were the most popular in the United States and the firm had several branches in Europe, for example in Berlin. Worthington's steam pump was patented in 1841.24 In 1859 he invented the duplex pump, which consisted of pumps placed side by side. As one pump neared the end of its stroke the other one discharged water to keep up the flow while the first reversed. In 1885, Worthington brought out his high-duty engine, built exactly as the duplex engine but with the addition of a compensating cylinder to use steam expansively.25 Although Worthington's pumps were popular, several engineers at the end of the nineteenth century tried to enhance their efficiency.26 One of them was the German Professor Alois Riedler in Berlin. The principal feature of Riedler's pump was an improved mechanically-operated valve. The valve and valve seat were circular in form and made of high-grade bronze. At the beginning of the stroke the valve opened automatically and remained open practically during the entire stroke. When near the end, it was positively closed at the proper moment by the controller.27 As the valve was large, all throttling of the water through the valve passages was avoided. The mechanical controller, closing the valve at the proper moment, prevented slip and allowed the pump to run at any desired piston speed. Also, the injurious effects of strains due to 'water hammer' or 'hydraulic shocks' were eliminated. These pumps were designed by Riedler and his assistant Johannes Stumpf, of whom more will be said later. The engineers of Fraser & Chalmers in Chicago, many of whom were German-Americans, knew about Riedler's pumps. In 1884, the same year in which Riedler and Stumpf constructed the pump, several German-American engineers from Fraser & Chalmers examined it working in mines in Germany and obtained the manufacturing rights for the United States.28 Riedler and Stumpf, who continued improving the pump, were later also visited by representatives of other American firms. In 1897, for example, the East Jersey Water Company wrote to Riedler inquiring about details of his pump. He sent them detailed descriptions and a drawing of it. After this, the company sent one of their engineers, the German-American Clemens

Hans-Joachim Braun

19

Herschel, to Berlin. He received instruction for building the pump in the United States, and it was later built by the Dickson Manufacturing Company in Scranton.29 At the beginning of the twentieth century pumps were built in America, under licence from Riedler and Stumpf, and were called 'Express Pumps'. These were normally driven directly by an electric motor without the use of gears to cut down the speed.30 Johannes Stumpf, although not a member of the Association, spent the years 1893 to 1896 as a chief engineer at Fraser & Chalmers in Chicago, where he was employed in the construction department for pumps, compressors and Corliss steam engines. Stumpf s career is interesting indeed. He was a student of Riedler when the latter was teaching mechanical engineering at the Technical Institute at Aachen. When Riedler was called to the chair of mechanical engineering in Berlin, he took Stumpf with him as his assistant. At that time Stumpf improved Riedler's steam pump and, together with Riedler, started developing a steam turbine. Riedler and his assistants visited the Chicago World Exhibition in 1893. During their stay they visited Fraser & Chalmers, and Stumpf decided to leave the Technical Institute in Berlin to become head engineer there. In 1896, however, he was offered a chair of mechanical engineering, especially steam engines, at Berlin, which he accepted.31 He took Oscar Lehmann, a member of the Association who had worked under him at Fraser & Chalmers, to Berlin as his assistant.32 The Riedler-Stumpf steam turbines were described by Riedler in the Technologist of 1904. They had constructed this turbine in the engineering laboratories of Charlottenburg. The Riedler-Stumpf turbine was an impulse turbine like the well-known de Laval turbine; however, the rotors were larger and therefore the speed of revolution was decreased.33 Also, another advantage of the Riedler-Stumpf turbine was that the blades were not bolted on the wheel but milled.34 This turbine was built in Germany by Krupp for the AEG (German General Electric Company) and Riedler wanted to make them known in the United States as well. However, he was not very successful. In 1902 the AEG and General Electric founded, together with the Curtis Steam Turbine Co. and the International Steam Turbine Co., a new company which was called Vereinigte Dampfturbinen Gesellschqft*h (United Steam Turbine Company). It was the Curtis steam turbine which became popular in the United States, not the Riedler-Stumpf type.36 Even the AEG in Germany preferred Curtis turbines to RiedlerStumpf steam turbines because the shafts of the latter proved too thin, and there were also difficulties with the construction of the speed regulator.37 So this attempt at technology transfer, in which the National Association of German-American Technologists played a role, failed. Two mechanical engineering firms also had close connections with the Association, the firm of Korting & Schiitte in Philadelphia and the Otis Elevator Company in Yonkers, New York. Both firms used the Association and its journal as a means of advertising their products in America and Germany. The mechanical engineer A. Schiitte was a friend of Ernst Korting, who built gas engines in Hanover. A speciality of his was injectors. Korting tried to sell them in the United States and sent Schiitte there to establish a

20

Technology Transfer between Germany and the US, 1884-1930

plant in Philadelphia. In the beginning, Schiitte was unsuccessful because of patent difficulties. William Sellers had obtained a licence to build the Giffard injector and resisted Schutte's attempt to capture the American market with the Korting injector. Schiitte, who had worked in Sellers' firm, later established his own firm, which produced a new kind of injector, Korting's 'universal injector'. This firm seems to have done very well.38 Several of the Association's members who lived in New York worked for the Otis Elevator Company in Yonkers. In Berlin, the firm of Carl Flohr was working in the same field, although the Otis company seemed to be more advanced technically, and Flohr received information on the development of elevator manufacturing from Otis. In 1878, the first hydraulic elevator was installed in New York and a year later Flohr started building them. The Otis Electric Company built its first electric elevator in 1892; Flohr built his in 1893.39 Although there is no evidence that all the know-how of the German engineers had come from Otis, there were close contacts which, after a long period of deliberation and negotiation, led to the merger of the subsidiary of Otis in Berlin and the Carl Flohr Company in 1951. Ill So far, only a few brief remarks have been made on the role of members of the National Association of German-American Technologists and technology transfer from Germany to the United States. It was pointed out that German-American engineers, who worked in various American mechanical engineering firms, were active in transferring technology from Germany to America and also, at times, the other way round. The journal of the Association, The Technologist, also played a central role in technology transfer. The following three cases, in which the Association or some of its members were active in technology transfer, will be presented in more detail. They deal with the Stumpf uni-flow steam engine, by-product coke-oven technology and waterless gasholders. In these case studies some of the general problems of technology transfer mentioned at the beginning will also be examined. The emphasis of this section of the paper shifts from the role members of the Association played in technology transfer to transfer of technology between German and American firms, because the information available on the role members of the Association played in this process is scanty. After some work had been done in the nineteenth century to develop a uni-flow engine, especially by the Englishman Leonard Jennett Todd, who in 1885 obtained a patent, it was the German Professor Johannes Stumpf of the Institute of Technology in Berlin-Charlottenburg who designed and in 1908 patented an efficient valve gear to suit the engine. It was taken up by the Erste Briinner Maschinenfabrik and licences were taken out by several continental and British firms. By the outbreak of the First World War, several hundred 'uni-flow engines' had been built for driving mills, fans and electric generators and for propelling ships and locomotives.40

Hans-Joachim Braun

21

Contrary to normal steam engines, which are counter-flow, the steam in the uni-flow engine, after passing through steam-jacketed heads, enters the cylinders at the ends and then, after cutoff and expansion have taken place, it is exhausted through ports around the centre of the cylinder, which are uncovered by the piston at the end of the stroke. So the steam has a flow in one direction only, hence the name uni-flow engine.41 In the counter-flow engine, the steam returns along its path at the end of the stroke and is exhausted at the same end of the cylinder by which it entered. As a result, the cold expanded steam washes the cylinder walls and head during 50 to 75 per cent of the return stroke, thereby cooling them to such an extent that the high-pressure steam, when it is again admitted, is cooled or condensed by coming into contact with the head clearance spaces of the cylinder, which have just been cooled by the expanded exhaust steam. It was to remedy this fundamental defect of the counter-flow engine that successive expansion stages were resorted to, as in compound or tripleexpansion engines. Superheating was also employed, but this could not be effected without high cost in installation and operation. The advantages of the uni-flow engine were that it was more economical than multi-cylinder horizontal engines, occupied less floor space and was lower in upkeep because of the small number of working parts.42 In 1896, Stumpf was commissioned by the Pope Manufacturing Company of Hartford, Conn, to set up two Riedler-Stumpf pumping engines. The pumps were driven by vertical triple-expansion engines with Corliss valves and a central condensing system. Stumpf tried to simplify this complicated construction and started research on this matter, which led him to the construction of the uni-flow engine. In his research he had the internal combustion engine and especially the steam turbine as his models. The construction of steam turbines of several stages, which began at that time, was developed along the lines of pure unidirectional flow, and this brought up the question of whether it would not be possible to raise the reciprocating steam engine to the same thermal level as the turbine by the use of the uniflow principle.43 The first Stumpf uni-flow steam engine was built in 1907. It was important for its further development that research in German institutes of technology concentrated on the improvement of the engine. In Dresden, Dr Mollier—of Mollier-Diagram fame in thermodynamics—worked on the thermodynamic aspects of the uni-flow engine. He and Stumpf approached the VDI (German Association of Engineers) with a request for financial support for investigating the temperature conditions in the uni-flow cylinder. This request was granted. At about the same time, the government of the German state of Saxony provided considerable funds for the completion of a testing plant. The uni-flow cylinder used for this purpose in the engineering laboratory of the Institute of Technology at Dresden was built by the Maschinenfabrik Augsburg Numberg (MAN) and took the place of the lowpressure cylinder of the existing triple-expansion engine. It was put into operation in September 1911.44 The greatest credit for the commercial introduction of the uni-flow engine

22

Technology Transfer between Germany and the US, 1884-1930

is due to Sulzer Brothers of Winterthur, Switzerland, and Ludwigshafen, Germany. Extensive experiments enabled Sulzer Brothers to find the proper alloys for cylinder and piston-castings, which ensured reliable operation of these parts without the use of a tailrod.45 The Sulzer firm also played a prominent role in the transfer of the Stumpf uni-flow engine from Germany (and Switzerland) to the United States. As is well known, Sulzer was engaged in the transfer of Diesel engines and had close connections with the Nordberg Manufacturing Company of Milwaukee, Wisconsin.46 The American concern was the first to take up the manufacture of uni-flow engines with the constructional features proposed by Stumpf, combined with changes resulting from their own practice.47 There was another route of transfer. In the Technologist, the journal of the National Association of German-American Technologists, A.D. Skinner, Vice-President of the Skinner Engine Co. of Erie, Pa., had read about the uni-flow engine. He not only made enquiries about it in the United States, but went straight to the country of origin, Germany. In his report, he praised power plant engineering and efficiency in Europe and especially in Germany.48 This superiority, he maintained, was not the result of there being better engineering brains in Europe than in America, but because fuel was generally much more expensive in Europe. Therefore, the pursuit of high economy rather than low first cost was the factor that ultimately decided the selection of prime movers. Skinner's observations in Europe were correct and it can generally be said that the availability of cheap fuel and raw material was a decisive factor in choosing a particular machine technology.49 It is only logical that the uni-flow engine never played such an important role in America as in Europe, although its rate of diffusion was remarkable in the United States too. On his journey in Europe, Skinner visited the most important manufacturers of the uni-flow engine. He went to Sulzer in Ludwigshafen, where everything in connection with the uni-flow engine was willingly shown to him, an attitude not typical of German machinery firms at that time. He also went to the plant of Ehrhard & Sehmer at Saarbriicken, where a uniflow engine of 6,000 hp nominal capacity was being built to be installed in the Rochling Iron and Steel Works at Volklingen.50 Back in the United States, Skinner did something which was quite typical: he went to his plant and, according to drawings, descriptions and impressions he had gained in Germany, built a Stumpf uni-flow engine without a licence. This engine, which seems to have worked well, was sold to the A. Schreiber Brewing Company of Buffalo. Of course, the American representatives of Stumpf brought a law suit for infringement of Stumpf's patent, which they won.51 Another path of transfer of the uni-flow engine was provided by W. Trinks, a German-American mechanical engineer and metallurgist and Professor at the Carnegie Institute. In 1914, he gave a 'Technical Report of his last trip to Europe' to the Pittsburgh branch of the National Association of German-American Technologists. Besides other machinery plants, he had also visited the engineering firm of Erhardt and Sehmer. He gave photographs, drawings and some details of the performance of the engine to

Hans-Joachim Braun

23

the Pittsburgh engineering firm Mesta Machine Co., who then obtained a licence for building Stumpf steam engines in the United States.52 After the end of the First World War, Stumpf uni-flow engines were not only used for stationary purposes as they had been, almost exclusively, earlier, for example for driving compressors, pumps etc., but also as marine engines and for driving steam cars. Stumpf himself pointed out that in cases where the output of high-pressure compressors must vary, a directly-connected steam engine was often preferable to a gas engine or to geared turbine drives. At that time neither the internal combustion engine, in this case mainly the Diesel engine, nor the steam engine was susceptible to any great speed change without sacrificing efficiency.53 Stumpf convinced members of the National Association of German-American Technologists who worked for the Worthington Pump and Machinery Corporation in New York to obtain a licence to build uni-flow engines.54 By the end of the First World War, numerous American firms had licences to build Stumpf engines. However, Stumpf did not seem to be satisfied with this. Realizing that the prospects for his engine in the United States were good, he established his own company, the Stumpf Una Flow Steam Engine Company, in Syracuse, New York, in 1920.55 This process is quite common nowadays, but different from the normal cases of technology transfer in the eighteenth and nineteenth centuries. In those days the prospective donor of a new technology often moved to the foreign country to manufacture his products there, if they were not imported from the country of origin. Later, if often happened that firms in the country to which it had been transferred, adopted the new technology. In the case of the uni-flow engine, production in a foreign country by the donor took place at the end of an extended process of technology transfer. After Stumpf had realized that uni-flow engines produced by American firms were in high demand, he decided to found his own company in the United States. The next case study deals with coke-ovens, the transfer of coke-oven technology from Germany to the United States and the role Friedrich Schniewind, a chemical engineer and member of the National Association of German-American Technologists, played in this process. In the second half of the nineteenth century, 'beehive ovens' were almost exclusively used in the United States. They were small and relatively cheap to build, but had several disadvantages. One oven required two or three days to produce five or six tons of coke, and this coke lacked the quality and uniformity necessary for the best blast-furnace performance. Beehive ovens were also wasteful because in their operations valuable chemicals were burned or wasted away.56 It was soon realized in Europe, where the development of by-product coke-ovens progressed quickly, that North America provided a lucrative market for these ovens. In 1892, the erection of twelve by-product ovens, the first to be introduced into the United States, was begun at Syracuse by the Semet-Solvay Company, a subsidiary of the Belgian Semet-Solvay Process Company. More important for this paper was the attempt from 1886 onwards of the German Dr C. Otto Company of Dahlhausen, near Bochum,

24

Technology Transfer between Germany and the US, 1884-1930

to gain entrance into the American coke-oven market. This company was very successful with the vertical flue Otto-Hoffmann oven. In this oven the regenerative principle, which Friedrich and William Siemens had used in the blast furnace, was applied to coke-oven operation, and patented in 1883. The patent was bought by the firm of Dr C. Otto and the ovens built according to them gained popularity in Germany. By 1894 more than 1,200 of them had been built on the continent.57 Dr Friedrich Schniewind, a chemical engineer working for Otto, was commissioned to investigate the possibilities of establishing an Otto subsidiary in the United States. He left for the United States in 1889 and soon became a member of the National Association of German-American Technologists in New York and was one of its leading chemical engineers. It was due to his initiative that the Association concerned itself with the problem of coking and with the help of its journal, The Technologist, American chemical engineers were informed about the latest developments of cokeoven technology in Germany.58 The Cambria Steel Company ofJohnstown, Pa. became interested in the Otto-Hoffmann system of carbonization and in 1893 sent John Fulton, a prominent mining and chemical engineer, to Germany to find out whether the Connelsville coal might be successfully coked in a Otto-Hoffmann plant. The results were excellent. As a matter of fact, this coal was much better suited for coking than the coal generally available in Germany because the moisture content (1.3 per cent) was lower.59 This was a case in which technology transfer was enhanced by the more favourable raw material situation in the recipient country. A contract was signed between the Cambria Steel Company and the firm of Dr C. Otto for the construction of two batteries of thirty ovens each, together with a full by-product recovery system. To erect this plant and to build other plants in the United States, the Otto Coke Chemical Company was founded in Pittsburgh, Pa. in 1894 with Schniewind as Superintendent. The Otto Company was, however, reluctant to invest in the United States and preferred to give out a licence for 108.000 marks to the Otto Coke and Chemical Company, a company independent of the German Otto plant. This company was afterwards changed into the United Coke and Gas Company, New York, and after that into the Schniewind Coke Oven Company.60 In the period 1894 to 1905 the company, which controlled the German Otto patents, built 2,780 Otto ovens in the United States and Canada. In the beginning, Schniewind built the Otto-Hoffmann oven, but afterwards developed his own model, which was a modification of the Otto-Hoffmann oven. It was designed to allow accurate and uniform distribution of fuel gas to the combustion chambers and had the further advantage that the regenerators were built independent of the oven structure and could not therefore affect the latter by expansion. Although the construction of Schniewind's coke-oven was quite successful from the technical point of view, the management of the firm seems to have been poor. Besides, in the beginning it suffered from a chronic shortage of capital, but this was soon remedied as orders for coke-ovens became frequent.61

Hans-Joachim Braun

25

The German firm of Dr C. Otto employed an able engineer, Heinrich Koppers, who got into trouble with the Otto firm over patent rights. This was the reason why he left Otto in 1891 and founded his own firm, which built regenerative by-product coke-ovens. In 1903, Koppers went to the United States in order to examine the coke-ovens built there by the United States Coke & Gas Company and by Semet-Solvay. What he found out encouraged him to try to sell his own products there. The Otto-Hoffmann and Semet-Solvay ovens he saw were in a bad, neglected state, which was probably due to the fact that there was, at that time, not much competition in this field, so that the firms involved were not obliged to take much care of their equipment and look for the most efficient working processes.62 Koppers established contact with the editor of Iron Age, who introduced him to some engineers of the 'United Steel Trust'. Koppers praised the coke-ovens which he had manufactured in Germany, described them as superior to anything to be found in North America and gave the impression that in Germany all newly-erected ovens were exclusively built according to his prescriptions. He supported his claims by an impressive reference list.63 So in 1906 the United Steel Corporation sent a commission to Europe to study the coke-oven situation. The American Commission was pleased with what it saw and wrote a report praising especially the Koppers' ovens.64 In 1907 the Illinois Steel Corporation sent a second delegation to France, Belgium and Germany. This visit resulted in an invitation to Koppers to come to America and build a coke plant of 280 ovens at the Illinois Steel Company's factory at Joliet, Illinois. The report on the Otto-Hoffmann ovens in Germany had been rather favourable and Illinois Steel approached Schniewind with the question whether he would be prepared to build these ovens for them in the United States. This would have been probably cheaper for Illinois Steel. Schniewind, however, did not want to build Otto-Hoffmann ovens and recommended his own system. The commission proposed some modifications of the Schniewind process, but Schniewind declined their proposal. So Illinois Steel invited Koppers, an invitation which was also due to the endeavours of Louis Wilputte, a coke-oven engineer from Wales, who had heard of Koppers' interest in the United States, visited him in Essen and established contacts with Illinois Steel in order to convince this company of the superiority of Koppers' ovens. He later became Koppers' general manager in the United States.65 In 1907, Koppers came to America. In Germany he had been used to using fire-clay bricks (chamotte stones) for building his coke-ovens, but this was not possible in the United States because chamotte was available only in very small quantities and therefore expensive. It would have been too costly to import chamotte stones from Europe. So Koppers had no other choice than to use silica, which was abundant in the United States and had already been used by Schniewind. Koppers' disappointment at not being able to apply the material he was accustomed to working with—a material to which, he thought, he owed a large part of the success of his ovens— soon changed into delight. To his great surprise his research on the structure and

26

Technology Transfer between Germany and the US, 1884-1930

rigidity of silica stones showed that they, when properly used, were even superior to the chamotte stones. They were resistant to higher temperatures and pressures and also to erosion caused by salts. There was, however, a disadvantage, but Koppers managed to overcome this. When heated the silica stones expanded by about 1.4 per cent, and this could cause disintegration of the coke-oven. By an ingenious device—joints in which the stones could expand—Koppers succeeded in solving this problem. Silica stones were later also introduced into Germany.66 So, in the case of the Koppers' coke-ovens, we have an interesting case of technology transfer. A German engineer comes to the United States to introduce a superior technology. Because of the raw material situation in the country, he is obliged to apply a different process which, in the beginning, he considers to be inferior to the one he had used in Germany. As it happens, however, the new process proves to be even superior to the one in use in Germany and is transferred to that country. His success in the United States is enhanced by the fact that the coal normally used for coking is better than the one in Germany. After the successful transfer of his coke-oven process to the United States, Koppers founded a firm in Joliet, Illinois, and later, in 1912, in Chicago. The Chicago firm, in which German and American engineers worked together, was founded with a capital of 500,000 dollars. With the outbreak of the First World War, Koppers sold his company to Andrew Mellon for 300,000 dollars. With this sum he founded a new company in Pittsburgh. During the war, Koppers put into operation an average of one complete coke-oven plant every two months. The coke-oven capacity of the United States was doubled during the war years and Koppers constructed most of it.67 Already, during the war, the company used a strategy of diversification and integration. They bought gas works in New York, Chicago, Boston and Philadelphia. In 1917, after the United States' entry into the war, Koppers' shares, which amounted to 20 per cent of all shares of Koppers, Pittsburgh, were confiscated and later bought by Andrew Mellon. After the war, however, the German Koppers firm and American Koppers, which was then a completely independent company, continued to exchange patents and transfer technological know-how by other means.68 Members of the National Association of German-American Technologists entered the scene again when Koppers bought the Baltimore Machinery Company of Bartlett & Hayward. After having acquired various gas works and bituminous coal mines in Kentucky, West Virginia and Pennsylvania—large amounts of bituminous coal were needed to feed all the cokeovens Koppers built—and after he had bought railways to transport the coal, it is not surprising that Koppers found the idea of a metal fabricating and machinery concern attractive. Many of the leading engineers of Bartlett & Hayward were GermanAmerican members of the National Association of German-American Technologists.69 These engineers kept themselves informed about the latest developments in building gasholders in Germany and discussed them at their meetings.70 By the turn of the century, gasholders were in high demand

Hans-Joachim Braun

27

because gas lighting played an important role in the cities. Therefore gasholders were needed which, especially in big cities, were quite sizeable. These were either single lifts, i.e., simple bells inverted in tanks of water, or were constructed on the telescopic principle, in which case much ground space was saved as holders of much greater capacity could be contained in the same-sized tank. The tank for the gasholder was usually made by excavating a circular reservoir somewhat larger in diameter than the proposed holder. The tank had to be water-tight and was normally lined with concrete; the holder was made of sheet iron riveted together.71 These gasholders were to be found both in the United States and in Germany. They had their disadvantages, however, because with the large amounts of water needed for sealing, the gasholders had to be large and used a lot of iron for their construction. Therefore, in 1910, the engineer Konrad Jagschitz of the MAN developed his waterless gasholders. This was a rigid steel shell constructed with a regular polyground or cross-section of 10 to 28 sides according to its capacity. It was covered with a ventilated roof for protection against the weather and for structural rigidity. The great advantage of this gasholder was that, for a size of 250,000 cbm contents, only about 100 tons of water were necessary to seal the gas, as opposed to about 65,000 tons of water necessary for the traditional telescopic gasholder.72 Although a few of these gasholders had been built during the First World War, production really picked up only after the end of the war. The Maschinenfabrik Augsburg-Nurnberg, which produced the waterless holder at its Gustavsburg plants near Mayence in Hesse, was interested in giving licences to foreign plants in England, France and Holland. In this context, the United States played an especially important role. In 1923, Vice-President Hockley and Chief Engineer Wagner of Bartlett & Hayward Co. in Baltimore visited MAN's Gustavsburg plant. They had heard from GermanAmerican employees of Bartlett & Hayward, who were members of the National Association of German-American Technologists, of the improved MAN gasholder.73 Also, the engineer Boyer of Humphreys and Glasgow Ltd., London, a firm which produced waterless gasholders, had told them about the innovation.74 The licence negotiation between MAN and the representatives of Bartlett & Hayward were unproblematic. Hockley and Wagner were obviously very keen on obtaining a licence and the agreement was signed in October 1923. MAN managed to get a sizeable royalty for every waterless gasholder produced and sold by Bartlett & Hayward.75 It is not surprising that the Bartlett & Hayward representatives were keen on obtaining a licence. The waterless gasholder could be mass-produced and could also be built in very large sizes because most of the sealing water was superfluous. Therefore it was very suitable for marketing in the United States. Konrad Jagschitz, inventor of the waterless gasholder, went to Baltimore to advise Bartlett & Hayward with the first construction of the gasholder.76 Bartlett & Hayward wasted no time in manufacturing it. By the end of 1924 they had already produced seven gasholders, the two largest ones of 425,000 cbm. For these seven, MAN received royalties of about 800,000

28

Technology Transfer between Germany and the US, 1884-1930

marks, so they were very satisfied with the deal. Already in 1924, Bartlett & Hayward approached MAN with the wish to prolong the licensing agreement, which had originally been valid for ten years.77 Six months later it turned out, however, that the prospects of the waterless gasholders in the United States were not all that good. As to the very big sizes, there was not much competition, but the market for small- and medium-sized gasholders was very tight. Besides, there had been some difficulties with the gasholder, because gas leaked out of some of them. So, in June 1925, two representatives of Bartlett & Hayward visited Gustavsburg to negotiate a reduction of the MAN royalties. In the end, MAN agreed. Although relations between the two firms were generally good, Bartlett & Hayward had one recurring complaint to make about the German company: in the licensing agreement, MAN agreed to inform Bartlett & Hayward of all improvements to the gasholder which they made. This was also extended to any difficulties which the German firm might experience with gasholders manufactured by it. It seems, however, that MAN deliberately held back news of both kinds. Bartlett & Hayward gained this impression from other German sources.78 Here we have an example of behaviour of a German firm which often occurred in cases of licensing agreements: to keep the recipient firm dependent as long as possible on the technology of the donor firm, the latest developments of this technology were often held back. In this particular case, however, MAN was not successful in its strategy, because Bartlett & Hayward undertook research of its own. But this will be shown later. As mentioned before, there was much competition in the market for gasholders in the United States. Around 1926 the German firm of Kloenne from Dortmund in the Ruhr tried to enter the American market. MAN began a patent law suit against Kloenne, because their managers believed that Kloenne had infringed their patent on waterless gasholders. However, Kloenne managed to install a few improvements in the gasholder and, more important, tried to seal the gas by means of oil, reducing the amount of water needed to a great extent. The Stacey Waterless Gas Holder Company in Syracuse, New York, obtained a licence from Kloenne to manufacture their gasholder in the United States. Bartlett & Hayward asked the Western Gas Construction Company, Fort Wayne, Indiana, with which it had business contacts, to ask Kloenne about the conditions for obtaining a licence. A representative of Western Gas went to Dortmund and as a result Kloenne contacted Bartlett & Hayward. The latter firm informed MAN of its contacts with Kloenne and MAN did its best to convince Bartlett & Hayward that it was not worthwhile to change over to the Kloenne system of waterless gasholders. After that Bartlett & Hayward broke off negotiations with Kloenne.79 The connection between Bartlett & Hayward and Kloenne was a warning for MAN. After this, MAN was eager to disclose all its latest technological improvements of the waterless gasholder in order not to loose Bartlett & Hayward as a licensee. MAN made experiments with a new type, which resembled the old telescopic gasholder. The problem, however, was the water sealing which could, at that time, not be avoided when using the

Hans-Joachim Braun

29

telescopic gasholder. Although MAN worked hard on this problem, there was no immediate solution. In June 1931, Bartlett & Hayward contacted MAN and informed it of the possibility of obtaining a contract to erect a giant 10,000,000 cubic feet gasholder in Philadelphia. There was, however, severe competition from Kloenne's licensee Stacey, who owned a Waterless Gas Holder Company in Syracuse, New York. In order to obtain the contract, Bartlett & Hayward told MAN, it was necessary to underbid Stacey. This could, however, only be done if MAN did without their royalty. This was, undoubtedly, a shrewd move on Bartlett & Hayward's part, because this company knew that MAN would do their best not to let Kloenne, their chief rival in Germany, receive royalties from this contract. So MAN agreed. However, it was stipulated that the gain from the erection should be equally divided between Bartlett & Hayward and MAN. Bartlett & Hayward finally got the contract. It was quite in MAN's interest that in 1931 Bartlett & Hayward and the Stacey Company started negotiations in order to settle peacefully the patent quarrels between the MAN and the Kloenne systems in the United States. Another purpose of these negotiations was to bar other gasholders producers from entry into the American market.80 However, during the next year, 1932, Bartlett & Hayward rapidly acquired the dominant position in gasholder construction in the United States. When building the 10,000,000 cubic feet capacity gasholder in Philadelphia and even one of 15,000,000 cubic feet capacity, a gasholder that surpassed anything previously built in the world, the company redesigned the original MAN method of construction and developed special machinery to fabricate the shapes and plates for these giants. One of the machine tools created by Bartlett & Hayward was a special 'bending brake' to fabricate the shell plates, the piston plates and the roof plates. In 1932, the company succeeded in materially improving the construction and efficiency of the waterless holder, changing its shell from a polygon shape to a circular one. This modernized design was called the 'New Bartlett & Hayward Co. Waterless Gas Holder'. 81 This gasholder was later adopted on the continent.82 So, in this case, a German innovation was improved in the United States and, in this improved form re transferred to Germany. Although the precise weight of the National Association of GermanAmerican Technologists in technological transfer between Germany and the United States is difficult to determine, some general conclusions can be drawn. There were different routes of technological transfer. A member of the Association, an engineer, industrialist, or both, could learn about an advanced technology in Germany, introduce it into the United States and use the Association as a promotion centre. This happened, for example, in the case of refrigeration and Fred W. Wolf. Also, members of the Association, like the Pittsburgh Professor W. Trinks, visited Germany and especially German factories regularly, gave talks to the Association, thereby informing its members about the latest technological developments and sometimes inducing them to transfer technology from Germany. Engineers like Trinks also got American firms with no special connection with the Association

30

Technology Transfer between Germany and the US, 1884-1930

interested in transferring technology from Germany. Apart from this, the Association's journal Der Techniker and later The Technologist was used by German-American engineers and entrepreneurs to promote products originating in Germany in the United States. Cases in point are the Korting injector and the Riedler-Stumpf steam turbine. As already stated at the beginning, 'personal transfer' played the dominant role. This could be shown in almost all cases studied, especially in the cases of refrigeration, steam pumps, uni-flow engines and coke ovens. (German)-American engineers had to examine the new technology used in Germany in situ. It was normally not sufficient to use descriptions and drawing. Adaptation to American conditions was important. This sometimes meant more simplified constructions, which often did not last as long as the ones used in Germany but, also, larger constructions and, if possible, those which could be mass-produced. Apart from adaptation, some American firms also improved German technological innovations. This sometimes even resulted in a retransfer of an improved technology to Germany. Licensing was most common. Sometimes the donor firm did not disclose everything relating to the innovation in order to retain a technological advantage. Although fuel costs were not as high as in Germany, American engineers and companies did have an eye on them as was shown by the case of the Stumpf uni-flow steam engine. This was natural, as both individual engineers and companies are, or always should be, concerned with efficiency and the attempt to bring down the total cost, of which fuel cost is a part. Notes This paper is a revised and enlarged contribution to the Conference of the Society for the History of Technology in Toronto, October 1980. Previous versions were given at the University of Delaware, Massachusetts Institute of Technology and the University of British Columbia at Vancouver in March and April 1980 and at Princeton University and McGill University Montreal in October 1980. The author is grateful to numerous American, Canadian and German colleagues for helpful comments and wishes to especially mention Professors John Beer, George Basalla, Eugene Ferguson and David Hounshell of Delaware, Professor Merritt Roe Smith of MIT, Professors Robert Allen and Donald G. Paterson of UBC and Mr Edmund N. Todd, University of Pennsylvania. Thanks are also due to the Stiftung Volkswagenwerk, which supported me with a research and travel grant, without which the present work would have been impossible. I am also grateful to Mr Werkmann, Archivist of MAN, Niirnberg, who was of great help during my visit to the Nuremberg archives. 1. For the years until 1976 there is an exhaustive bibliography by Taghi Shaghati-Nejad and Robert Belfield, Transnational Corporations, Technology Transfer and Development, Worldwide Institutions Research Group, The Wharton School, University of Pennsylvania, Philadelphia, 1976. As the title says this bibliography concentrates on multinational corporations. Recent books on technology transfer include Jack Baranson, Technology and the Multinationals. Corporate Strategies in a Changing World Economy, Lexington u. Toronto, 1978; Jack N. Behrmann and Harvey W. Wallender, Transfers of Manufacturing Technology within Multinational Enterprise, Cambridge, Mass., 1976.

Hans-Joachim Braun

31

2. Historical research has concentrated mainly on technology transfer from England to Continental Europe in the period of early industrialization and from England to the United States in the first half of the nineteenth century. Some recent relevant publications dealing with the United States are Brooke Hindle, 'The Transfer of Power and Metallurgical Technologies to the United States, 1800-1880', in Centre National de la Recherche Scientifique: Vaquisition des Techniques par les Pays Non Initiateurs, Paris, 1973, pp. 407-28; Darwin H. Stapleton, The Transfer of Technology to the United States in the Nineteenth Century, University of Delaware Ph.D. thesis, 1975, in which a good bibliography can be found; and Nathan Rosenberg and Walter G. Vincenti, The Britannia Bridge: The Generation and Diffusion of Technological Knowledge, Cambridge, Mass. and London, 1978. Some articles in Nathan Rosenberg, Perspectives on Technology, Cambridge, 1976 are also relevant. 3. Mira Wilkins, The Role of Private Business in the International Diffusion of Technology, in Journal of Economic History, Vol. 34, 1974, pp. 166-88. There are two relevant books by the same author, The Emergence of Multinational Enterprise: American Business Abroad from the Colonial Era to 1914, Cambridge, Mass., 1970, and The Maturing of Multinational Enterprise. American Business Abroad from 1914 to 1970, Cambridge, Mass., 1974. Lawrence G. Franko has tried to look the other way round: 'The Origins of Multinational Manufacturing by Continental European Firms', in Business History Review, Vol. 48, 1974, pp. 279-302, and The European Multinationals. A Renewed Challenge to American and British Big Business, Stanford, Conn., 1976. There are some remarks on historical developments in Edward M. Graham, Oligopolistic Imitation and European Direct Investment in the United States, Harvard DBA Theses, 1974. 4. Mira Wilkins, 'The Role of Private Business in the International Diffusion of Technology Transfer' in Journal of Economic History, Vol. 34, 1974, pp. 166-88. 5. Everett M. Rogers and F. Floyd Shoemaker, Diffusion of Innovations, London and New York, 1971.* 6. For example, some of the works of Douglas North. On the historical literature on technology transfer, see Darwin H. Stapleton, The Transfer of Technology to the United States in the Nineteenth Century, University of Delaware Ph.D. Thesis, 1975; also the works of Nathan Rosenberg cited in Note 2. A competent historical study on technology transfer is David J. Jeremy, Transatlantic Industrial Revolution. The Diffusion of Textile Technologies between Britain and America, 1790-1830s, Cambridge, Mass. and Oxford, 1981. 7. I have just completed an article on 'A Technological Community in the United States: The National Association of German-American Technologists 18841941'. This paper deals mainly with the organizational history of the Association of which, so far, nothing is known in the literature, and with its members' views on political, social and economic problems. 8. 'Deutsche und amerikanische Industrieverhaltnisse. Deutscher und amerikanischer Kraftmaschinenbau' in Die Turbine, Vol. 1, 1904/5, p. 220. 9. Jiirgen Kocka, 'The Rise of Modern Industrial Enterprise in Germany' in Alfred D. Chandler Jr. and Herman Daems (eds), Managerial Hierarchies. Comparative Perspectives on the Rise of the Modern Industrial Enterprise, Cambridge, Mass. and London 1980, p. 104. 10. Cf. Franko, 1974, p. 285. 11. This point is particularly elaborated in H.J. Habakkuk, American and British Technology in the 19th Century. The Search for Labour-Saving Inventions, Cambridge, 1962 and the—sometimes controversial—literature following this publication. 12. There are some classificatory notes in Mira Wilkins, 'The Role of Private Business ...' See also S. H. Hymer, The International Operations of National Firms: A

32

Technology Transfer between Germany and the US, 1884-1930

Study of Direct Foreign Investment, Cambridge, Mass., 1976, p. 33. Also Jack Baranson, 'Technology Transfer through the International Firm' in American Economic Review, Papers and Proceedings, Vol. 60, 1970, pp. 435-40, and Titus O. Adeboye, International Transfer of Technology: A Comparative Study of Differences in Innovative Behavior, Harvard DBA Thesis, 1977. There is a helpful article in German: Peter Hertner, 'Fallstudien zu deutschen multinationalen Unternehmen vor dem ersten Weltkrieg' in Norbert Horn und Jiirgen Kocka, Recht und Entwicklung der Grofiunternehmen im 19. undfruhen 20. Jahrhundert, Gottingen, 1979, pp. 577-89. 13. A case in point, the Stumpf uni-flow steam engine, will be treated below. 14. This point is stressed by Stapleton, as note 6, p. 21. 15. Mitteilungen des deutsch-amerikanischen Techniker-Verbandes, Vol. 1, 1885-6,

16. W. R. Woolrich, The Man who created Cold. A History of Refrigeration, New York, 1967, p. 169. 17. The Technologist, Vol. 17, 1912, p. 52f. (in memoriam Fred W. Wolf), Woolrich, pp. 28, 121 f. As a German-American, Wolf had, of course, various contacts with German-American brewers in the United States. 18. Carl Linde, Aus meinem Leben und von meiner Arbeit, repr. of the 1916 edition, Miinchen, 1979, p. 62. 19. The Technologist, Vol. 17, 1912, p. 53. 20. Woolrich, as note 16, p. 90. 21. There are some interesting trade catalogues published by the De La Vergne Machine Company, especially De La Vergne Oil Engine Type CFH3 166 (1917); Hornsby-Ackroyd Oil Engines (1905); Koerting Four-Cycle Gas Engine (1906); Refrigerating and Ice Making Machinery (1908). 22. Mitteilungen des deutsch-amerikanischen Techniker-Verbandes, 1885-95, p. 1. 23. Ibid., p. 13. 24. There is a detailed history of pumps in Arthur M. Greene Jr., Pumping Machinery. A Treatise on the History, Design, Construction and Operation of Various Forms o Pumps, New York, 1919. For Worthington pumps, see p. 55 ff. 25. Worthington Pump and Machinery Corporation 100 years, 1840-1940, Harrison, NJ, 1940, is a short business history which should be updated. The business history aspect is also treated in M.J. Fields, 'The International Steam Pump Company: An Episode in American Corporate History' in Journal of Economic and Business History, Vol. 4, 1932, pp. 637-46. See also Clarence E. Searle, Henry R. Worthington (18171880) and his Influence upon American Industry, New York, 1947. Much technical information can be found in Henry R. Worthington, Worthington Steam Pumping Engine, History of its Invention and Development, New York, 1876, and by the same author, The Worthington High Duty Pumping Engine at UExposition Universelle de 1889. Relevant trade catalogues are Worthington Pumping Engines, Steam Pumps and Hydraulic Machinery (1881, 1893); The Worthington High Duty Pumping Engine (1889, 1893); Worthington Pumping Machinery (1900). For general technical literature dealing with pumps at the turn of the twentieth century, see Philip R. Bjorling, Pumps: Their History and Construction, Manchester, 1892; Henry Robinson, Hydraulic Power and Hydraulic Machinery, London, 1893; Julius Weisbach, The Mechanics of Pumping Machinery, London and New York, 1897; Irving P. Church, Hydraulic Motors, New York^ 1905; Charles A. Hague, Pumping Enginesfor Water Works, New York, 1907 and Heinrich Berg, Die Kolbenpumpen, enischliefilich der Fliigel- und Rotationspumpen, Berlin, 1921. 26. Worthington pumps were, especially on the Continent, notorious for consuming much fuel. This was stressed by Franz Reuleaux, Vber Neuerungen an Dampfpumpen und Dampfpumpenwerken, 1886. For further information, see Hans-Joachim Braun,

Hans-Joachim Braun

33

'Wissenschafts- und Technologietransfer zwischen Deutschland und Nordamerika am Ausgang des 19. Jahrhunderts. Das Beispiel Franz Reuleaux' in Technikgeschichte, Vol.48, 1981, pp. 112-30. 27. Praser & Chalmers Catalogue, Riedler Pumps, Chicago and Erith, Kent, 1900, p. 13ff;Alois Riedler, 'Neuere Wasserwerks- Pumpenmaschinen fur stadtische Wasserversorgung-Anlagen' (offprint from A. Riedler, Schnellbetrieb. Erhohung der Geschwindigkeit und Wirtschaftlichkeit der Maschinenbetriebe), Munich and Leipzig, 1900, p. 12f. 28. Fraser & Chalmers Catalogue, Riedler Pumps, p. 14. 29. A. Riedler, Neuere Wasserwerks-Pumpenmaschinen, 1900, p. 67. 30. A. Riedler, 'Express-Pumpen mit unmittelbarem elektrischen Antrieb' (offprint from Schnellbetrieb), Munich and Leipzig, 1900, passim; Greene, p. 165. 31. Otto Schone, 'Johannes Stumpf in Forschungen und Fortschritte, Vol. 13, No. 1 1937, p. 15f. 32. The Technologist, Vol. 2, 1897, p. 118. 33. Carl MatschoB, Die Entwicklung der Dampfmaschine, Vol. 2, Berlin, 1908, p. 617. The early Laval turbines were difficult to handle because of their high speed, which was impractical. On early steam turbines, see Wilhelm Gentsch, Steam Turbines, trans, from the German by Arthur R. Liddell, London and New York, 1906; Max Dietrich, Die gebrauchlichsten Dampfturbinensysteme, Rostock, 1906; Walter S. Leland, Steam Turbines, Chicago, 1917. A review of steam turbines in America is in Ernst A. Kraft, Amerikas Dampfturbinenbau, Berlin, 1927; of steam turbines in Germany, F. E. Junge and E. Heinrich, 'The Steam Turbine in Germany' in Engineering Magazine, Vol. 40, 1910/11, p. 439f., and AEG, Berlin, Steam Turbines, Berlin, 190723. 34. A. Riedler, 'Uber Dampfturbinen' in The Technologist, Vol. 9, 1904, p. 193. 35. 25 Jahre AEG Dampfturbinen, ed. by Allgemeine Elektrizitats-Gesellschaft, Berlin, 1928, p. 2. 36. The Curtis turbine was an impulse turbine which, however, in the division of the heat drop or pressure drop into stages followed Charles Parsons. 37. 25 Jahre AEG Dampfturbinen, 1928, p.*9. There is general information on the history of the steam turbine in Edgar T. Westbury, Turbines, Steam, Water and Gas, London, 1964. 38. Johannes Korting, Ernst Korting 1842-1921. Ein Ingenieur und Unternehmer im kaiserlichen Deutschland, Diisseldorf, 1975, p. 39ff.Also Korting 1871-1971, Hanover, 1971. 39. The First One Hundred Tears. Otis Elevator Company, 1953, p. 15ff;ParallelenKonsequenzen. Betrachtung einerfast 100-jdhrigen parallelen Entwicklung 1852-1951 und der 2 Jahrejungen Fusion 1951-1976, Berlin, 1976, passim; 50 Jahre deutscher Aufzugsbau. Eine Festschrift zum 50jahrigen Geschaftsjubilaum der Carl Flohr A.G.. Berlin, 1929, p. 10ff.On Otis, see L.A. Peterson, Elisha Graves Otis, 1811-1861, New York, 1945. There is an unpublished manuscript on the history of elevators in the Baker Library, Harvard Business School: David Shannon, The Annals of Vertical Transportation, 1953. 40. H. W. Dickinson, A Short History of the Steam Engine, London, 1965, p. 153 ff 41. B.V. Nordberg, 'The Steam Engine' in Power, Vol. 53, No. 1, 1921, p. 14. 42. 'The Universal Uniflow Engine, in Power, Vol. 40, No. 22, 1914, p. 777. 43. J. Stumpf, The Uni-Flow Steam Engine, Syracuse, New York, 1922, Preface. 44. Die Hundertjahrige Geschichte der ersten Brunner. Maschinen-Fabrikgesellschaft in Briinn von 1821 bis 1921, Leipzig, 1921, p. 168 f., ^eitschrift des Vereins Deutscher Ingenieure, 1913, pp. 1074-8. 45. J. Stumpf, as note 43, p. 150 ff.

34

Technology Transfer between Germany and the US, 1884-1930

46. W. Robert Nitske and Charles M. Wilson, Rudolf Diesel, Pioneer of the Age of Power, Norman, Oklahoma, 1965, pp. 253-7. 47. J. Stumpf, as note 43, p. 168. 48. A.D. Skinner, 'The Uniflow Engine in Europe,' in Power, Vol. 40, 1914, pp. 558-60. 49. As Habakkuk and others have pointed out, high labour costs were also of significance. 50. Skinner, as note 48, p. 559. 51. Power, 45, 1917, p. 570. 52. The Technologist, Vol. 17, 1912, p. 53. 53. J. Stumpf, 'Die Gleichstrom-Verbund-Schiffsdampfmaschine' in Werft, Reederei, Hafen, Vol. 8, 1927, pp. 245-9; J. Stumpf, 'A 1250 Horsepower Compound Marine Engine' in Power, Vol. 66, 1927, p. 509 f.; 'Uni Engine drives Steam Car', in Power, Vol. 45, 1917, p. 177 f.; J. Stumpf, 'Uniflow Engines drive High Pressure Compressors' in Power, Vol. 62, 1925, p. 940 f. 54. Cf. Stumpf, as note 43, p. 218. 55. Ibid., Preface. 56. Fred C. Foy, Ovens Chemicals and Men. Koppers Company Inc., New York, Newcomen-Society in North America, 1958, p. lOf. 57. Franz Michael Ress, Geschichte der deutschen Kokereitechnik, Essen, 1957, p. 589; C. S. Finney and John Mitchell, 'By-Product Ovens in the United States' in History of Iron and Steelmaking in the United States, The Metallurgical Society, American Institute of Mining, Metallurgical, and Petroleum Engineers, New York, 1961, p. 48. 58. H.A. Wasmuth, 'Entwicklung der Koks-Industrie in Verwerthung der Nebenproducte', in Der Techniker, Vol. 15, No. 4, 1893, pp. 61-4; Walter Stern, 'Fabrikation von Coke und Trocken-Destillation der Kohlen' in Der Techniker, Vol. 17, No. 1, 1898, pp. 6-8; The Technologist, Vol. 2, 1896, p. 215. 59. 'Auszug aus dem Bericht iiber amerikanische Koksofenverhaltnisse Ende September 1913 von G.E.Junius' in Otto-Archives, Bochum. 60. Ress, as note 57, p. 589. 61. Cf. Horace C. Porter, Coal Carbonization, New York, 1924, p. 211; F.M. Ress, 'Vorarbeiten zur Firmengeschichte' in Otto Archives, Bochum (CIII—6 d 5). The German-American Association of Technologists continued to deal with byproduct coke-ovens: R. W. Hilgenstock, 'Uber Koksofen mit Gewinnung der Nebenprodukte und Verwendung der gewonnenen Gase zum Betriebe von Gasmaschinen' in The Technologist, Vol. 8, 1903, pp. 73-82; Wm. Meyn, 'Koksofen mit Gewinnung der Nebenprodukte in Amerika' in The Technologist, Vol. 9, 1904, pp. 63-80. Other regional American engineering associations, too, discussed the merits and disadvantages of different by-product ovens: W.H. Blauvelt, 'The By-Product Coke Oven' in Journal of the Western Society of Engineers, Vol. 10, 1905, pp. 477-99; Edwin A. Moore, 'By-Product Coke Ovens in America: Past, Present, and Future' in Proceedings of the Engineers Club of Philadelphia, Vol. 23, 1906, pp. 151-73. 62. Koppers' work in Germany and the United States is described in Koppers, Ein halbes Jahrhundert im Dienste der Kohleveredelung, 1901-1951, Essen, 1951, and Ress, as note 57, p. 409 ff. 63. Ress, as note 57, p. 595. 64. Foy, as note 56, p. 10. 65. Koppers, Ein halbes Jahrhundert, p. 44; Ress, 1957, p. 448. 66. Cf. The Refractories Journal, June 1955, p. 266. Also H. Hock, Kokereiwesen, Dresden and Leipzig, 1930, p. 74.

Hans-Joachim Brawn

35

67. Foy, as note 56, p. 14. 68. Cf. Koppers Company, By-Product Coke and Gas Oven Plants, Benzol Recovery Plants; Motor Fuel Recovery Plants, Tar Distilling Plants; Ammonia Recovery Apparatuses Designed and Built by the Koppers Company, Pittsburgh, Pa., 1913; Koppers Company, Coke and Gas Ovens, Papers, 1911-44; Koppers, Ein halbes Jahrhundert, p. 56. 69. 'Mitglieder-Verzeichnis des deutsch-amerikanischen Techniker-Verbandes 1903-1904' in The Technologist, 1904, p. 16. 70. For example, Hubert Krekel, 'Gasbehalter von 10 Mio KubikfuB Inhalt' in The Technologist, Vol. 15, 1910, p. 172. 71. 'Gasholder' in Encyclopaedia Britannica, 11th edition, Vol. 11, 1910, p. 488. On gasholders and gastanks, see also George Wehrle, American Gas Works Practice, Standard Practical Methods in Gas Fitting, Distribution and Works Management, New York, 1919, pp. 168-70; Hubert Krekel, 'Der neue wasserlose Gasbehalter' in The Technologist, Vol. 31, 1926, pp. 71-7; Hermann Holler and W. Fink, Rohrleitungs-und Behalterbau, Berlin, 1932. 72. Ferdinand C. Latrobe, Iron Men and their Dogs, Baltimore, 1941, p. 134; FritzBiichner, HundertJahreGeschichte der MaschinenfabrikAugsburg-JViirnberg, 1840-1940, 1940, p. 215; 'lOOJahre MAN Werk Gustavsburg' in MAN Werkszeitung, May 1960, p. 18. 73. Wagner, although not a member of the Association himself, was in close connection with some of its members. In 1914 he gave a paper on coal gas and its by-products at the Baltimore branch of the Association (F. H. Wagner, 'Uber Kohlengas und seine Nebenprodukte' in The Technologist, Vol. 19, 1914, pp. 41-3). He was, besides his work at Bartlett & Hayward, a lecturer in thefieldof gas by-products at Johns Hopkins University in Baltimore and published a book Coal and Coke (New York, 1916). 74. MAN-Archives, Nurnberg, 132. 3.2/356. 75. MAN, 132.3.2.1 ('Wasserlose Gasbehalter, Gasbehalterlizenz Amerika', 5 Oct. 1923). 76. MAN, 132.3.2.1 (25 Feb. 1924). 77. MAN, 132.3.2.1 (2 Dec. 1924). 78. MAN, 132.3.2./356 (22 Jan. 1928; 'Reiseinstruktionen an Herrnjoerger'). 79. MAN, 132.3.2/356. 80. MAN, 132.3.2/356 (28 July 1931; 12 November 1931). 81. Latrobe, as note 72, p. 138f. 82. Cf. Fritz Wehrmann, Die Gasspeicherung, Miinchen, 1954.

E d i s o n

i n

t h e

M a g n e t i c

M o u n t a i n s :

O r e

V e n t u r e ,

t h e

S e p a r a t i o n

1 8 7 9 - 1 9 0 0

W. B E R N A R D C A R L S O N

'I shall make either a gigantic success or a gigantic failure.'1 —Thomas A. Edison on ore separation in 1894 By the age of thirty-five, Thomas A. Edison had produced the inventions for which he would be long remembered: the quadruplex telegraph, the phonograph and his system of incandescent lighting. Yet in his later years, he embarked on several equally interesting and ambitious projects. Edison dabbled with early motion-picture photography, developed an improved storage battery, promoted low-cost concrete houses and searched for synthetic rubber. However, the largest and most complex project that he undertook following the introduction of his electric lighting system was his magnetic ore separation venture. In the late 1880s and 1890s, Edison developed an industrial process which concentrated low-grade magnetic iron ore (magnetite) into compact briquettes. To put this process into commercial production, he constructed a huge plant at Ogden in the highlands of north-western New Jersey. Confident that the briquettes would be purchased by eastern blast furnaces, Edison invested over three million dollars in the plant and devoted ten years to perfecting his ore separation process.2 Edison embarked on his ore separation venture because he perceived a vast market for iron ore in the declining eastern iron and steel industry. During the 1880s the production of steel grew rapidly in America, which stimulated the development of the rich iron deposits of upper Michigan and Minnesota. Most of the ore from these new deposits was consumed by the blast furnaces of Pittsburgh, Illinois and Ohio; due to high transportation costs, it could not be used economically in the furnaces of eastern Pennsylvania, New York and New Jersey. Partly because of their proximity to these new sources of ore but also because they were technologically more advanced and better organized, the mid western iron and steel firms soon surpassed their eastern counterparts. Some eastern plants lingered through the 1890s by using ore imported from Cuba and Spain, but most were forced to shut down. Edison, along with several other inventors, responded to this situation by perfecting a magnetic ore separation process. Hoping to revitalize the eastern industry, Edison designed his plant and process to convert the low-

38

Edison: the Magnetic Ore Separation Venture, 1879-1900

grade ores found throughout the Appalachians into a marketable quantity that would be consumed by the eastern blast furnaces.3 Edison's perception of a potential market for his briquettes was an essential factor in the development and operation of the Ogden plant. Because of this, the ore separation venture is an ideal case study for examining how an inventor's perception of the technological and economic environment shapes his inventive style. In narrating the story of the ore separation project, this paper will describe how factors such as the availability of capital and Edison's penchant for efficiency and automation affected his perception and ultimately determined the viability of the enterprise. Edison first became interested in ore separation in early 1879 while working on his electric lighting system. Assuming that he was going to employ a platinum filament lamp in his system, he studied the available sources of platinum and experimented with extracting it from various ores. Intrigued by these experiments, Edison tried extracting gold, silver, lead and tin from different ore samples.4 In the course of this work, Edison invented a magnetic separator for iron ore (see Fig. 1). This separator consisted of a hopper suspended above an electromagnet and two bins placed below the magnet. Finely crushed magnetic iron ore flowed from a narrow slit in the bottom of the hopper past the electromagnet in a wide, fine stream. The iron particles were attracted by the magnet, drawn out of the vertical stream and fell into one bin. The non-magnetic waste or tailings fell into the other bin. Patented in 1880, this separator embodied the design principles upon which the separators at Ogden were later based.5 To exploit this invention and finance further experimentation, Edison and several investors from the Edison Electric Light Company organized the Edison Ore-Milling Company in 1879. Through this company, Edison set up two plants on the beaches of Long Island and Rhode Island to remove iron from black magnetic sand. Unexpectedly, the project suffered from storms which washed this sand in and out to sea and from difficulties in marketing the ore; consequently, in 1883, Edison dropped the project. Trying to process the sand, he later remarked, was like 'taking a mortgage out on a school of herring'.6 During the next several years, Edison was busy improving and marketing his electric lighting system, but in 1887 he returned to ore separation. His decision to move from the electric lighting field was probably motivated by a desire to work in a new and less competitive field. By 1887 the electric lighting field was becoming crowded with inventors and firms who were competing directly with Edison. Several companies, including United States Electric Lighting, challenged him with extended litigation over the incandescent lamp patents. Still others, especially George Westinghouse, competed with Edison's direct-current lighting system by introducing alternating-current systems. Rather than develop his own a.c. system, Edison instead fought the 'battle of the currents' from 1886 to 1888, in which he attempted to discredit alternating current by arguing that it had a greater potential for fatal electrocution than direct current. With this competition,

W. Bernard Carlson

39

i^i^^^^ilii

Figure 1. Edison's basic magnetic separator. Key: B—hopper holding crushed ore; b—slit in bottom of hopper; C—electromagnet; D—stream of crushed ore passing magnet; F—bin for tailings; F'—bin for concentrated ore. From US Patent 228, 329 (1 June 1880), Magnetic Ore Separator.

40

Edison: the Magnetic Ore Separation Venture, 1879-1900

Edison came to believe that he was losing his pre-eminence in the electric lighting field.7 In contrast to electric lighting, the ore separation field was relatively open and potentially lucrative. Having worked earlier on a separator, Edison was already familiar with the technology of the field and aware of the challenges that it offered. Like other inventors, he found that 'There is something very fascinating in the production of lean ore of concentrate in which only very few particles of foreign matter can be detected.'8 Furthermore, by 1887, Edison probably had noticed that the eastern iron and steel industry had stagnated, which led him to surmise that there was a potential market for iron ore in the east. In the early phases of the ore separation project, Edfton employed a method of invention and development similar to the one he had used in developing his electric lighting system.9 To finance his research, he reorganized the Edison Ore-Milling Company in 1887 and increased its capitalization from $350,000 to $2,000,000. At his new West Orange laboratory, he fitted up rooms for ore separation experiments, equipped with separators, crushers, grinders and assay furnaces.10 Edison complemented the knowledge of his laboratory assistants by consulting two experts from the iron industry—Walter S. Mallory, a Chicago iron manufacturer, and John Birkinbine, a prominent Philadelphia mining engineer.11 Just as he had collected data on the gas industry while developing his lighting system, so Edison acquired books on mining and minerals, government geological reports and maps. He also gathered an extensive collection of ore samples by shrewdly exploiting the newspaper coverage he received; he announced in an interview that he was stocking his new laboratory, sending out orders for 'All ores, metals, &c &c. Also everything from an Elephant's hide to the eyeballs of a United States Senator.' Following this announcement, Edison was soon inundated with numerous samples of metals and ores which people asked him to assay.12 With these resources, Edison and his staff improved the ore separator and tested it at experimental plants in Humboldt, Michigan and Betchelsville, Pennsylvania in 1889 and 1890. At these two plants they transformed the ore separator (Fig. 2) from a laboratory prototype to a small-scale commercial process.13 At the Humboldt plant, Walter Mallory ran large test batches through the separator and subsequently made adjustments and improvements in order to maximize the amount of iron recovered from different types of ore. Mallory also estimated the general costs of operating an ore separation plant.14 Edison, H. M. Livor and others worked at Betchelsville and developed the machinery needed to process and handle large amounts of finely crushed ore: crushers, grinders, screening apparatus and conveyor belts.15 Although neither of these plants reached a level of commercial production, the experience of constructing and operating them further convinced Edison that ore separation was technically feasible. To develop commercially his ore separation process, Edison needed accurate information about iron deposits of low-grade magnetic ore in New York, New Jersey and Pennsylvania. Travelling through the highlands of

W. Bernard Carlson

41

Figure 2. Blower separator, as illustrated in Iron Age, Vol. 41,6 Dec. 1888, p. 847. The pipe above the bar magnet is the air current outlet. these states, his surveyors mapped ranges of iron ores using a special dipping needle compass; other assistants collected samples at various local mines which they brought back to the laboratory and tested in the separator. Based on these surveys and tests, Edison eventually purchased or leased over 16,000 acres of low-grade deposits, including some of the largest in the east.16 Relying on the results of this survey and advice from Mallory and Birkinbine, Edison formulated a marketing strategy. He had noticed that with the development of the rich iron deposits of Michigan and Minnesota, the iron and steel industry was moving westward. As a result of this migration, numerous eastern blast furnaces were left idle, unable to compete with the midwestern plants. To Edison, these eastern furnaces could be revitalized if provided with an abundant supply of inexpensive iron ore. Using his ore separator to process extensive low-grade deposits, he believed that he had the ideal means of supplying the ore needed in the east. However, to produce

42

Edison: the Magnetic Ore Separation Venture, 1879-1900

the ore cheaply, it was necessary to process it on a large scale; for this reason, Edison designed his plant to have an enormous capacity and to utilize labour-saving and automated machinery. Edison assumed that by using the latest and most efficient equipment his plant would remain competitive even if iron prices dropped substantially. Furthermore, by locating his plant near many of the eastern blast furnaces, he calculated that his concentrate would have lower transportation costs than ores shipped from Michigan and Minnesota. While the average freight rate for ores from Minnesota to the eastern furnaces was $3.00 per ton, Edison estimated his concentrate could be shipped for $2.32 to most of his potential buyers. By following this strategy, he was confident that he could produce and ship concentrate competitive in price with Minnesota ores.17 For the most part, Edison's market strategy was similar to that of others intent on promoting ore separation; anticipating a need for ore by eastern furnaces which would be filled by utilizing nearby lean ores, inventors and mining engineers had installed by 1890 over twenty ore separators in small experimental plants. While most of these separators were used to process lean ores produced at mines in the course of mining rich veins, inventors at two northern New York mines set up larger plants for the express purpose of concentrating ore from large lean deposits.18 Yet, Edison differed from his contemporaries by being far more bold; while other inventors put up ore separation mills with a daily capacity of no more than 1,200 tons, Edison planned to process 5,000 tons. In constructing such a gigantic plant at Ogden in north-western New Jersey, he was a revolutionary since established mining wisdom previously dictated that maximum profits were to be made by mining rich veins of ore with a minimum amount of processing equipment.19 Undoubtedly, Edison was willing to take such chances because he strongly believed that a large and ready market existed for concentrated ore; in engineering jargon, he perceived a 'market pull' for his product. Supremely confident of his potential market, his process and his control of the key low-grade deposits, Edison stated: This venture has all the elements of permanent success—all the factors are known—The deposits cannot be duplicated Like Crops or paralleled like RR Neither is it dependent upon or subject to errors as in merchandizing or conditioned upon good management for success.20 Knowing that his plans required substantial capitalization, Edison courted outside investors. To attract attention to ore separation generally and to his process in particular, he had Birkinbine read several papers before the American Institute of Mining Engineers. Just as he had brought investors to Menlo Park to see his electric lights, so he brought capitalists to Bechtelsville to show them his ore separation process. Edison also demonstrated the ore separator in New York for the leading iron ore dealers.21 However, the investors and businessmen involved in iron mining tended to prefer exploiting rich sources of ore rather than develop efficient means of processing. Consequently, Edison found no outside supporters and was left to finance the Ogden plant using his own resources. In 1889, with a few

W. Bernard Carlson

43

close associates, he organized the New Jersey and Pennsylvania Concentrating Works for the purpose of constructing the Ogden plant.22 Independent of outside investment, Edison was effectively isolated from external control and advice. Outside investors would not necessarily have compelled Edison to construct a substantially different plant at Ogden, but their presence probably would have influenced his efforts. For instance, had he been dependent on outside capital and failed to secure it, Edison would have had to redesign his process or alter his marketing strategy in order to make his venture appealing to investors. Similarly, outsiders who did invest in his ore separation venture would have protected their investment by funding the project incrementally; this would have kept Edison from building a plant before he had perfected his ore separation process. External investment would also have prevented Edison from becoming isolated from expert advice. In obtaining capital he probably would have interacted with businessmen cognizant of trends in the iron market and with engineers familiar with the technical aspects of iron mining and smelting. Instead, after 1890, Edison ignored the mining engineers who were aghast at his confidence that he could solve the formidable problems involved in ore separation.23 Thus, because he relied on his own financial resources, Edison formulated his market strategy and built Ogden according to his own ideas. Curiously, the experts that Edison consulted in the early stages of the ore separation project had little effect on his perception of the industry. Birkinbine, probably the most knowledgeable consultant Edison had, was fired in 1890 just as the Ogden plant was getting under way. There is no indication that Mallory, though a close associate of Edison and an officer in the ore separation company, ever tried to shape the market strategy. His role appears to be strictly limited to technical matters. Pilling and Crane, iron ore dealers, advised Edison later on the problems involved in marketing the concentrate, but there is nothing to suggest that they questioned his basic assumptions about the iron and steel industry. Rather, they appear to have provided specific marketing information that only reinforced Edison's perception of the industry.24 Failing to secure outside capital, Edison was not bothered by having to finance his ore separation venture alone. He apparently liked having complete control over his projects, and this feeling was reinforced by his recent experiences with his electric lighting companies. As his firms for the manufacture, construction and operation of electrical systems grew in size, their operations were supervised more by his assistants and financial backers than by Edison himself. Moreover, with the consolidation of these firms as Edison General Electric in 1888 and as part of General Electric in 1892, Edison relinquished his remaining control of these enterprises. Believing that the financiers had unfairly removed these companies from his control, Edison had little desire to have outside investors involved with Ogden. To reporters, Edison adamantly stated that ore separation was the 'one thing I am now working on out of which I shall make money and out of which nobody can get any share except the boys here who own the thing with me'. 25

44

Edison: the Magnetic Ore Separation Venture, 1879-1900

Figure 3. Part of the Edison Concentrating Works at Ogdensburg, New Jersey. From 'The Edison Concentrating Works' in Iron Age, Vol. 59, 28 Oct. 1897, p. 4. Key: 1-Boiler and engine house; 2-Ore mill; 3-Stock house, concentrates; 4-Concentrates conveyor; 5-Mixing house; 6-Furnace house; 7-Briquetting house; 8-Stock house; 9-Crusher house; 10-Rock stock house; 11 -Conveyor to ore mill; 12-Machine shop; 13-Power station The location and design of the Ogden plant (Fig. 3) reflected Edison's marketing strategy. Located in the north-western corner of New Jersey, Ogden was within seventy-five miles of most eastern blast furnaces. Although most of the rich ore had been previously removed from the Ogden mine, the remaining low-grade iron constituted the largest deposit that his surveyors had found. The ore, which averaged 20 per cent magnetite, was distributed in a narrow strip 2 miles long and 200 yards wide. Edison took advantage of the geological form of the deposit by employing a set of travelling cranes and later two steam shovels (one of which was the largest in America) which moved along the seam, stripping away all the orebearing rock.26 The travelling cranes or shovels broke off 6 ton chunks of rock and loaded them onto a cableway or railroad skips which carried them several hundred

W. Bernard Carlson

45

Figure 4. Giant and intermediate rolls (from 'The Edison Magnetic Concentrating Works' in Scientific American, Vol. 78, 22 Jan. 1898, p. 54). yards to the main plant.27 At the plant, the chunks were unloaded by another crane and fed into the giant rolls (Fig. 4). The most prominent feature of the plant, these rolls were 6 feet in diameter, 5 feet long and weighed 35 tons each. The chunks fell between the two rolls and were crushed by the momentum of the heavy rolls. Special clutches protected the 700 horsepower steam engine driving the rolls by releasing them at the moment of crushing. Because the momentum of the rolls did the actual work, a smaller steam engine could be used than was normally employed with machinery of this scale.28

46

Edison: the Magnetic Ore Separation Venture, 1879-1900

From the giant rolls, the rock passed through three smaller sets of rolls which reduced it to pieces \ inch in size. An elevator then carried the crushed ore to the top of the 50 foot high dryer. Heated from below by an open-air furnace, the dryer contained an elaborate series of iron plates arranged alternately at 45 degree angles. The crushed ore was dried by the hot air that rose from the furnace as it fell from plate to plate. After drying, the ore was stored in a stockhouse, which provided a supply of ore for other sections of the plant to draw upon at their own rate, independent of the crushing mill.29 The second phase of the process, the separation phase, began with grinding the ore to 14 mesh or slightly smaller than ^ inch in maximum crosssection. This was done by the three high rolls specially designed by Edison. In these machines, the lowest roller was set in fixed bearings while the top two rollers moved up and down freely. The exact position of the rollers and the degree of fineness to which the ore was crushed was controlled by a high-pressure air cylinder. With this arrangement, Edison claimed that the rolls were substantially more power efficient than other commercially available grinders.30 From the three high rolls the finely crushed ore fell next through a series of screens which eliminated any pieces larger than 14 mesh. Then the ore was delivered to the separators proper. At Ogden, 480 electromagnets were employed in three groups. The ore first streamed past a battery of relatively weak magnets, which removed particles of high iron content. Stronger magnets, which attracted the less pure particles, followed. By regrinding and reseparating the tailings several times, additional iron was recovered and the iron content of the final concentrate was progressively increased. This staged separation process could capture about 95 per cent of the magnetite available in the ore, making the final concentrate 68 per cent pure iron.31 In addition to iron, however, the concentrate also contained phosphorus. Since traces of phosphorus make Bessemer-processed iron and steel brittle, the amount in a particular iron ore frequently determined its commercial value. Hence, one of the critical problems Edison encountered was how to lower the phosphorus content of the concentrate. Since the phosphorus compounds in the ore tended to be crushed smaller than the magnetite particles, Edison overcame this difficulty by using a gentle air current to blow the phosphorus away as the concentrate streamed through a special chamber. These phosphorus compounds were then recovered and sold for use in paints. Similarly, the other waste product of the separation process, the tailings, was marketed as high-grade sand to building contractors.32 During the first several years at Ogden, after the finely crushed concentrate passed through the separation mill, it was temporarily stored in a second stockhouse or shipped directly to blast furnaces. However, because blast furnace operators found the finely crushed ore material difficult to handle and process, Edison was forced to add a bricking section to the plant. In this phase of the operation, the finely crushed concentrate was mixed with a binder, moulded under pressure into a cylindrical briquette (1 inch

W. Bernard Carlson

47

high and 3 inches in diameter) and baked to a hard finish. The briquettes were stored in a third large stockhouse, located near a spur of the New Jersey Central Railroad. Edison estimated that it took 130 minutes to process an entire railroad car of ore, from crushing to bricking, with most of the time (110 minutes) consumed by the baking of the briquettes.33 To co-ordinate the many steps of this large plant (at its peak, Ogden could process 5,000 tons daily), Edison introduced several novel and ingenious features. Most notable were the large conveyor belts which connected different sections of the plant. With the help of a rubber-goods salesman, Thomas Robins, Edison developed a durable rubber-covered belt and enlarged its carrying capacity by putting angled rollers under either side of it. Since the conveyors ran on thousands of stationary bearings, he designed a special 'dust-proof bearing and installed a central lubricating system. From one main tank, oil was pumped continuously through underground pipes to every bearing and gearbox, thus eliminating the constant maintenance these components normally required. Another feature was the special safety bolts which Edison inserted at critical points in the large crushers and grinders. By shearing when a piece of machinery was unduly strained, these bolts protected the machine from being completely ruined. To synchronize the operations of the plant, telegraph wires and messengers linked the different buildings, which allowed different phases of the plant to be appropriately speeded up or slowed down and crews to be dispatched quickly to perform repairs. To monitor the efficiency of the process, Edison introduced a system of time accounting for each machine in the plant. At each machine, the operator noted the exact starting and stopping times as well as any reasons why the machine was not running. This data was subsequently collected and from it Edison and his staff identified bottlenecks in the plant. Gradually added to the Ogden plant, these features show that Edison worked hard to integrate the components of his ore separation process so that it functioned smoothly like a very large, well-oiled machine.34 In addition to integrating the components of his ore separation plant, Edison emphasized efficiency and automation in its design and operation. This was certainly in keeping with his market strategy; to remain competitive even if iron ore prices fell substantially, Edison assumed that laboursaving and automatic equipment was essential. Accordingly, he introduced features such as the conveyor belts, the gravity-fed dryers and screens, and the central lubrication system in order to minimize the labour, time and energy required to process the ore. When interviewed about Ogden (Fig. 5) Edison explained that he employed coal-powered shovels rather than dynamite because the energy provided by dynamite at $250 per ton was more expensive than the energy provided by coal at $2.00 per ton. He took pride in the fact that the three high rolls were 82 per cent efficient and in showing that the giant rolls were designed to consume only 700 horsepower. But most of all, Edison frequently emphasized that Ogden was completely automated, that in converting the raw ore into briquettes it was never touched by human hands. Every step, from mining to loading the railroad cars, was done by machines with his

48

Edison: the Magnetic Ore Separation Venture, 1879-1900

Figure 5. View of a portion of the Ogden plant looking north-east. On the left were the buildings containing the giant and intermediate rolls; in the centre was the powerhouse; and the buildings across the rear housed the crushing and separating machinery. The main separators were in the tower on the right. (From the photographic print collection, Edison National Historic Site, West Orange, New Jersey.) workers only operating and maintaining the equipment. To automate fully his operation was a costly and time-consuming undertaking but throughout his years at Ogden, Edison regularly improved and replaced machinery with this goal in mind. So caught up with automation was he that his associates remarked that he was happiest when, after installing a new automatic device, he could say, 'Boys, let five more men go tonight—don't need them.'35 Enormous and efficient as it was, how did the Ogden venture fare? While the newspapers heralded great triumphs in the mountains, in reality Edison encountered only prolonged difficulties. Upon completing construction of the basic plant in 1891, he received orders for concentrate from the Bethlehem Iron Company and North Branch Steel, but was soon beset by problems relating to the operation of the plant and to the overall condition of the iron and steel industry.36 Shortly after Ogden went into production, Edison discovered that he could not control the amount of phosphorus in the concentrate. Despite his de-phosphorizing chamber, the concentrate ranged from 0.028 to 0.07 per cent phosphorus while most commercially acceptable ores contained less

W. Bernard Carlson

49

than 0.03 per cent. Partly as a result of the high phosphorus content, Bethlehem Iron cancelled their order in 1891 after purchasing only a few thousand tons.37 Another difficulty which Edison immediately encountered was that the finely crushed concentrate failed to function predictably in the blast furnace (remember that when Ogden first went into production, Edison did not bother to mould the concentrate into briquettes). Instead of becoming molten pig iron, the concentrate frequently came out of the furnace as red hot chunks which could not be used to make steel. Blast furnace operators generally attributed this phenomenon to the failure of the concentrate to react properly with the furnace gases. Although the concentrate could be used successfully in blast furnaces having bosh angles between 83 and 90 degrees, most American furnaces had smaller bosh angles. Consequently, without extensive redesigning and rebuilding, most eastern blast furnaces were unable to use Edison's concentrate as their main ore charge. Warned of this difficulty as early as 1884, Edison never seems to have understood it. Instead, after his first orders were discontinued, he defined the problem in terms of shipping and handling the ore. In response, Edison decided to mould the concentrate into briquettes and, accordingly, purchased bricking machinery, developed a binder and set up the bricking section of his plant. Unfortunately, while bricking made the concentrate easier to ship, it did little to improve the behaviour of the concentrate in the blast furnace.38 It was not until January 1897 that Edison finally conducted his own tests of the briquettes at the Crane Iron Works in Catasauqua, Pennsylvania. In these highly publicized tests, he found that the output of pig iron for a furnace fully charged with briquettes was 33 per cent greater than normal. The pig iron produced in these tests was considered to be exceptionally strong and ideal for use in the Bessemer process. In addition, it was estimated that in employing the briquettes they used less fuel, limestone and labour. During the tests, it was reported that 'Nearly all the furnace men in the East visited Catasauqua and expressed their satisfaction at the working of the ore, and the prospective addition of briquettes to their sources of supply.' Edison, however, was not able to show that the briquettes could be used as readily and dependably as any other available ore and in any blast furnace. Consequently, blast furnace operators, who were working in a highly competitive and rapidly changing industry, preferred to employ more familiar ores rather than risk using the briquettes.39 Besides these technical difficulties, Edison encountered strong price competition. Initially, the Ogden concentrate appears to have been competitively priced; in 1891, Edison filled the North Branch Steel order at $4.75 per ton while Minnesota ore prices ranged from $4.25 to $5.00 and Cuban ore was $5.82 per ton. However, with the continued development of the iron-rich Minnesota ranges, the panic of 1893 and the emerging domination of several large midwestern steel firms, the price for iron ore steadily declined through the 1890s. By 1899, the price of Lake Superior ores had dropped to between $2.25 and $3.90 per ton.40

50

Edison: the Magnetic Ore Separation Venture, 1879-1900

Because of high maintenance costs and the expense of developing new equipment, Edison could not lower the price of the concentrate and remain competitive. In crushing and handling thousands of tons of ore daily, machinery at Ogden wore out and had to be replaced frequently. In June 1891, Samuel Insull estimated that Edison was losing about $6,000 a month trying to maintain and operate Ogden. Edison was also faced with the costs of developing and installing the equipment necessary to eliminate phosphorus from the concentrate and for moulding the concentrate into briquettes. While it is difficult to determine how much these two innovations cost, the expense of developing them undoubtedly increased Edison's operating costs and the price of the concentrate. Likewise, it is impossible to estimate how much Edison spent in upgrading Ogden to full automation. After 1895, he (Fig. 6) consistently claimed that he could sell his briquettes at $3.50 per ton but, given the high costs of replacing and developing equipment at Ogden, it is improbable that Edison made any profit at that price. Furthermore, as iron prices continued to fall through the 1890s, the briquettes were unmarketable at $3.50 per ton.41 The high phosphorus content of the briquettes, their poor performance in the blast furnace and the declining price of iron ore seriously impaired the Ogden venture. Primarily because of these difficulties, Edison had only five orders during the ten years that Ogden was in operation.42 In addition, however, Edison failed in his ore separation venture for a more fundamental reason: he simply did not accurately perceive the condition of the iron and steel industry. In formulating his market strategy, Edison correctly perceived the decline of the eastern blast furnaces, but he did not comprehend that this decline was but one aspect of a series of larger changes occurring in the industry. As the iron and steel industry moved from the east to the Trans-Allegheny or midwestern region, important technological and organizational changes took place. During the 1890s, iron and steel producing firms in the midwest surpassed those in the east not only because they had access to the cheaper Lake Superior ores, but also because they were technologically more advanced. Using coke instead of anthracite coal as fuel, employing laboursaving devices, being generally larger and more rationally designed, midwestern blast furnaces steadily increased their annual output of pig iron from 5.8 million tons in 1890 to 9.5 million tons in 1900.43 Organizational changes accompanied these technological changes. In the 1890s, the midwestern sector of the iron and steel industry came to be dominated by large firms with the investment capital necessary to expand and improve their plants. Led by expert managers, these firms acquired railroads, ore boats, iron mines and coal fields in order to vertically integrate their operations and further reduce costs. Highly competitive, the larger midwestern firms created a turbulent environment in which the smaller eastern firms were unable to survive. Since they could not compete with the larger and more technologically sophisticated firms of the midwest, eastern blast furnaces stagnated and declined. In short, the decline of the iron and steel industry in the east was not caused simply by a lack of inexpensive ore

W. Bernard Carlson

51

Figure 6. Edison at Ogden in 1895. (From the photographic print collection, Edison National Historic Site, West Orange, New Jersey.)

52

Edison: the Magnetic Ore Separation Venture, 1879-1900

(as Edison had assumed) but rather was a part of a larger pattern of technological and economic changes affecting the industry.44 As a consequence of these broad developments, no amount of inexpensive concentrate could have revitalized the eastern iron and steel industry. In reality, there was no significant market for Edison's briquettes. If the price of Lake Superior ores had been higher and if the eastern blast furnaces could have been adapted to handle the concentrate, then Edison might have been able to market the briquettes. Given the actual conditions of the industry, however, ore separation was not the 'market pull' situation Edison had assumed, but more of a 'technological push', requiring the cultivation of a market for concentrate and major alterations in the technological structure of the industry. For these general economic reasons, other inventors in the ore separation field appear to have fared little better than Edison. The only difference in their experience seems to have been that they lost far less since they did not erect plants as large as Ogden. Magnetic ore separation did not achieve commercial success in America until the 1920s when Edward Wilson Davis of the University of Minnesota perfected a process for economically concentrating taconite ores.45 Thus, the ore separation venture was unsuccessful because Edison failed to recognize the broad developments occurring in the iron and steel industry and instead assumed that a large market existed for iron concentrate. But why did Edison misperceive the industry? While his misperception was influenced by many complex changes that occurred in his life during the 1880s and 1890s, two factors in particular seem to have contributed directly to his misjudgement. First, as noted above, because Edison financed the ore separation venture largely on his own, he was isolated from outside control. Unlike his successful electric lighting project, where he was assisted by Grosvenor Lowrey and other investors in adapting his invention to the prevailing market, there were no outside advisers through most of the ore separation venture.46 Not constrained by the necessity of obtaining outside capital to build Ogden, Edison was not influenced by the opinions of others concerning the iron and steel industry, leaving him free to follow his own ideas. Second, when success was not immediately forthcoming at Ogden, Edison did not revise his market strategy but rather intensified his efforts to increase the efficiency of the plant. Confident that a market existed for his product, he introduced more efficient equipment and attempted to automate the plant's operations in order to produce competitively priced concentrate.47 Implicitly, he equated efficiency with profitability. Unfortunately, increased efficiency does not guarantee profits in every case; because of the additional capital required to design and construct a more efficient process and because of changing patterns of the market and general technological environment, a seemingly efficient process can become quickly obsolete. Preoccupied with the pursuit of efficiency and automation, Edison failed to realize that changes in the iron and steel industry were making his ore separation process uncompetitive. Instead, engaged by the challenge of automating Ogden, he preferred to retain his own misperception of the industry.

W. Bernard Carlson

53

Edison never seems to have understood fully the problems with Ogden and only gradually abandoned his ore separation venture. Towards the end, he made two intense but futile attempts at making Ogden commercially viable. In 1897, he devoted all his energy to reorganizing and automating the plant. He threw himself into his work, moved to Ogden, and seldom visited his family at Glenmount and his West Orange laboratory. Despite his efforts, he was still unable to deliver usable, competitively priced briquettes to his only customer at the time, Bethlehem Steel. In 1899, with a sudden rise in the price of Minnesota ore, Edison returned briefly to Ogden only to be discouraged by problems with the plant's decrepit machinery. With this attempt, his ore separation venture at Ogden ended.48 Yet Edison never publicly admitted defeat. As late as 1904, he told interviewers that he planned to rebuild Ogden and that he was going to employ a gold separator at a gold mine he had purchased in New Mexico. A group of British investors financed a large ore separation plant using Edison equipment in Dunderland, Norway, but Edison had little to do with this unsuccessful undertaking. Instead, he turned to manufacturing his phonograph and kinetoscope, began to develop a Portland cement plant and worked on an improved storage battery.49 This study of Edison suggests that an inventor's perceptions may play a significant role in the innovative process. It is well known that inventors succeed not because they create ingenious devices but because they successfully take into account the environment in which they function. Frequently, they invent in response to economic and technological needs and watch for opportunities that will permit their inventions to be successfully introduced.50 More subtly, however, the success or failure of a particular invention may depend not only on the nature of the environment but also on the inventor's perception of it. Perceiving the environment in a particular way, an inventor will proceed to work in certain directions giving his invention unique qualities. If he correctly assesses his environment, the inventor may succeed, but if he does not, he is likely to fail miserably. The case of Edison's ore separation venture provides a clear example of this; misperceiving the changing patterns of the iron and steel industry, Edison developed his ore separation process in a certain way and as a result was unable to market his briquettes. Thus, although it is not always examined explicitly, an important component of an inventor's* style is his perception of the environment. Acknowledgements Much of the research for this paper was done at Edison National Historic Site in West Orange, New Jersey with the help of A. Reed Abel and Leah Burt. Various drafts of this paper were critically read by Dr Robert Belfield, Dr James Brittain, Christopher S. Derganc, Dr Thomas P. Hughes, Dr Reese V. Jenkins, Regina B. Kampf and Dr Mark Rose. I am grateful to each of these individuals for their patience, inspiration and support.

54

Edison: the Magnetic Ore Separation Venture, 1879-1900 Notes

1. 'What Edison is Doing' in New York Sun, 1 July 1894 in Batchelor Scrapbook No. 1346, Edison National Historic Site Archives, West Orange, New Jersey. Hereafter, these archives will cited as ENHS. 2. An undated sheet in 1896 Ore-Milling, New Jersey and Pennsylvania Concentrating Works file, ENHS, gives the total expenditure for Ogden as of 1900 as $3,164,681.13. 3. In 1897, the editors of Iron Age estimated that eastern blast furnaces were paying approximately 37 cents more per ton of ore for rail freight than were plants in Pittsburgh. See 'The Edison Concentrating Works' in Iron Age, Vol. 59, p. 1 (28 October 1897) (hereafter cited 'Edison Works', Iron Age, 1897) and Axel Sahlin, 'Introduction and Development of Magnetic Separation of Iron Ores' in Engineering and Mining Journal, Vol. 53, p. 616 (11 June 1892). 4. Edison discussed his need for platinum and his plans for erecting a plant to extract platinum in a long memo dated November 1879 in 1879 General Electric Lighting File, ENHS. See also Laboratory Notebook N781122(1), 149-151, ENHS. His further experiments with other experiments are described in Notebook N800809, 2-3, ENHS. 5. US Patent 228,329 (1 June 1880) and Frank Lewis Dyer, Thomas Commerford Martin and William Henry Meadowcroft, Edison: His Life and Inventions, New York, 1929, pp. 941-2. 6. For information on the Edison Ore-Milling Company, see 'Proposed Plan of Reorganizing the Edison Ore-Milling Company, Limited', c. 1887, 1887 Ore-Milling file, ENHS (hereafter cited as 1887 Reorganization Plan) and Bryon Michael Vanderbilt, Thomas Edison, Chemist, Washington, 1971, p. 143. Details of the activities, staff and financing of the two plants are given in 'Memorandum of Cash disbursements by S.B. Eaton on account of the Edison Ore-Milling Company, Ltd., 28 May 1881-21 November 1881', 1881 Ore-Milling file, ENHS. For information on the problem of marketing the ore, see 'To the Stockholders of the Edison Ore Milling Co., Limited', c. 15 January 1884, 1884 Ore-Milling file, ENHS. Quote is from Engineering and Mining Journal, Vol. 52, pp. 733-4. 7. Matthew Josephson, Edison: A Biography, New York, 1959, pp. 299, 346-8 and 354-8. Particularly bitter about his experiences in the electric lighting industry and anticipating his success with ore separation, Edison is supposed to have said T am going to do something now so different and so much bigger than anything I've ever done that people will forget my name was ever connected with anything electrical.' From A.O. Tate, Edison's Open Door, New York, 1938, p. 278. 8. Quote is from an editorial titled 'Magnetic Ore Separation' in Iron Age which was partly reprinted in John Birkinbine, 'Progress in the Magnetic Concentration of Iron Ore' in Transactions of the American Institute of Mining Engineers (hereafter cited as Trans. AIME), Vol. 19, p. 673 (1890-1). 9. Christopher S. Derganc, 'The Role of Style in Invention: Edison's Electric Light and Power System' in IEEE Spectrum, Vol. 16, February 1979, pp. 50-9, and Thomas P. Hughes, 'Edison's Method' in William B. Pickett, ed., Technology at the Turning Point, San Francisco, 1977, pp. 5-22. 10. A.O. Tate to W.S. Perry, 27 June 1887, 1887 Reorganization Plan, W.S. Perry to Edison, 12 October 1887, all in 1887 Ore-Milling file, ENHS, W.K.L. Dickson appears to have had most of the responsibility for organizing the ore separation laboratory; see W.K.L. and Antonia Dickson, The Life and Inventions of Thomas Alva Edison, New York, 1894, p. 323 and W.K.L. Dickson, 'Changes Made in the Ore-Milling Outhouse', 24 November 1888, W.K.L. Dickson File, ENHS.

W. Bernard Carlson

55

11. Walter S. Mallory took an active part in improving the separators, and from 1888 to 1890 he personally invested in and supervised Edison's experimental ore separation plant in Humboldt, Michigan. See W.S. Mallory to Edison, 24 January 1889, 1889 Ore-Milling file and 'Preliminary agreement between Edison Ore-Milling Company and Walter S. Mallory', 25 July 1888, 1888 Ore-Milling Mineral Assayfile,ENHS. In the late 1890s, Mallory became vice-president of the New Jersey and Pennsylvania Concentrating Works. Another expert was John Birkinbine, a prominent Philadelphia mining engineer and President of the American Institute of Mining Engineers in the early 1890s. Birkinbine had studied extensively the distribution of blast furnaces and the supply of iron ore in America; indeed, he may have suggested to Edison that ore separation could produce the ore needed by the stagnating eastern iron industry. See, among others, the following articles by Birkinbine: 'Influence of Location Upon Pig-iron Industry' in Trans. AIME, Vol. 21, 1892-3, pp. 473-91; 'Prominent Sources of Iron-Ore Supply' in Trans. AIME, Vol. 17, 1888-9, pp. 715-28; and 'The Distribution and Proportions of American Blast Furnaces' in Trans. AIME, Vol. 14, 1885, pp. 561-75. Retained as a consultant by Edison from 1888 to 1889, Birkinbine tested other separators, gathered samples and introduced Edison to iron ore dealers. See J. Birkinbine to Edison, 30 July 1888, 1888 Ore-Millingfile;J. Birkinbine to Edison, 13 April 1889, 1889 Ore-Milling, Birkinbine file; and J. Birkinbine to Edison, 17 July 1890, 1890 Ore-Milling, Birkinbinefile,ENHS. 12. For lists of books and maps on mining in the Edison library, seeJ.R. Ash to Painter, 18 August 1888, 1888 Ore-Milling file and notebook N900215. Edison secured samples by sending out assistants to gather them and by purchasing them from mineral dealers; see P.D. Dyer to Edison, 19 September 1887, G.L. English and Company to Edison, 17 June 1887, and G. A. Simmons to Edison, 8 December 1887, all from 1887 and 1888 Ore-Millingfiles.Quote is from R.E. Eaton to Edison, December 1887, 1887 Ore-Millingfile,ENHS. 13. 'The Edison Magnetic Separator' in Iron Age, Vol. 41,6 December 1888, pp. 847-8 and the following US Patents: 377,518 (7 February 1888), 396,356 (15 January 1889), 476,991 (14 June 1892), 430,280 (17 June 1890), 434,588 (19 August 1890). 14. The Humboldt plant was operated from August 1888 to December 1890 when it was destroyed byfire.See W.S. Mallory to E.G. Thomas, 13 October 1888, 1888 Ore-Milling, Mineral Assay file; W.K.L. Dickson to W.S. Mallory, 14 February 1890, letterbook LB900131, pp. 306-7; W.S. Mallory to Edison, 30June 1890 and 5 December 1890, 1890 Ore-Milling, Malloryfile,ENHS. 15. For a general description of the Bechtelsville plant, see Thomas W. Leidy and Donald R. Shenton, 'Titans in the Berks: Edison's Experiments with Iron Concentration' in Historical Review of Berks County, Vol. 23, Fall 1958, pp. 104-10. See also the following clipped articles in Scrapbook No. 48, ENHS: Philadelphia Record, 29 March 1889; New York Sun, 18 July 1889; and Patterson Press, 15 July 1889. In addition, Edison operated a separator in London from September 1888 to March 1889; see Edison to J. Dredge, 14 February 1889, A.O. Tate to J. Dredge, 15 February 1889, A.O. Tate to S. Insull, 28 February 1889, and Edison to J. Dredge, 22 March 1889, all in letterbook LB890130, pp. 993, 237, 428, and 762, ENHS. 16. For general information about the surveys, see Dyer, Martin and Meadowcroft, as note 5, pp. 479-80, Vanderbilt, as note 6, p. 149, Josephson, as note 7, p. 370 and 'Edison Works,' in Iron Age, 1897, p. 2. Specific information about the areas surveyed is given in notebooks N891208, N900412 and N900925 and the following letters: C.J. Reed to Edison, 3 February 1890, 1890 Ore-Milling, Searches, Maps and Surveys file; R.D. Casterline to Edison, 21 August 1891 and 30 August 1891,

56

Edison: the Magnetic Ore Separation Venture, 1879-1900

1891 Ore-Milling, Survey file, ENHS. For information on the dipping needle compass, see J.C. Smock, 'The Use of the Magnetic Needle in Searching for Magnetic Iron Ore' in Trans. AIME, Vol. 4, 1875-6, pp. 353-62. 17. Untitled memorandum on Edison's plans for ore separation, c. 1892-1893, Eisenman Memorial Collection, American Society of Metals, Metals Park, Ohio (hereafter cited as Edison Marketing Memo, c. 1892). I am grateful to Cyril Stanley Smith for providing me with a typed transcript of this document. Edison also elaborated on his marketing strategy in several interviews: see Engineering and Mining Journal, Vol. 52, 26 December 1891, p. 734; 'A Talk with Edison' in Scientific American, Vol. 66, 2 April 1892, p. 216; and 'The Edison Magnetic Concentrating Works' in Scientific American, Vol. 78, 22 January 1898, p. 57. 18. For description of other ore separators, see John Birkinbine, 'Progress in Magnetic Concentration of Iron Ore' in Trans. AIME, Vol. 19, 1890-1, pp. 656-74; Clinton M. Ball, 'The Magnetic Separation of Iron-Ore' in Trans. AIME, Vol. 25, 1895, pp. 533-51; 'The Buchanan Magnetic Ore Separator' in Engineering and Mining Journal, Vol. 47, 15 June 1889, p. 542; Harvey S. Chase, 'The Chase Magnetic Separator' in Trans. AIME, Vol. 21, 1892-3, pp. 503-12; Robert Anderson Cook, 'The Wenstrom Magnetic Separator' in Trans. AIME, Vol. 17, 1888-9, pp. 599606; John C. Fowle, 'Magnetic Concentration at the Michigamme Iron Mine, Lake Superior' in Engineering and Mining Journal, Vol. 50, 29 November 1890, p. 628; F.H. McDowell, 'Magnetic Concentration at Tilly Foster' in Trans. AIME, Vol. 21, 1892-3, pp. 519-21; and J.P. Wetherill, 'The Mine Hill Ore Deposits in New Jersey and the Wetherill Concentration Plant' in Engineering and Mining Journal, Vol. 64, 17 and 24 July 1897, pp. 64-6 and 98-9. 19. James B. Ross, 'Technological Convergence in Taconite Concentration', unpublished paper, Department of Mechanical Engineering, University of Minnesota, 1982. I am grateful to Mr Ross for making this work available to me. 20. Edison Marketing Memo, c. 1892, p. 9. 21. John Birkinbine and Thomas A. Edison, 'The Concentration of Iron Ore' in Trans. AIME, Vol. 17,1888-9, pp. 728-45; John Birkinbine, 'Progress in Magnetic Concentration of Iron Ore' in Trans. AIME, Vol. 19, 1890-1, pp. 656-74; and clipped articles from New York Sun, 18 July 1889 and Patterson Press, 15 July 1889 in Scrapbook No. 48, ENHS. After demonstrating his separator to the iron dealers, Edison reportedly declared: Why the men engaged in it [the iron business] haven't any sand. They don't know what they want. All of them agreed that the machine was just the thing that was needed, and the want of something of this kind had been greatly felt for years. But I am through with the iron men. They are the worst set of fellows to introduce an invention to that I ever met. From 'Edison and the Iron Trade', undated newspaper clipping attached to F.T. Morton to Edison, 2 January 1890, 1890 Ore-Milling, Ore Separator file, ENHS. 22. See 'Certificate of Organization of the New Jersey and Pennsylvania Concentrating Works', 3 January 1889, 1889 New Jersey and Pennsylvania Concentrating Works file, ENHS. Edison provided most of this company's capital, but some of his employees and close associates also invested in the venture. Robert L. Cutting, a New York financier associated with the electric lighting companies, invested $225,000, while Walter Mallory, Charles Batchelor, William Perry, Francis Upton, A. O. Tate and J. Hutchinson contributed smaller amounts. See W. S. Perry to Edison, 3 October 1893, 1893 Ore-Milling, New Jersey and Pennsylvania Concentrating

W. Bernard Carlson

57

Works file and 'The Present State of Mr. Edison's Undertaking is as Follows:', memorandum c. 1895, 1895 Ore-Milling, Edison, N.J.file,ENHS. 23. See, for example, E.G. Spilsbury, 'Improvements in Mining and Metallurgical Applicances during the Last Decade' in Trans. AIME, Vol. 27, 1897, p. 457. 24. Edison to J. Birkinbine, 23 April 1890, 1890 Ore-Milling General file; Pilling and Crane to Edison, 12 February 1894, and W.S. Pilling to Edison, 6 March 1894, both in 1894 Ore-Milling, Mineral Surveys, Pilling and Cranefile,ENHS. 25. For examples of managerial decisions made by Samuel Insull, Charles Batchelor, Francis Upton and Edwin H.Johnson, see Batchelor Daybook, No. 1336, 1886-7, ENHS. Quote is from 'A Talk with Edison' in Scientific American, Vol. 66, 2 April 1892, p. 216. Another example of Edison's unwillingness to have outside investors associated with ore separation may be taken from the way in which a friend of Edison described hisfinancialposition: Edison owns it all. There will be no corporation, no stocks, no bonds, no debentures, absolutely nothing that the kings offinancecan get hold wherewith to create, by a vote of directors and a stroke of the pen, enormous obligations which can be trafficked at the Stock Exchange. Edison is now a capitalist. If he needs to borrow money he has his own securities to get it from, and he has become so disgusted with and suspicious of those operations which are called financing, that he wants never more to be associated with them. From 'Edison Planning Victories Anew' in Philadelphia Press, 5 February 1893, in Batchelor Scrapbook No. 1346, ENHS. 26. 'Edison Works' in Iron Age, 1897; 'Edison's Latest Feat' in New York Sun, 31 October 1897, 'Edison Ore Mines' in Electrical Review, 27 October 1897, both in Batchelor Scrapbook No. 1246; and 'The Ogden-Edison Mine', in Newark Sunday Call, 3 June 1894, in Batchelor Scrapbook No. 1346, ENHS. 27. The cableway, which was used until 1894, is described in 'The Lidgerwood Suspension Cableway and the Edison Ore Concentrating Works at the Ogden Mines, New Jersey' in Engineering and Mining Journal, Vol. 55, 10 October 1891, pp. 425-6. After 1894, all descriptions of the plant mention the use of railway skips to transport rock to the crushing mill, so presumably the cableway had been discontinued. See 'The Ogden-Edison Mine' in Newark Sunday Call, 3 June 1894, in Batchelor Scrapbook No. 1346, ENHS. 28. Vanderbilt, as note 6, p. 155. 29. 'Edison Works' in Iron Age, 1897, p. 3. 30. Dyer, Martin and Meadowcroft, as note 5, pp. 948-50. 31. A thorough explanation of the complete ore separation process can be found in 'Edison Works' in Iron Age, 1897, p. 5. The percentages are from Vanderbilt, as note 6, p. 158. 32. For reference to the phosphorus dust being used to make paints, see G.N. Morison to Moriarty, 27 September 1894, 1894 Ore-Milling, Edison, N.J. file, ENHS. The sand, which proved to be a very profitable by-product,* is discussed in Theodore Waters, 'Edison's Revolution in Iron Mining' in McClure's Magazine, November 1892, pp. 91-2. 33. 'Edison Ore Mines' in Electrical Review, 27 October 1897, Batchelor Scrapbook No. 1246, ENHS. 34. Thomas Robins, Jr., 'Notes on Conveying-Belts and their Use' in Trans. AIME, Vol. 26, 1896, pp. 78-96; 'Magnetic Ore-Separation at Edison, New Jersey, U.S.A.' in Engineering, 12 November 1897, pp. 580-1; and 'Edison Works' in Iron Age, 1897, pp. 3 and 7.

58

Edison: the Magnetic Ore Separation Venture, 1879-1900

35. 'Edison's Wand Defies a Trust' in New York Journal, 29 November 1898, Batchelor Scrapbook No. 1346, ENHS; 'Edison Works' in Iron Age, 1897, p. 5; Dyer, Martin and Meadowcroft, as note 5, p. 487; and Waters, as note 32, p. 80. Edison also had some difficulty with his work force, which may have contributed to his interest in automated machinery. In 1895, 105 machinists went on strike after Edison refused to pay them double for Sunday work and overtime. Rather than negotiate with the strikers, Edison closed the plant 'for the purpose of a general weeding out'. See 'Edison Works Close Down' in Philadelphia Declaration, 30 August 1895, 1895 Ore-Milling file, ENHS. Quote is from 'Edison Ore Mines' in Electrical Review, 27 October 1897, Batchelor Scrapbook No. 1246, ENHS. 36. For examples of laudatory newspaper accounts, see 'What Edison is Doing' in New York Sun, 1 July 1894, Batchelor Scrapbook No. 1346; 'The Edison Iron Works' in Boon ton Bulletin, 9 January 1896, 1896 Ore-Milling file; and 'Edison Triumphs with Iron Magnet' in New York Press, 29 October 1897, Batchelor Scrapbook No. 1246. In 1891, Edison shipped 36,000 tons of concentrate to Bethlehem Iron and 5,000 tons to North Branch Steel. See Robert P. Linderman to Edison, 14 May 1891, 1891 Ore-Milling, Ogden, N.J. file; 'The Present Situation of Mr. Edison's Undertaking is as Follows:', c. 1895, 1895 Ore-Milling, Edison, NJ. file; and A.W. Howe to North Branch Steel Company, 12 November 1891, 1891 OreMilling, Ogden, N.J.file,ENHS. 37. During the early years at Ogden, the phosphorus content ran very high, ranging from 0.036 to 0.072 per cent; see Charles Batchelor's experiments to reduce the phosphorus content in Batchelor Daybook No. 1337, 1887-92, pp. 188-98, ENHS. In the first shipments to the Bethlehem Iron Company, the phosphorus content ranged from 0.036 to 0.046 per cent; see Bethlehem Iron Company to Edison, 9 June 1891, 1891 Ore-Milling, Ogden, N.J.file,ENHS. By 1897, even with the dephosphorizing chamber, the ore still contained between 0.028 and 0.033 per cent phosphorus; see 'The Edison Magnetic .Concentrating Works' in Scientific American, Vol. 68, 22 January 1898, p. 57. 38. The problems of using concentrate iron ore were frequently dismissed by proponents of ore separation; see, for example, Axel Sahlin, 'Introduction and Development of Magnetic Separation of Iron Ores' in Engineering and Mining Journal, Vol. 53, 11 June 1892, p. 53. Other sources, however, are quite explicit about the problems involved in using iron concentrate; see J. Wiborgh, 'The Use of Fine Concentrates in the Blast Furnace' in Engineering and Mining Journal, Vol. 68, 9 September 1899, pp. 305-6 and W.S. Pilling to Edison, 23 January 1893, 1893 Ore-Milling, Briquettesfile,ENHS. In an 1884 report, it was noted that operations were discontinued at the Rhode Island separation plant because the concentrate could not be smelted and was consequently unmarketable; see 'To the Stockholders of the Edison Ore Milling Co., Limited', c. 15 January 1884, 1884 Ore-Milling file, ENHS. For the development of the briquetting machinery, see entries for 15 April, 27 May, and 8 June 1892 in Batchelor Record Book No. 1337, pp. 210-11 and 21517, ENHS. 39. L.J. Peckitt to Edison, 22 January 1897, 1897 Ore-Milling, General file, ENHS; Wiborgh, see note 38, pp. 305-6. 40. For the prices of Lake Superior iron ores, see American Iron and Steel Association, Statistics for American and Foreign Iron Trades for 1891, Philadelphia, 1892 and Statistics for American and Foreign Iron Trades for 1900, Philadelphia, 1901. For the price of Cuban iron ore, see 4 December 1891 entry, Batchelor Record Book No. 1337, p. 181, ENHS. 41. Insull's remark is from the entry for 21 June 1891, Batchelor Record Book

W. Bernard Carlson

59

No. 1337, p. 154, ENHS. The problem of perfecting the binder for the briquettes seems to have been particularly expensive and time-consuming; one article noted that Edison performed over 6,600 experiments before finding a suitable binder compound. See 'Edison's Latest Feat', undated clipping, Batchelor Scrapbook No. 1246, ENHS. In 1891, Edison calculated his operating costs to be $2.66 per ton and that he could sell the concentrate at $5.28 per ton. Since thesefiguresfail to take into account any replacement costs, it is hard to imagine that Edison could have made any profit by selling his ore in the late 1890s at $3.50 per ton. See entry for 8 March 1891, Batchelor Record Book No. 1337, p. 143, ENHS. 42. In addition to the Bethlehem Iron and North Branch Steel orders, Edison had three others: Reading Iron Company, unknown amount in 1893; Carnegie Steel, 1,700 tons in 1897; and Bethlehem Steel, 1,797 tons in 1898. See A. Broder to Pilling and Crane, 16 January 1893, 1893 Ore-Milling, Briquettes file; 'Testing Edison's Iron Ore' in Electrical Review, Vol. 31, 1 December 1897, p. 31; and Batchelor Scrapbook No. 1345, 1898-1902, p. 1, ENHS. 43. For the general development of the iron and steel industry in the late nineteenth century, see W. David Lewis, Iron and Steel in America, Greenville, Delaware, 1976, pp. 43-8. The figures were compiled from statistics of blast furnaces given in Iron Age, Vol. 45, 12 June 1890, p. 1003 and Vol. 65, 8 March 1900, p. 22. In this case, the Midwestern furnaces were those in western Pennsylvania, Ohio and Illinois. 44. Lewis, as note 43, pp. 49-51. In an 1898 interview, Edison acknowledged that his adversaries were the large midwestern steel firms, particularly the newlyformed Federal Steel Company. However, still confident about Ogden, Edison described his project with populistic overtones—that the revitalization of the eastern iron industry would provide 25,000 jobs and that the midwesternfirmswere unfairly exercising monopolistic control over the western ore market. See 'Edison's Wand Defies a Trust5 in New York Journal, 29 November 1898, Batchelor Scrapbook No. 1346, ENHS. 45. Ross, 'Technological Convergence in Taconite Concentration'. 46. C.S. Derganc, 'The Role of Style in Invention'. 47. In an interview in 1898, Edison explained his marketing strategy in the following way: Why can we lead the Mesaba ores in the market? That is simple. The secret lies in the fact that from the time the rocks are drilled by compressed air, blasted out, loaded into cars by a steam shovel and hoisted to the breakers in mechanical skips and derricks, it is unnecessary for human labor to intervene until the finished briquettes are delivered by an automatic carrier to the railway cars. From 'Edison's Wand Defies a Trust' in New York Journal, 29 November 1898, in Batchelor Scrapbook No. 1346, ENHS. 48. For details concerning the 1899 attempt, see entries for 24 October 1899, 9 July 1900 and 10 January 1901, Batchelor Record Book No. 1345, pp. 81, 155 and 195, ENHS. 49. 'Edison Will Mine the Gold' in Omaha Bee, 22 January 1898 and 'New Edison Project' in New York Press, 25 February 1898 in 1898 Ore-Milling, Ortiz mine file, ENHS. For the Dunderland project, see Robert Conot, A Streak of Luck, New York, 1979, pp. 345-6. 50. For discussion of the market strategy of inventors, see Nathan Rosenberg, Technology and American Economic Growth, White Plains, New York, 1977, pp. 55-7.

S a m u e l t h e

B r o w n : D e s i g n

h i s o f

I n f l u e n c e

o n

S u s p e n s i o n

B r i d g e s T H O M A S DAY Introduction It is intended that this paper should complement Paxton's account1 of the manner in which Telford developed his design for the Menai Bridge and that it should outline briefly Brown's approach to the design of the same type of structure. No attempt will be made here to present details of individual designs because this has been done by Professor E.L. Kemp in the pages of this journal in 1977.2 The emphasis will now be on Samuel Brown's ideas and concepts with a view to assessing in broad terms the extent of his practical experience and theoretical knowledge. In Britain during the second and third decades of the nineteenth century, he was arguably the most prolific builder of catenary bar-chain bridges and was associated with the construction of at least twenty-four bridges or piers and made designs and estimates for many more. Sadly his work and influence as an engineer have until lately been neglected and it is now hoped that this paper— together with the earlier one by Professor Kemp—will go some way to redeeming the situation. Samuel Brown (Fig. 1) was born in London on 10 January 1774.3 He was the eldest son of William Brown, merchant of London and formerly of Borland, County Galloway, and Charlotte, the third daughter of the Reverend Robert Hogg, Minister of Roxburgh.4 Nothing is known about his childhood, his upbringing or his education. On 8 June 1795 he entered the Navy as an able seaman. Within four weeks of this date he was promoted to the rank of midshipman, later becoming a master's mate before applying to sit his lieutenancy examination in June 1801.5 Brown saw active service during the Napoleonic Wars and in 1805 was present, as first lieutenant of the British frigate Phoenix, during the particularly furious single-ship action in which the larger and more heavily-armed French frigate Le Didon was forced to surrender to her British adversary. Subsequently, while still serving in the Phoenix, Brown took part in the naval action when Sir Richard Strachan's squadron captured the remnants of the French fleet which had escaped after Nelson's victory at Cape Trafalgar. His active naval career ended with a number of short appointments and on 1 August 1811 he was advanced to the rank of commander and, having been unable to obtain further promotion, he accepted the rank of Retired Captain on 18 May 1812.6

62

Samuel Brown: The Design of Suspension Bridges

Figure 1. Samuel Brown (1774-1852). Before his enforced retirement from active naval service, Brown had registered his first patent7 in which he defined his proposals for equipping ships with iron standing rigging. In total, he registered twelve patents. They show a diversity of topics and while many, including the early patents, were concerned with nautical matters, later ones included ideas for the construction of inclined planes, lighthouses and breakwaters and proposals for improvements to ship construction. The two patents for which Brown is best

Thomas Day

63

known specified new methods for manufacturing wrought-iron chains8 and constructing bar-chain suspension bridges.9 Initially, Brown's interests turned towards the active promotion of chain cables and he established two small manufactories at Narrow Street, Ratcliff, and the Borough near Waterloo Bridge.10 These premises were soon working to capacity and in 1812 Brown set up another chain works at Mill Wall, Poplar, which was in close proximity to the Naval Dockyard at Deptford and the London and India Docks. Finance for the establishment of the chain-making business was provided by Brown's cousin Samuel Lenox, a successful merchant, and in 1808 the two men formed a partnership which traded under the name of Samuel Brown and Company.11 The manufactory at Poplar was unable to meet the increasing demand for chain cables so in 1818 Brown and his partner opened a new works at Newbridge (Pontypridd), South Wales. It was his success as a manufacturer of chain cables and the knowledge of iron gained in this context that gave Brown the basis on which to develop and promote his interest in bridge building. The Prototype Brown concentrated his attention on the design and construction of flatdeck suspension bridges sometime between 1808 and 1813.12 A date closer to the latter would fit neatly into a theory that he was directly influenced by reports of Finley's work in America13 and Douglas' lectures at the Military College,14 but there is no evidence to this effect. He knew of the contribution made by the Chinese and was fully aware of the problems associated with bridges whose decks followed the catenary curve of suspension chains; he wrote . . . Such an arrangement is evidently a bad one, inasmuch as we must ascend to the point of suspension, then descend, and rise according to the curve of the chain, . . . This is hardly practicable, and my earliest attention was employed to remedy this evil.15 He constructed a prototype bridge in 181316 or 1814.17 It is not possible to determine whether the inspiration for this was his need of a model to back proposals for his plan for a suspension bridge at Runcorn then being considered by Telford,18 his entrepreneurial zeal or his need to verify his design by experimentation.19 In 1825, when promoting an unsuccessful proposal for St Catherine's Bridge across the Thames, Brown described his prototype bridge as follows: . . . I erected a bridge with the road or platform perfectly horizontal, on my premises at Mill Wall, where it still remains. This is effected by introducing perpendicular rods through the joints of the main suspending bars, and adjusting their length to the curve above, so that they formed a series of straps for the reception of a row of bars on each side, placed edgewise, and extending the whole length of the bridge parallel

64

Samuel Brown: The Design of Suspension Bridges to the entrance. The beams being laid across these bars, the platform or road becomes quite horizontal; or an ascent may be given from the sides to the middle in the same plane as with the roads leading to the bridge.20

It may have been similar in construction to that shown later by Dutens (Fig. 2). The prototype had a span of 105ft (31.98m), and 37cwt (1.88 tonne) 21 'dt. Jhind .j-hoJt

d'oJrr JanJ LtJ dti/un/ ^c j]lj laniuil

l]r»#n}

Figure 2. Duten's illustration of Brown's prototype bridge. of iron was used for its construction. Although built either for demonstration or for experimentation, this bridge stood for more than ten years. It was tested with loaded wagons and was subject in its early life to inspection by Telford22 and Rennie,23 who at different times drove across it in carriages. Both engineers seemed to have found it satisfactory. Later Rennie, when being questioned before the Parliamentary Select Committee considering Telford's Menai Bridge proposals, gave evidence that he had had his coach driven across several times while he inspected the structure and that there had been very little vibration.24 B r o w n ' s Patent for S u s p e n s i o n B r i d g e s Brown's main ideas for suspension bridge building were contained in the patent 'Invention or Improvement in the Construction of a Bridge, by the Formation and Uniting of its Component Parts in a Manner not hitherto Practised' which he registered in London in 181725 and in Edinburgh a year later.26 In the former he described briefly some of the factors that influenced the development of his design, and outlined the general features of the construction techniques he proposed to use by making specific reference to a design for a suspension bridge 1,000 ft (304.6 m) long. To assess Brown's contribution to the development of suspension bridges it is necessary to study in detail the ideas expressed in these patents and show how they were related to current practices and ideas. The first point that should be noticed is that Brown made no claim for a patent for the idea of suspension bridges but to a new method of constructing them. . . . I, the said Samuel Brown, do hereby declare that my said Invention or Improvement consists of a bridge or bridges of suspension, in the construction of which, instead of using metal chains formed of links,

65

Thomas Day

wires, or by other methods, as heretofore practised, I employ a combination of straight bars, bolts or rods, joined or united at their ends either by side plates and bolts, coupling boxes, welding, or other suitable methods, so that these bars, bolts or rods so joined become in effect one entire length (of the bridge) . . . and these constitute my main lines or means of suspension.27 He acknowledged past practice of using chains and the recent use of wire (Galashiels Bridge) as main load-bearing agents and concluded that In all bridges of the Catenary order there must be some similarity in the principle of forming pieces, the points of suspension, and securing the extremities of the chains .. .28 It was his interest in chain cable making coupled with the results obtained from a series of tests on the strength of iron carried out at his chainworks29 that caused Brown to opt for straight bars rather than conventional chains. Anderson30 and Telford31 reached the same conclusion. To support this choice, he quoted the results of tests on two straight bars and a chain all made from No. 3 Welsh iron. These are shown in Table 1.32 TABLE 1 Dimensions Bars 2 in. dia. x 12 ft 6 in. long (51 mm x 3.80 m) 3 in. deep x 1 in. thick x 12 ft long (76 mm x 25 mm x 3.66 m) Chain \\ in. dia. bar links x 12 ft 6 in long (38 mm dia. bars)

Wtight

Breaking Load

1321b (60 kg) 1241b

82.75 ton (84.3 tonne) 72.25 ton

18.5 in. (0.47 m) 13 in.

(56 kg)

(73.5 tonne)

(0.33 m)

2411b (110kg)

79 ton (80.4 tonne)

23 in. (0.58 m)

Elongation

Brown's main argument for the use of rods or bars in lieu of chains was based on an examination of the weight ratio of the test pieces. The results were selective. They demonstrated clearly that although the rods and chain carried comparable loads, the chain was nearly twice as heavy. Brown argued that this made the use of chain unfavourable in terms of cost, because of the increased quantity of iron and workmanship and the loss of carrying capacity resulting from its additional weight. He stressed the fact that a majority of bridges were public works, that the supply of chains for them was liable to be put out to public contract and that it was problems with inferior-quality chains that had led him and his rival, Brunton, to recognize the importance of testing all chain cables before they left their manufactories. Possibly favouring his own case, he quoted the strength of cables obtained

66

Samuel Brown: The Design of Suspension Bridges

^ ^

J

RECONSTRUCTED ) /

(^

OF CHANNEL (~*\

"""]

PRESENT

PROFILE

CHANNEL

WALL

PROFILE BED

I E. SIDE)

OF MILL

(E. SIDE)

OF AQUEDUCT SLOPING Q i

AND

LIMITS

BANK

OF

ROCK

CUTTING

BUSHES -MASONRY PLANTS

—\-—

f—WALL

COURSES

3ZH

AND

STONES

END

ARLES

g^*L-LARGE

3 AQUEDUCT

ROCK

SURFACE OF

LIMIT

PATCHES

ON

FLOOR

ROCK

OF CHANNEL ,

WALLS

/

UPPER

r

CHANNEL

ISOLATED

"w

1

LINING

OF CHANNEL

LINING

I

£7« ROCK CUTTING

PEIRO

TR0UCAD0

LINING

96

The Large Roman Water Mill at Barbegal (France)

Figure 6. View of south end of rock cutting 'Peiro troucado' near aqueduct junction with mill.

Robert H. J . Sellin

97

Figure 7. View of north end of rock cutting. Right-angle turn in Aries aqueduct is below bush at right-hand side of photograph. site more carefully, thought that it represented a reservoir divided into three major basins and a number of minor ones and thought that it had formed one of six Imperial cloth factories which were known to have been established. The same writer noted fragments of basalt mill wheels lying at the foot of the slope and concluded that these had served for grinding colouring pigment for cloth dying. Another local historian writing a little later recognized these millstones as of a type suitable for grinding corn and he therefore concluded that the site had had a double function—cloth finishing and corn grinding. A German archaeologist writing about the aqueducts of Nimes and Aries in 1910 accepted this double-function theory but J. Formige writing in 19242 took the view that the site's sole function had been flour milling. A systematic excavation of the mill was planned in 1935 and carried out in 1937 and 1938 by F. Benoit. He published a full report1 of his investigations in 1940 in which he concluded that this structure was definitely not a reservoir built on the side of a hill but a mill building designed to be powered by water wheels arranged in two parallel cascades down the slope, one on each side of the building. His ability to reconstruct a detailed ground plan and elevation (Figs 3 and 4) of the mill was due to his careful excavation of the lower parts of the mill channels where, in order to get the maximum benefit from the available drop in head, the water channels and the chambers containing the mill wheel axle supports had been constructed below ground level. These water passages and chambers below ground level had become filled with silt and debris soon after the mill fell into disuse and this

98

The Large Roman Water Mill at Barbegal (France)

had not only preserved the masonry structure there in good order but it had also preserved the calcareous encrustations which were found to have formed on the water passage walls, giving clear confirmation of their use, and also by the manner in which these deposits had been formed and eroded, suggesting ways in which the waterwheels had been controlled by the deflection of water through openings in the mill chutes. C.L. Sagui, writing in 1948,3 made estimates of the hydraulic power available at this site and also of the flour output. The two important quantities to be determined in arriving at the power availability are the head and the discharge. The total head across the mill can be measured today and is approximately 20 m but the discharge value is much more difficult to determine. Sagui measured the width of the approach aqueduct and found it to be 0.8 m (and this can be confirmed, see Fig. 7) but he also states that he saw deposits which suggested that the depth of flow in the aqueduct had been 0.5 m. Today such marks are only visible in the Aries aqueduct where it appears that the depth of flow was nearer 1 m. He then goes on to base his calculations on a discharge of 1 m3/s, which implies that he is adopting a mean velocity of 2.5 m/s in the mill aqueduct, which is very high. The problem of water consumption will be discussed again later but it is sufficient to say here that the high discharge value adopted by Sagui leads to a power availability at each of the sixteen water wheels which he believed would have been sufficient for each to drive two grindstones and his estimate of flour output is therefore based on thirty-two units. This he calculates to be 28 tonnes in 24 hours or enough to feed 80,000 people. As well as the water consumption and flour output figures, there remains another problem of interpretation in connection with the upper part of the mill site where the remains of two massive masonry walls set at an angle to the mill building enclose a triangular space between the termination of the aqueduct channel and the upper wall of the mill (see Fig. 3). These two inclined walls together with the missing upper end wall of the building have always been accepted as forming a distribution reservoir feeding the two mill channels but this interpretation is now in doubt as a stability analysis carried out for the missing third wall—the foundation bench cut in the rock is clearly visible and measurable—shows that it could not have retained water safely to the required depth. T h e Mill Aqueduct: N e w E s t i m a t e s o f D i s c h a r g e a n d Flour Production Of the two aqueducts which cross the valley to the north of Barbegal (see Fig. 5), the easterly one leading to the mill was supported for the most part on a continuous masonry substructure or wall. The more westerly Aries aqueduct had a channel bed which ran at a higher level than that of its companion which made it more economic to be constructed as a continuous arcade, a form commonly associated with Roman aqueducts although used in reality over a very small proportion of their total route length. It was thought that the Aries aqueduct was constructed first because the

Robert H. J . Sellin

99

line of the mill aqueduct can be seen (Fig. 2) to deflect noticeably to the west where it emerges from the rock cutting and joins on to the mill structure. This change of direction allowed the mill to be built at its most advantageous orientation on the hillside. Benoit argues that, had the mill aqueduct been constructed either before or at the same time as the Aries aqueduct, a route would have been chosen across this valley that would have obviated the need for a last-minute change of direction. However, the advantages would have been very small and it cannot be claimed that this argument is conclusive in establishing precedence of construction. Probably because the mill aqueduct was constructed on a solid substructure in the region close to the rock cutting, the Teiro Troucado', much of this remains today. Enough in fact to make the measurement of aqueduct bed gradient a possibility. Figure 5 shows in plan the aqueduct remains in this region as they are today and it is quite easy to make out at gaps the precise upper level of the fine red concrete used here to line the aqueduct channel. Where it remains today, this lining is covered by a thick calcareous deposit quite different in colour, texture and thickness from the concrete. This in its turn is covered by a layer of soil, stones and plants. Bed levels were determined at four stations along this part of the mill aqueduct where the upper surface of the lining material was accessible and also at one point on the bare rock surface. The position of the five level stations is indicated in both Figs 3 and 5 and the results in Table 1. TABLE 1 Measurements of relative level and gradient for points on mill aqueduct bed as indicated in Fig. 5. Level Station Horizontal Location (m) Relative Levels (m) Gradient Between Stations Indicated

1

2

3A

3

4

+ 63.50

+ 49.70

+ 34.44

+ 9.00

0.00

-0.101

+ 0.041

-0.136

Zero

-0.368

-0.010

-0.005 + 0.001

Level Taken on surface

-0.002 In Table 1 a positive gradient indicates a falling bed level in the assumed downstream flow direction. The results achieved are erratic which could be explained by a number of possibilities: (a) the aqueduct builders did not work to a very precise bed gradient

100

The Large Roman Water Mill at Barbegal (France) and here it should be remembered that a short length of adverse gradient would not seriously affect the functioning of the aqueduct; (b) parts of the aqueduct substructure built on alluvial material may have settled more than other parts, probably immediately after construction; (c) the whole area may have tilted to the north due to crustal deformation along the margin of the Mediterranean basin during the last 1,700 years; (d) the level measurements may be in error.

The level traverses all closed within acceptable limits and were repeated a number of times so that (d) seems unlikely, but it is possible that some of the measurements did not in fact lie on the original bed of the channel due to local damage or repair of the lining material. The substructure under station 1 is far enough away from the ridge to have suffered settlement but it is difficult to see why 3A should have gone down when 2 did not. The possibility of a general tilting of the whole area can be resolved in the future if levels can be taken of the aqueduct bed at some point on the north side of the valley where the aqueduct is once more resting on bed-rock and these related to levels at station 3. It is quite possible that one day these levels will be seen to have been affected by more than one of the above factors and maybe others as well. It will now be seen that it is not possible at present to establish the discharge by first determining the aqueduct gradient. However, from experience with sewers of about the same dimensions, it would seem that a mean velocity of 1 m/s is much more probable in this case than the 2.5 m/s implied in Sagui's calculations. Another quite separate aspect of the discharge problem which should be considered is the discharge capacity of an overshot waterwheel of the dimensions used here. A wheel diameter of about 2.1m fits in with the measurements of the mill channels and an effective (inner) wheel width of 0.7 m is the greatest that could be accommodated with safety in the stone channels 1 m wide. Assuming a speed of rotation of lOrpm and an equivalent uniform bucket depth of 0.2 m gives a usable discharge of 0.15 m3/s although the actual value may have been considerably less. This gives an aqueduct discharge of 0.3m3/s which results in a mean velocity of 0.8 m/s and a gradient (using a Manning n value of 0.015) of 0.001. It has been assumed here that the waterwheels were of the overshot variety. Benoit1 clearly supports this view, citing the shape of the wheel chambers as well as the dowel holes in the superior sill stones as evidence. Landels4 reviews the evidence available for water power under the Greeks and Romans and concludes that the wheels at Barbegal were overshot rather than undershot. On the evidence of the site it is difficult to see how undershot wheels, which are so suitable for use by a river bank requiring the minimum of construction, should have been fitted at Barbegal. Nevertheless, the grounds for assuming one sort in preference to the other are

Robert H. J. Sellin

101

meagre and further archaeological or literary evidence may appear which in the future would alter the balance of opinion. It now appears that the potential power available from this aqueduct was 60 kW, not Sagui's value of 196 kW and, if his 65 per cent efficiency is accepted, each waterwheel could now have generated 2 kW compared with the previous figure of 6.6 kW. However, a millstone of 0.9 m diameter—the size suggested by the fragments found—has been estimaed to need 3.6 kW at an assumed speed of 53 rpm so that it now looks as though this speed of rotation may not have been realized here and the estimated output of milled flour must be reduced accordingly from 40 to 24 kg per hour per stone. It is also quite clear that there was never any possibility of each waterwheel powering two grindstones (as Sangui calculated) so the maximum flour production now stands at 9 tonnes per 24 hours' continuous output for the whole mill. Time lost due to the maintenance needs of the aqueduct and the mill machinery must have been large and in addition there must have been interruptions due to inadequate supplies of grain or water as well as normal management and labour disruptions. It would therefore seem reasonable to assume a 50 per cent load factor for the mill and this cuts its average flour production down to 4.5 tonnes per day or enough to feed a population of 12,500 based on a consumption unit of 350 g per day. This corresponds closely enough to estimates of the size of population of Aries in the fourth century AD SO that it would now appear that Sagui was incorrect in assuming a substantial export trade in milled grain from Aries at that period due to Barbegal alone. T h e Mill a n d the Water D i s t r i b u t i o n A r r a n g e m e n t s One of the more obvious features of the Barbegal site is that water was supplied by a single channel cut through the crest of the ridge (Fig. 6) at the upper end and was discharged by a double-channel arrangement at the bottom. The arrangements at the lower end wer*e made clear by Benoit's excavations and these still remain substantially open to inspection (Fig. 8). It was by a process of backward extrapolation from these low chambers and water channels that Benoit was able to arrive at his reconstruction of the whole mill (Figs 3 and 4). There is nothing in the remains today (Fig. 2) to contradict his scheme of a double eight-stage water mill arrangement with each waterwheel contained in its own step in the channel and each millstone assembly housed in the corresponding internal chamber in the mill building. The mill structure can be considered as five longitudinal divisions. The two outer ones, shown in Figs 9 and 10 respectively, being essentially masonry walls 3.5 m thick, each containing a stepped mill water course. The two next to these consist of stepped rooms containing the mill machinery, probably of several stories each, while the remaining central division consisted originally of an access passage including timber platforms and timber and masonry stairs. This permitted the movement of grain and flour between the entrance to the building (centre front) and the individual mill

102

The Large Roman Water Mill at Barbegal (France)

Figure 8. Lowest waterwheel chamber (1) in eastern watercourse of mill.

Robert H. J . Sellin

103

Figure 9. Western water channel structure. rooms. The lower courses of many of these internal walls are standing today, the plan being fairly complete on the east side of the building and rather less so on the west. A systematic arrangement in four rows of fourteen large stones, or their foundation platforms, are thought to have supported a system of pillars which probably carried a timber and tile roof across the whole mill building. This may well have been stepped in elevation in line with the internal mill chamber arrangements as shown in Sagui's reconstruction of the mill, Fig. 11. Whether the roof covered the three central divisions of the building in a single pitch or was constructed separately over the two mill chamber divisions with an auxiliary unpitched roof, following the slope of the hill, over the central access corridor cannot now be determined. To judge from Fig. 11, Sagui had experienced some difficulty with the roof, or had not appreciated the problem, because his drawing leaves the matter unresolved. One point is clear: he had access to Benoit's drawings when he prepared his own reconstruction. If Benoit's positioning of the eight waterwheels is correct (Fig. 4), there are some interesting points which arise and which he discusses in his report. A millstone must be mounted with its shaft vertical and a waterwheel of this type with its shaft horizontal. This means that a satisfactory right-angled gearing is required and that the millstones must either be on a floor above the waterwheel shaft or on a floor below. Looking at the profile and section in Fig. 4 it seems most probable that the millstones in chambers 1 -3 were mounted above their wheels and those in 4-8 below. There are certainly small recesses constructed below the floor level in chambers 1-3 (on the far side of the shaft opening in Fig. 8)

104

The Large Roman Water Mill at Barbegal (France)

Figure 10. Lowest three chambers of eastern water channel.

Robert H. J . Sellin

105

106

The Large Roman Water Mill at Barbegal (France)

Figure 12. Western diagonal wall at top of slope. which would have housed the inboard waterwheel axle bearing, the millstone shaft lower bearing and the gearing. For chambers 4 and upwards, this arrangement would have been reversed, which explains why the floor of chamber 4 is without any such low-level recess and why it is benched into the hillside to a greater extent than its neighbours. Sagui is probably correct then in assuming that the central part of the mill building would have risen to several stories although his drawing does not agree with the site, or with Benoit's elevation, in his positioning of the lower waterwheels. The lowest ones (chamber 1) would certainly have been out of sight, probably below ground level (see Fig. 8). The triangular space at the top of the site has always been assumed to have formed a distribution reservoir for the two mill channels. Of the three walls enclosing this space, substantial parts of the two diagonally-inclined ones remain (Figs 12 and 13), indicating that these were at least 2.5 m thick. All that remains of the third wall, however, which also formed the upper end wall of the mill building, is a carefully cut and clearly visible foundation bench some 1.1m wide shown in Fig. 13 and running the full width of the

Robert H. J. Sellin

107

Figure 13. Foundation bench for missing end (north) wall of mill building. building. If this space did in fact contain a distribution reservoir, then the missing wall must have successfully withstood a hydrostatic pressure distribution on its uphill side corresponding to the level of the water surface in the reservoir relative to the base of the wall. Whether or not the reservoir bottom was formed on a horizontal platform constructed behind the end wall is not likely to have affected significantly the horizontal loading on this wall unless the reservoir was essentially watertight and the supporting structure, now vanished, very well drained. What can be easily demonstrated is that if this wall, with a base width not exceeding 1.1m, had been subjected to an upstream hydrostatic pressure distribution over its full height due to a water level equal to the inferred level of entry to the mill channels, then its factor of safety against overturning would have been 0.3 and against failure by sliding 0.85. This situation would have been impossible. Even if a sufficiently shallow and watertight reservoir had been successfully constructed on an adequately drained base, it seems very probable that at sometime during the useful lifetime of the mill (thought to have been 100 years) the drainage would have become blocked and again failure of the wall would have occurred. This dangerous situation could easily have been avoided by providing a simple channel fork where the aqueduct leaves

108

The Large Roman Water Mill at Barbegal (France)

the rock cutting (the rock walls are in fact cut back at the correct angle for the last 2 m of the cutting) and running two separate channels out over the very substantial diverging walls still remaining and shown in Fig. 12. This reconstruction of the water distribution arrangements avoids altogether the need for a reservoir and was first suggested by A.J. Parker.5 It would appear to be supported strongly by the difference in thickness of the walls under consideration: 2.5 m for the diverging walls and 1.1 m for the foundation of the transverse wall. The only requirement that this scheme makes on the transverse wall is that there should have been some provision for drainage to prevent rain water and seepage from the aqueduct from accumulating in the enclosed space formed by the three walls. A feature of the diverging walls which, perhaps more than any other single factor, was taken to support the, reservoir theory is the existence of an opening high up in each (see Fig. 12), with their lower sills 1.4 and 0.7 m respectively below the present-day levels of the tops of the walls. These were interpreted as drainage openings serving a reservoir with its bottom at the sill level and used from time to time to flush out silt. If this was so it is difficult to see why an opening was provided on each side as a single one would have served one reservoir well enough and made the provision of a drainage channel to carry the water from here to the foot of the slope and clear of the mill area much simpler. There are clear signs of a boundary wall having enclosed the mill building and a surrounding ground area. The regular stepped foundation can be seen on the left of Fig. 2 and the full plan has been drawn in by Benoit, Fig. 3. This wall could have caused further problems if water had been released from these openings in large quantities. Alternatively, it is possible that they were part of a flow control or diversion structure in the twin channels that it is suggested ran along the top of these diverging walls. Finally, on this point, it is not clear what purpose a distribution reservoir could have served at this location. Its water storage capacity would have been insignificant and if a settlement tank had been required to trap sediment carried by the water an appropriate structure could have been provided much more easily further back along the aqueduct. The Gallo-Roman mill at Barbegal has clearly been of great interest to local historians and archaeologists, particularly in relation to the economic and industrial organization of the late Roman Empire, and it can also prove to be of interest to hydraulic engineers concerned with the early history of man's control over water and its useful application. This site must be one of very few constructed and operated and fallen into ruin before the Industrial Revolution in the west on which it is possible to make numerical estimates of discharges, power availability and output which help to put it and other early engineering works into context today. Notes 1. F. Benoit, 'L'usine de meunerie hydraulique de Barbegal (Aries)' in Revue Archeologique, Serie 6, Vol. 15, No. 1, 1940, pp. 19-80.

Robert H. J . Sellin

109

2. J. Formige, Les monuments romains de la Provence, Champion, Paris, 1924, p. 18. 3. C.L. Sagui, 'La meunerie de Barbegal (France) et les roues hydrauliques chez les anciens et au moyen age' in Isis, Vol. 38, Feb. 1948, pp. 225-31. 4. J.G. Landels, Engineering in the Ancient World, Chatto and Windus, London, 1978, pp. 16-26. 5. A.J. Parker, University of Bristol. Private communication.

T h e

U s e

M i n i n g :

A

o f

G u n p o w d e r

D o c u m e n t

o f

i n 1 6 2 7

G. H O L L I S T E R - S H O R T The history of gunpowder blasting in mining operations has attracted little attention from British or American historians, at least in respect of the invention and diffusion of the technique.1 However, the appearance in the last few years of two studies, both of which are concerned with throwing light on the origins of blasting, provide an occasion for reviewing what is currently known about the matter. 2,3 These and other sources for the early history of the use of gunpowder in mining will be the subject of a subsequent article. Here, by way of foretaste, I reproduce the earliest unimpeachable document recording the use of the new technique, a translation of it, and an account of what is known about the engineer, Caspar Weindl, who carried out the trials in question. On 8 February 1627 Caspar Weindl demonstrated his blasting technique (Sprengkunst) before the members of the Emperor's Court of Mines from Schemnitz.4 The trial took place in the Daniel crosscut of the Oberbieberstollen mine near Windschacht (now Stiavnicke Bany) a few kilometres from Schemnitz (now Banska Stiavnica), one ofseven mining towns in a region then known as Lower Hungary (now Central Slovakia). The trial was a success. Within five years gunpowder blasting was being used throughout the region, and indeed far beyond it. A report of the trial drawn up by Christopher Spilberger, Secretary to the Court of Mines, is preserved in the State Central Mine Archive at Banska Stiavnica.5 Text Adi 8. Februari dits 1627 Jars hat die Ganz Loblich Gwerkschafft beim Haubtperkhwerch Ober BiberstoUen, Ihro Kai. Mai. perggericht zur Schembniz zur Einfart wegen des Casper Weindlss Sprengwerch solches in Augenschein zunemen, ob es dem Gezimerwerch durch dass schiessen schedlich sein mechte in beratschlagung zu siehen begruesst. Uber solchem eingenomenen augenschein, vnd in Gegenwart der Ambleut, sowohl des Perggerichts beschehenen Schuss hat sichs befunden, dass dises Sprengwerch wol fiirzunemen sei, vnd nichts schedlichs causirn werde, ob zu Zeitten gleich ein Rauch entstehet, vergeet er doch in ainer Viertl Stundt vnd ist den hewern ohne schaden, nimbt auch vil boses Wotter mit sich wegckh, aber offt zu schiessen wurde es nit thuen, dann es wurde die andern Khuren im Arzthauen vnd Gefol, wenn Sie offt sollen stilhalten verhintern, aber fur Rathsamb war, die weillen im Danielschlag schone Anbruch vorhanden, die aber zimblich fesst, doch keine heiier, die man zuelegen mechte, vorhanden sein, daselbst

112

The Use of Gunpowder in Mining: A Document of 1627

sowol in den Schachten vnd Stolwenten auch der Soolen, liesz sich dass Sprengwerch gar wol an. Weiter ist damallen Caspar Sprenger befragt worden, ob Er dise Ortter im Danielschlag wolte zu Lehenschafft annemben, weil das ainzige Ortt im Tieffisten den Vncosten mit dem Sprenger nicht ertragen wurde, hieniber meldt solcher, wenn man Ime 40 oder 50 guette heuer gibt, so traue er Ihme dise orter gar wol mit der herrn Gwerckhen nuzen guetten zu Lehenschafft anzunemben. Auf solch sein erpieten wirt Ime Caspar darauf geantwort. Weil im Tieffisten vil Ortter aus Mangl Heier feieren miiessen vnnd dits Orts allein ain 40 heier von Noten vnd sein doch keine vorhanden, ob man nit mitl haben konne, souil heier etwa von andern ortten herzubringen. Darauf meldt Caspar, wann man den Vncosten, der darauf geen wurde, nit ansehen, noch sparen wolt, vnnd Ime ainen Passbrieff von Ihr Kai. Mai. ausbringen vnd ertailen wiirde, Trauet Er Ime gar wol auss Tyroll ain anzal guetter Heier zu notturfft an solche Ortter als in das Tieffeste, Danielschlag, hintern Kiisten, Schachten, Stolwant an der Sool vnnd andre Ortter zuezuweiten, vnd ins werckh zusizen, herein zu bringen. Souil thuet das Kaiserlich Perggericht ain Ganze Lobliche Gwerckhschafft berichten, welche ohne massgeben auf solcher Verern beratschlagungen des Casparn Sprengers zuesagen vnnd erpietten ins Werckh zusezen wissen werden. Datum Schembniz den 16. Februari Ao 1627. Georg Putscher Pergm [aister] Caspar Pistorius Chri. Spilberger Perggerichts schreiber Translation On February 8 of the year 1627 the most excellent mining company at the main mine of Ober Biberstollen invited the Emperor's mine chamber at Schemnitz to come to the mine for a demonstration of Caspar WeindPs gunpowder blasting device (Sprengwerch) to see whether it would damage the timber work. After an explosion had been demonstrated in the presence of the mine chamber and the mining officials it was established that this gunpowder blasting device could be used without fear of damage; although it produced smoke, this dispersed within a quarter of an hour, and did not do the faceworkers any harm; and moreover it got rid of a good deal of bad air (Vil boses Wotter). There would not be many explosions so as not to cause the other ore-extracting teams to break off too often. But the use of explosives could solve the problem of how to exploit the ore bodies in the Daniel cross-cut which were of very good quality, but very hard, and there were no faceworkers available to work them. The gunpowder blasting device could be used in the shafts and the horizontal galleries leading from the soles of the shafts. So Caspar Blaster (Caspar Sprenger) was asked whether he would be

G. Hollister-Short

113

willing to lease these parts of Daniel's cross-cut, since it would be too costly to use the explosive for just one place at the deepest part of the mine. He replied that if he were given 40 or 50 faceworkers he would take on the lease of these places, with the backing of the chamber officials. Caspar received the following reply: since many places in the deepest parts of the mine had to remain unworked because of the lack of faceworkers, and this place alone required 40 faceworkers, when there were none available, would it not be possible for faceworkers to be brought from other places? Caspar replied that if the expense of such an initiative were forthcoming, and he were given an imperial warrant (ie letter of credit) he was confident that he could bring a fair number of good faceworkers from the Tyrol, and set them to work in the places where they were needed—in the deepest part of the Daniel cross-cut, and the further workings, shafts and galleries leading from the soles of the shafts, and other places further afield. So the imperial chamber informed the most excellent mining company of Caspar's worthy suggestion, who gave their agreement without more ado, and undertook to set the work in train. Dated Schemnitz 16th February 1627. Georg Putscher Mine Surveyor Caspar Pistorius Chri. Spilberger Secretary to the Mine Chamber Caspar Weindl Weindl came from Rottenberg in the Tyrol where his father held an appointment as mine manager (Verwalter) under the Crown. It seems reasonable to assume that Caspar acquired some acquaintance with mining matters during his youth. In the 1620s he was serving, probably as an officer with the sappers, in the army of General Raymond Montecuccoli, then on campaign in Italy. He himself was to assert that it had been in Florence (probably in 1625 or 1626) that the plan was formed for him to go to Schemnitz to see whether his technique would avail against the hard and almost unwinnable rock that the miners encountered there.6 Whether it was at the behest of his former commander that he made the journey is one probability. The fact that Count Jerome Montecuccoli, the General's brother, was a major shareholder in the Brenner Company which at that time controlled the Oberbieberstollen mine certainly lends colour to this idea. It is also possible that Weindl had worked in Schemnitz before the Italian expedition. The Tyrolean mines were becoming worked out, and numbers of miners from there had made their way eastwards to Slovakia. That Schemnitz still suffered from an acute labour shortage the report of 1627 itself makes clear. At all events, Weindl had settled down in Schemnitz well before 8 February 1627. In May 1629 Weindl married his third wife, Ursula Poly, daughter of a former commander of Schemnitz Castle. He had by that time secured an important position as an inspector on the staff of the Brenner Company.

114

The Use of Gunpowder in Mining: A Document of 1627

Weindl was evidently a haughty and overbearing personality. He was quarrelsome, and got drunk more often than was considered decent. When charges were brought against him by the town's law officers, however, he denied that they had any authority over him. In 1632 or 1633 he petitioned the Crown for reward in view of the fact that his technique had made possible a doubling of silver production while he had received up to that time only 600 Florins in recognition of his services. He died in 1646.7 As for the trial blast of 8 February 1627, it can hardly have been the first that Weindl had carried out. There must, at the very least, have been a series of tests designed to determine how many shot holes were required in very hard rock, and how much powder was required in the charges. One should consider also that in the report of the trial, drawn up only eight days afterwards, Weindl was twice referred to by what was evidently his nickname 'Sprenger' (blaster). That he could be easily and naturally referred to in this way suggests that he had been perfecting his technique for some time (wherever it was that he did so) and that his fame had gone before. Like all innovators he encountered hostility and suffered from attempts to belittle his success, even to the extent of mine records being falsified so as to understate the efficacy of the new technique. A final point which calls for consideration is how to explain the speed with which the technique of gunpowder blasting was diffused far beyond Slovakia, so that by the 1650s it was known virtually throughout Europe. No doubt the answer is to be found in the economics of the matter. Firesetting, the technique it superseded, required the burning of prodigious quantities of timber to soften the rock at the work face. Many parts of Europe were experiencing something approaching a wood famine well before 1627, so that a technique which offered at least a partial escape from such physical and financial pressures could not fail to be attractive. Indeed, gunpowder blasting takes its place naturally among a number of other techniques already in use at that time whose object was to save wood altogether or significantly reduce the amount needed.8 At the same time blasting sharply raised productivity.

Notes 1. J.R. Partington, A History of Greek Fire and Gunpowder, Cambridge, 1960, scarcely mentions the matter. A few lines on p. 174 is all that he has to offer. A History of Technology, Oxford, 1957, Vol. 3 (ed. C. Singer et al) has nothing at all. 2. R. Vergani, 'Gli inizi dell'uso della polvere da sparo nell'attivita mineraria: il caso Veneziano', Studi Veneziani, NSIII, 1979, pp. 97-140. 3. J. Vozar, 'Der erste Gebrauch von Schiesspulver im Bergbau (die Legenda von Freiberg—die Wirklichkeit von Banska Stiavnica)'. Studia Historica Slovaca, X, Bratislava, 1978, pp. 257-80. 4. Schemnitz is so often confounded with Chemnitz that it seems worthwhile to state that they are not variant spellings of one and the same place. Chemnitz (now Karl-Marx-Stadt) is (or was) in Saxony in the D.D.R. 5. Statny ustredny bansky Archiv, Mine Chamber Records, Berg Protokoll

G. Hollister-Short

115

Buch, sub anno 1627. I must thank Mrs Heather Nicholas, GCES, U. of Sussex for invaluable help in translating this Report. 6. Weindl stated this in his petition (drafted c. 1632-33). He asked the Emperor for compensation in view of the fact that weekly silver production at Schemnitz had almost doubled, '... fast Toppelte Vermehrung der wochentlichen Sylbergefollen ...' since blasting began. Weindl was no ordinary miner. He employed a number of Latinisms in his plea: 'cuncitrn', 'tergiversirn'—'delay', 'tergiversation'. 7. I have drawn this account of Weindl's career from Josef Vozar's study cited above. 8. As for instance the preliminary concentration of brine before boiling down. Leckwerke, Strohkunste (graduation houses) are mentioned in Germany c. 1560, and that at Bad Nauheim was built in 1579. Proposals for economical stoves and furnaces are to be found in F. Kessler, Holzsparkunst, Nuremberg, 1618, and in the French edition Espargne bois... invention de certains et diversfourneaux artificiels, par Vusage desquels on pourra annuellement espargner une infinite de bois, Oppenheim, 1619. The English adaptation to mineral fuel at this time is part of the same process.

F e r m e n t a t i o n P r a c t i c e : P u r e

t h e

Y e a s t

E n g l i s h

T h e o r y

a n d

B e g i n n i n g s

C u l t i v a t i o n

B r e w i n g , M I K U L A S

o f a n d

1 8 8 3 - 1 9 1 3

T E I C H

On 12 November 1983 it will be exactly 100 years ago that in the Gamle [Old] Carlsberg brewery at Valby (a suburb of Copenhagen) the so-called pure culture yeast was employed in the production of bottom-fermented beer on a large scale for the first time. It happened at the behest of Emil Christian Hansen (1842-1909), the head of the Physiological Department at the Carlsberg Laboratory, who was able to bring the hitherto rather neglected botanico-physiological point of view to bear on investigations of brewery yeasts. It was part of a momentous change.that eventually led to the preferment of bottom yeasts over top yeasts in the brewing industry throughout the world. In Britain, however, despite an initial interest in pure yeast, top fermentation systems prevailed. The introduction of systematically-selected culture yeasts into brewing after 1883 constitutes an exceedingly good example of transformation by science of a traditional technique in a craft industry where operations to a considerable extent were still guided by a blend of empirical and scientific knowledge. At the same time, the circumstances that then made the English brewers reject the adoption of pure culture systems form an instructive illustration of the complex nature of the interaction between industrial techniques and science. They will be briefly examined in this account but, by way of preface, it will be convenient to consider, at least in outline, the process of brewing, including certain historical aspects of top and bottom fermentations. One further observation: the article draws on English rather than Scottish and Welsh material and this is reflected in its title. Brewing Operations; Top Yeasts and Bottom Yeasts Beer is commonly produced from barley, hops, yeast and water in a series of operations that include malting, mashing, wort boiling and fermenting. As a result of malting and mashing, the first two major processes, a sweet liquid termed 'wort' is obtained, containing principally malt sugar. The next step is the boiling of the wort with hops and this in turn sets the stage for fermentation in fermenting vessels by adding yeast to the hopped but cooled wort. Fermentation is the process by which malt sugar and other ferment-

118

Yeast Cultivation and English Brewing, 1883-1913

able sugar present in the wort are metabolized by yeast to ethyl alcohol and carbon dioxide. There are two types of yeast in brewing: top yeasts which rise to the surface of the fermenting wort and which ferment at temperatures from about 15 to 22°C, and bottom yeasts which sink to the bottom of the fermenting vessel and are adapted to lower temperatures from about 6 to 12°C. Bottom-fermentation beers are put into cold storage flagering') where they undergo further fermentation at about 0 to 4°C, whereby they mature and are saturated with carbon dioxide. Broadly speaking, topfermented beers are speedier to produce than the bottom-fermented ones. By and large, they are slightly more alcoholic (c. 4-6.5 wt % alcohol) than the bottom-fermented beers (c. 3-5 wt % alcohol). Except in Bavaria, top-fermentation systems were in universal use until about the 1840s. As to the origin of lager beer, there is no firm basis on which to build but there is some indication that it may have been transplanted from Bohemia to Bavaria during the second half of the fifteenth century and further developed there. Be that as it may, the principles of Bavarian bottom-fermented lager brewing had essentially been worked out by 1800.1 Among the first non-German brewers to direct his attention to bottom fermentation was Jacob Christian Jacobsen (1811-87) from Copenhagen. After personally collecting bottom yeast from the then leading Munich Spaten brewery in 1845, he introduced it into his own brewery.2 In order to expand the production of bottom-fermented beer he established a new brewery outside the city's boundary which he called after his son Carl. It began operating in autumn 1847 and became known as the Old Carlsberg brewery following the building of another brewery nearby during 1870-1.3 Carlsberg Laboratory Among brewers in particular and industrialists in general, J. C. Jacobsen was in a class by himself at that time in insisting that fundamental scientific research had an indispensable part to play in industrial development. His wealth enabled him to transform his vision into practice by founding, in connection with his brewery in 1875 (in his own words): a laboratory for chemical and physiological research and studies in those branches of the sciences which are particularly important for the processes of malting, brewing and fermentation, with the object not only of giving day-to-day assistance to the brewery, but also of giving the scientific staff the opportunity and resources to become qualified and active specialists in those fields which are particularly germane to the operations of brewing and the phenomena which they reveal.4 The creation of the Carlsberg Laboratory primarily for basic research into chemical and physiological aspects of brewing preceded the establishment of the Carlsberg Foundation, with the aim of promoting essentially all branches of science and humanities, which J. C. Jacobsen secured financially

MikuldS Teich

119

a year later. With the undoubted historical significance of the Carlsberg Foundation for the development of organized scholarship, we are not concerned in this essay. But Jacobsen's proviso expressly disallowing any secretiveness with respect to results of research obtained in the Chemical and Physiological Departments of the Carlsberg Laboratory is worthy of notice. Brewing Seasons Before the brewing industry began to benefit from the advantages of artifical refrigeration on a large scale from about 1880, it found it difficult to operate during the warmer months of the year. This applied above all to bottom-fermented beers that demand relatively low temperatures. In accord with this there existed in Bavaria the ordinance since the early sixteenth century prescribing that bottom-fermented barley beer was not to be brewed between St George's Day (23 April) and Michaelmas Day (29 September). On the other hand, rulers of Bavaria possessed the sole right and privilege of brewing top-fermented wheat beer throughout the year. Clearly, it was up to the ordinary brewers to prepare adequate supplies of bottom-fermented beer for the summer during the winter months when they were allowed to brew. If a brewer, for whaever reason, failed to produce enough lager beer for the late spring and the summer months, he suffered economically. The economic advantages that accrued to the Bavarian Court from the privilege of brewing top-fermented wheat beer during the whole year are not difficult to perceive. No doubt, in Bavaria these circumstances encouraged activities and also contributed towards the search for technically and economically efficient ways to brew by bottom fermentation. The time restriction on brewing operations in Bavaria (somewhat modified in 1805) was completely abolished in 1865, but the change-over to year-round brewing was not rapid. For instance, in the Spa ten brewery it did not take place until 1888.5 Outside Bavaria, non-brewing during the warmer season was apparently due to empirical custom rather than to official regulation. In the 1860s the breweries at Burton-on-Trent, the centre of English brewing, 'were almost completely shut down during the summer, the main brewing operations being carried on between the months of October and May'. This information comes from the eminent English brewing scientist Horace T. Brown (1845-1925), who started his career as junior brewer with the Worthington brewery at Burton in 1866. Later, there will be occasion to refer to him and also to his valuable 'Reminiscences of fifty years' experience of the application of scientific method to brewing practice'. They were presented by him to a meeting organized in his honour in London on 8 May 1916 and subsequently published in the Journal of the Institute of Brewing in the same year.6 We know from the already referred to letter by Jacobsen to Sedlmayr of 7 May 1884 that beer was brewed at Old Carlsberg during seven to eight months of the year until 1874. After this date the period was extended to nine months, from the beginning of October to the end of June, until 1882 when the brewery changed to all-the-year-round production.

120

Yeast Cultivation and English Brewing, 1883-1913 H a n s e n ' s A c h i e v e m e n t : P u r e Yeast Culture

But things were not going according to plan and large amounts of spoilt brews, for unaccountable reasons, had to be discarded as undrinkable. The solution to theoretical and practical aspects of this problem was supplied by Hansen. Like others connected with the theory and practice of fermentation at that time, Hansen had come under the spell of Louis Pasteur (1822-95). Actually, Hansen was so captivated by Pasteur that he chose the latter's pronouncement regarding the potential value of pure research in everyday life as one of the two quotations prefacing the volume in which his own practical and theoretical studies were brought together.7 But Hansen's admiration for Pasteur's accomplishments had not prevented him from examining them critically. Pasteur believed that the diseases in beer were caused by bacteria present in yeast and he prescribed a chemical treatment: cultivation in a solution of cane sugar rendered slightly acid. In opposition to this widely-favoured (bacteriologico-chemical) position of Pasteur, Hansen, as a botanist, adopted a botanical approach. Well conversant with fungal morphology and physiology, Hansen identified the contaminating culprit in three species of air-borne and insect-borne 'wild' yeasts, particularly abundant during the three summer months. It should be added that Hansen became alerted to these findings through his previous investigations into the occurrence of micro-organisms in the atmosphere in and around the Garlsberg brewery (1879, 1882). According to Hansen, the brewery yeasts, as long as they were propagated under the conditions obtaining in the respective breweries, were to be considered as constant types. In order to ensure this 'constancy', Hansen suggested cultivating appropriate species of yeasts proceeding from one individual yeast cell. In connection with this, the terms 'pure yeast' and 'pure cultivation' came into use and it is of more than semantic interest that Hansen himself was not too happy about it. He thought that they were misleading in the way they produced the wrong impression about the basic feature of the reform of brewing he was propounding. Hansen regarded selected suitable single species of brewery yeast for cultivation as its underlying principle. Here, undoubtedly, the source of inspiration was Charles Darwin (180982), the author of the other quotation that served as the guiding maxim behind his work. What Hansen found particularly illuminating was Darwin's emphasis that inasmuch as man was engaged in artificial selection, he had to work on natural variation due to minute changes in the conditions of life. Like Darwin, Hansen appreciated that a good deal was to be learned from practical plant breeders regarding useful or deleterious variations. Indeed he stated that he was led 'to introduce into the fermentation industries the same principles which have long been adopted in horticulture and agriculture for the cultivation of the higher plants.,8 There is no question that Hansen was keen on defending the priority of his achievement. He was especially sensitive to suggestions that he was preceded by Robert Koch (1843-1910), who described his improved

Mikuldi Teich

121

method of plate culture employing nutrient gelatine also in 1883. In Hansen's view, this procedure yielded uncertain results regarding pure culture because there was no guarantee of starting cultivation from an individual cell. Hansen's line was that it was precisely in this respect that Koch's course differed from his own, which rested on obtaining a single yeast cell from carefully diluted yeast solutions as the starting point of cultivation. Further supporting his claim that his method had no genetic relationship with Koch's, Hansen stressed that he had reported, albeit briefly, about his own researches in 1882. At that time the Tuborg brewery at Copenhagen was plagued with 'yeast turbidity', affecting beer in casks and bottles exposed to temperatures that were higher than that of the lager cellar. Such a product was not popular with the customer who demanded clarity in the beer he was drinking. The Tuborg brewery took up Hansen's warning about the need to avoid the presence of unsuitable ('wild') yeast species causing this particular beer irregularity and acted successfully upon it. This was not an unimportant factor contributing to a change ofJacobsen's opinion on the practical value of Hansen's researches with the effect that the latter was allowed to experiment with selected brewery yeast on a large scale in the old brewery in the autumn of 1883. While not doubting that culture yeast could be as effective as 'primary' yeast for the course of fermentation in fermenting vessels, Jacobsen was sceptical that it could bring about 'secondary' or 'afterfermentation' in casks where the beer was drawn off ('racked') after the completion of the main fermentation. For this purpose, in accordance with the commonly held view, Jacobsen accepted that wild yeasts are desirable as secondary yeasts, assumed to be responsible for the production of beer flavour. After Hansen's large-scale experiments showed that Jacobsen's apprehensions were unfounded and that the pure yeast culture system produced beer of good flavour, his change of mind was complete. A pure culture of what became known as Carlsberg bottomyeast Iwas introduced into the old brewery with a speed that even Hansen found unduly quick during 1883-4. Once Hansen's method had aroused Jacobsen's enthusiasm, he began to argue the case for it and Gabriel Sedlmayr, the owner of the Spaten brewery, was his first target: As my old master you shall be the first to whom as your pupil I report about my newest experience concerning yeast degeneration . . . from now on fermentation in my brewery will wholly be carried out by means of this pure yeast, produced from a single celll Truly a triumph of scientific research! When spoilage of beers occurred in brewery plants where standard conditions of cleanliness were maintained or even improved, it was attributed to 'degenerated' yeast. In such cases it was customary to resort to a change of yeast, acquired from another brewery, or even to utilize a mixture of yeasts from several breweries. The successful operation of Hansen's system led Jacobsen to break with the tradition and to accept the link between the

122

Yeast Cultivation and English Brewing, 1883-1913

plethora of anomalous species of yeast in the atmosphere during certain seasons of the year and the appearance of unpleasant flavours and other irregularities in beer. To quote from the letter to Sedlmayr: I now am convinced that in all breweries yeasts are infected, more or less, by species of wild yeasts because almost everywhere brewing is taking place during the dangerous summer months and the interminable changing of yeast, therefore, is of no avail. In previous times when as a rule, especially in Bavaria, no brewing took place during the summer months, the change of yeast in good breweries was also a rare exception. Sedlmayr was not only the first non-Danish brewer who obtained information about Hansen's efforts, but he also received from Jacobsen in repayment of a debt, as it were, a sample of Carlsberg culture yeast. It was expedited to Munich apparently by passenger train on 8 May 1884 and tried out three days later. Outside Copenhagen, this marked the beginning of the gradual recognition of the technical and economic advantages of using culture yeast on the European continent. Among the technical improvements which contributed materially to its spread was the apparatus for the continual production of pure culture yeasts devised by the collaborative efforts of A. Kuhle, the director of the Old Carlsberg brewery, and Hansen during 1885-7. Presumably because of the 'open' policy views held by Jacobsen, disapproving of secrecy operating to the disadvantage of the public, the two inventors had not taken out a patent. By 1900, the advantages of pure yeast culture regarding the safety of the fermentation process were widely recognized. Whereas formerly with the use of ordinary yeast the result was considerably subject to chance, the choice of a particular yeast determined by the desired type of beer and propagated by pure cultivation, in turn, enabled it to be produced. Although developed in connection with difficulties affecting the process of brewing bottom-fermented beers, Hansen's system found its way also into top-fermentation breweries on the continent. The new method, however, was not adopted in Britain and, as indicated by the title and stated in the opening paragraph, its reception and rejection by brewers in England at that time is the theme of the rest of this article.

Scientific R a m i f i c a t i o n s of H a n s e n ' s Work O u t s i d e Brewing: W. J o h a n n s e n and E. Fischer But before turning to it, it may be worthwhile to call attention to two consequences of Hansen's work outside brewing that appear to have been neglected by historians of science. The first case appertains to Wilhelm Johannsen's (1857-1927) concept of'pure lines', which he had developed early in the century in connection with his selection experiments on beans in order to find out whether variability was due to genetical or environmental factors. With its far-reaching effects strengthening the hand of those who

Mikuldf Teich

123

opposed the belief in the inheritance of acquired characteristics, we obviously cannot be concerned here. But what is of interest is that Johannsen had worked in the Chemical Department of the Carlsberg Laboratory as an assistant between 1881 and 1887. This was the period when Hansen was evolving and perfecting the practical and theoretical sides of his method for the preparation of pure-cultivated yeasts and it would be natural to presuppose a link between Hansen's endeavours in this area and Johannsen's work on pure lines. The intriguing thing is that Johannsen does not mention Hansen in the book where the pure line concept was elaborated: Ueber Erblichkeit in Populationen und in reinen Linien Ein Beitrag zur Beleuchtung schwebender Selektionsfragen (1903). However, he repeatedly pays tribute to the 'careful', 'comprehensive', 'laborious' investigations of Hansen, whom he regards as an 'excellent' researcher in Etemente der exakten Erblichkeitslehre (1909), the re-written and enlarged influential German version of Arwelighedslaerens elementer (1909). Very interestingly, it is here that Johannsen observed that whereas pure culture formed a basis of scientific microbiology, it had little impact on the study of heredity which came under the influence of statistics. In fact, Johannsen traced the pure line idea to Louis de Vilmorin (1816-60), the renowned French sugar beet breeder.9 The second example is more straightforward. It bears upon Emil Fischer's (1852-1919) researches culminating in the famous lock-and-key concept of interaction between enzyme and substrate. Familiar with the work of Pasteur and others, who demonstrated that micro-organisms were capable of acting specifically by consuming one or other optical antipodes, Fischer undertook to investigate whether yeasts also possessed specific fermentative activities. In this he was preceded and aided by Hansen, who supplied the German chemist and his co-worker H. Thierfelder (1858-1930) with eight different pure yeasts. Having altogether twelve species of yeast at their disposal, Fischer and Thierfelder obtained in 1894 an indication of a similarity between the stereochemical configuration of the fermentative agent (formed and made use of by the yeast cell) and the fermentable sugar.10 Following this, further work led Fischer to assume that specificity of enzyme action was determined stereochemically. In order to underline its restrictive character he employed the analogy that the interaction between enzyme and substrate depended on whether they correlated in the lock-and-key manner.11 H a n s e n ' s S y s t e m a n d E n g l i s h Brewing: Early T r i a l s It was natural that the news regarding Hansen's system of yeast culture aroused the curiosity of British brewers. On 9 July 1885, Horace Brown, who had previous contact with the Carlsberg Laboratory and especially with Hansen, wrote to Jacobsen requesting that his colleague, Dr George Harris Morris, be allowed to work with Hansen on pure yeast. The request was granted and from H. Brown's letter to Jacobsen on 30 August 1885 it is clear that Morris was well received at Copenhagen. It would appear that Hansen was contemplating leaving Carlsberg for an academic post at that

124

Yeast Cultivation and English Brewing, 1883-1913

time. Reacting to this report, H. Brown wrote in the same letter: Pure science will doubtless be the gainer by such a course, but to brewing technics such a loss would be a most severe one, for both here and on the continent our eyes are always turned upon the Carlsberg Laboratory and its accomplished botanist .. ,12 Meanwhile, having installed a yeast-propagating apparatus at Burton, Horace Brown and Morris began to search in earnest for ways and means of introducing pure yeast into English brewing. Embarking on this project they were, unquestionably, convinced that they would succeed and their belief was not shaken by the apparent lack of headway made through the years. This comes out distinctly in H. Brown's letters to Hansen. Thus on 13 June 1886 he stated: We are making further trials, and shall, I have no doubt, be able ultimately to obtain a yeast, or perhaps a mixture of pure yeasts, having the requisite qualities... I have not the least doubt that your discoveries will be of incalculable benefit to brewers in years to come. Almost three years later, on 5 February 1889, in connection with Hansen's planned visit to England, Horace Brown had this to say: I think on the question of the practicability of applying your beautiful methods to the English system of brewing we can give you more trustworthy information than anyone else, for I believe we are the only brewery which has attempted to carry out pure yeast culture in this country. I would suggest that you should come first of all to us. The brewery and laboratories shall be open to you and we will give you any information which is within our power. He then explained to Hansen that the conditions under which we have to work are totally different from those of the continent and that they are such as to render the application of pure yeast cultures much more different. Further, Horace Brown stressed that he and Morris have experimented in the brewery with many stocks of yeast, but hitherto we have not found one which, all things considered, has given us results as satisfactory as those attained with the ordinary mixed yeasts. Sensing that he perhaps might be misunderstood by Hansen, Horace Brown was at pains to assure the Danish scientist: Do not mistake me however. I think there is a greatfuture open to your methods in this country, but it will require long and patient work to adapt them to our requirements . . .

MikuldS Teich

125

Not the least absorbing parts of the letter are Brown's critical observations on the contemporary relations of science and technology in the brewing industry: We have unfortunately in this country, and especially amongst our brewing-technologists an immense amount of quackery and pseudoknowledge, which tends to retard very much the applications of science to our industry, and tends to bring our science into disrepute with our brewers, who, as a rule, cannot differentiate the true from the false . . . There are some indications at the present time of your methods falling into the hands of some English technologists of this class, and I am very afraid, knowing as I do the real difficulties of the case, that some delay in the recognition of your methods may accrue from this. We do not know whom Horace Browne had specifically in mind but the general thrust of his remarks was unmistakable. In the late 1880s, Horace Brown viewed British brewers, either on the commercial or the technological side of the industry, by and large as lacking the necessary background for proper appreciation of the benefits that could be gained from the adoption of new techniques based on science. But whatever the truth in this charge, Brown and Morris themselves, though most competent scientific brewers, continued to obtain disappointing results in their apparently solitary efforts to make the system of pure yeast cultures work in Britain. To quote from a well-known manual on brewing based upon a course of lectures given at the Finsbury Technical College by the highly-regarded Edward Ralph Moritz in 1889 (who in preparing it secured the co-operation of Morris): The practical application of pure yeast, has, we believe not yet obtained a footing in England . . . In this country, the question was taken up by Brown and Morris, who met with many difficulties in the application of this system to English brewing . . . Top-fermentation brewing as practised in England, differs in so many respects from that in vogue on the Continent and elsewhere, that it is impossible, from the satisfactory results said to have been obtained with pure yeast on the latter systems, to draw any conclusion as to the practicability of its application to the English system. This can only be decided by experiment; but should the difficulties alluded to above be overcome, there can be no doubt that the use of pure yeast would be attended with many advantages to the brewer.13 Trials with pure yeast cultures on a manufacturing scale were carried out by Brown and Morris for eight or nine years from 1885 to 1893 or 1894. They stopped after Morris left the Worthington brewery at Burton-onTrent but both investigators, curiously, appeared to be uncertain regarding the year when this actually happened.14 Despite negative results, there were a few individual brewing scientists in England who still felt like Horace's brother, Adrian J. Brown (1882-1910), who became the first holder of the Chair ofBrewing and Malting at the Mason University College, Birmingham

126

Yeast Cultivation and English Brewing, 1883-1913

(1899). On 28 September 1901, writing to Hansen of the reluctance to adopt pure yeast culture in Britain, he stated: . . . it is certainly not caused by want of good will, but due to the peculiar difficulties we meet with here in our breweries, and which so far we have not been able to conquer. Experiments, as you know, are being carried on at various breweries, but unfortunately we do not get all the advantage out of these we should, owing to trade secrecy, which cripples advance so much. This is deplorable of course but business and scientific research are in this country often very apart, I regret to say. You must always recognise that I personally have always most fully recognised the advantages we ought to derive from your noted system, and I never lose an opportunity to plead for experiment and investigation in connection with it. H a n s e n ' s S y s t e m and E n g l i s h Brewing: Early Objections It is noteworthy that shortly before this was written privately to Hansen, G.H. Morris decided to come into the open and explain publicly why he had abandoned his erstwhile favourable attitude to the possibilities of pure yeast employment in English brewing. He did it in the paper mentioned above, which he gave to a meeting of the Country Brewers' Society, the oldest English brewing trade organization, at Brewers' Hall on 8 May 1900.15 Morris held then the position of a Consulting Chemist to the Society and also the Lectureship in Technical Bacteriology at the Jenner Institute. He was prompted to take the step in response to a recently-expressed view by Hansen regarding the divergence of opinion in England on this issue and the lack of solid data which could help to resolve it. From Morris' paper and the discussion that followed, it emerged that the central problem was whether pure yeast cultures would perform the secondary as well as the primary fermentation. Morris emphasized that he did not oppose the use of pure yeast as such and admitted that Hansen's system was successfully operated on the continent to produce high-quality top- and bottom-fermented beers. Morris claimed that this was because the continental culture yeasts were not wholly eliminating the fermentable sugars during the primary fermentation. Under these circumstances, pure culture yeasts were able to act upon the residual unfermented matter during the secondary fermentation. According to Morris, this did not apply to the majority of English beers. In their case, he stated: . . . the amount of matter so left unfermented is not considerable, and I doubt very much whether, in any but the quickest running ales, this readily fermentable matter is sufficient to provide the after-fermentation which we find necessary. As a result, the beers Brown and Morris produced by means of culture yeasts were 'as a rule, sound and good in flavour, but remained thin and flat'. Secondary fermentation was induced only 'by the addition either of cold water malt-extract or of ordinary sugar priming'.

Mikulat Teich

127

About the same time, in 1902, a similar view was echoed by Franz Schonfeld, the leading German scientific specialist on top fermentation. He pointed to the difference existing between brewery top yeasts employed in Germany and England with respect to the degree of attenuation secured by them. For the most part, top yeasts in German breweries were yielding low-attenuated and in English breweries high-attenuated beers. As to the amount of alcohol, as a rule the German top-fermented beers contained less than their English counterparts but also, interestingly, less than the German lager beers. Schonfeld observed that under English conditions, pure culture yeasts lost the power to ferment after essentially having fermented out the fermentable material during primary fermentation. With their fermenting power in decline, pure yeasts were hardly capable of inducing a fermentation in racked beer ('first after-fermentation') and incapable of participating in the maturation of bottled beers ('second after-fermentation'). Schonfeld suggested: This was then the crux of the matter and decisive reason which brought about the downfall of pure yeast in England. Pure yeast was ineffective to actuate a second after-fermentation in bottled beers. Because such beers lacking after-fermentation and without sufficient reproduction of carbon dioxide (though from our point of view a weak one) are unsaleable, the fate of pure yeast thereby was once for all sealed.16 But the disapproval of pure yeasts because they could not bring about secondary fermentation believed to be essential to the production of better English beers was apparently not universal. Commenting on this, Julian L. Baker, one of the leading English brewing experts of the time, expressed himself as follows in 1905: This objection has not yet been fully proved, but if it be admitted there is no reason why pure yeasts should not be used for fermenting running beers where there is no true secondary fermentation. He furthermore had little doubt that Hansen's system would eventually be adapted 'to meet the peculiar conditions in English breweries'.17 The lack of success in wedding Hansen's methods to English fermentation systems was frequently explained by their distinctiveness. What was meant by this connotation appears to be less clear. There were some, like Baker, who associated it with the relatively high temperature of English top-fermentation systems. He also supposed that high temperatures favoured the development of wild yeasts and deteriorated and weakened pure yeasts, but the idea that high temperatures in themselves constituted the distinguishing features of English top-fermentation practice was questionable. In truth, there is no evidence for it because the range of temperatures at which top fermentations in English and continental breweries were conducted were substantially the same. Indeed, the yeast type and performance rather than temperature appeared to provide for others the clue to differences between the English and continental systems of fermentation, including the failure to domesticate

128

Yeast Cultivation and English Brewing, 1883-1913

Hansen's system in England. However, it is difficult to avoid the impression that at times the explanation of why brewers in England felt no need to resort to Hansen's method of isolation of pure yeasts was circuitous. They were thought to give beers with flavours that were unlike the ones that suited the English palate. Since for this purpose mixtures of brewery yeasts, it was held, performed adequately there was no point in replacing them by pure cultures. The following passage from a book on brewing by the President of the Institute of Brewing, A. C. Chapman, published in 1912, reflected this approach: A good many attempts to introduce the use of selected single-cell yeast into English breweries have, however, met with much less success. One reason for this is that the conditions obtaining in most English breweries are such as to result in the production of a definite type of yeast which gives the exact class of beer required, and which can without any special steps be kept practically pure, that is free from bacteria and other yeasts within the limits necessary for successful working. In the second place, there is a greater difference in character between the main fermentation and the secondary or cask fermentation in English high fermentation beers, than is the case in lager beers such as are brewed on the Continent; and it has not hitherto been found possible to obtain with a single-cell yeast the proper cask fermentation which is so important a feature in English brewing.18 In this connection a source of dissatisfaction to English brewers, identified by Horace Brown in his lecture in 1916, which made them wary to adopt pure culture yeast should be mentioned. Contrary to Hansen's findings at Copenhagen, Brown observed at Burton that English beers after the completion of the primary fermentation with pure culture yeasts were susceptible to turbidity. He ascribed the cause of the defect to wild yeasts lodged in the brewery,19 as by its very nature pure culture yeast servicing as pitching yeast was thought to be devoid of contamination. Experience of this sort of difficulty with a pure system actuated English brewers to persist in employing traditional top-fermentation systems in which wild yeasts as secondary yeasts were thought to have to play an integral part in endowing the beers with distinct flavours which were in demand.20 H a n s e n ' s S y s t e m a n d E n g l i s h Brewing: the E c o n o m i c Dimension What has emerged is that when English brewers voiced their distrust of Hansen's system they pointed out that it did not allow the production of beers suitable for the domestic sense of taste, by virtue of the fermentation temperature and qualities of yeast essential to it. Needless to say, there were other elements in addition to those mentioned, which acted together to put a break on its adoption in England. Clearly, it is beyond the scope of this article to attempt to deal with them comprehensively even if one could be certain of what they were. But there is one element we know of, the economic

Mikul&l Teich

129

one, which cannot be omitted from the story because of its close relation to the previously considered scientific, technical and subjective (i.e. the taste of the consumer) factors. These implied the existence of an economic dimension but, curiously, it was not referred to more directly by those who explored, but became critical of, the possibilities of using culture yeasts in relation to English beers. At this point, Pasteur's views comparing top- and bottom-fermentation systems deserve to be brought up. It should be remembered that the result of the Franco-Prussian War inspired Pasteur to improve French brewing practice on the basis of scientific research, with the aim of 'benefiting a branch of industry wherein we are undoubtedly surpassed by Germany'. 21 Pursuing this objective, he also sought to acquaint himself with the English brewing practice when he visited London in September 1872. By then, in continental but not in British breweries, the traditional top-fermentation systems were on the way to being squeezed out by bottom-fermentation systems generally derived from the process developed in Bavarian lager brewing. According to Pasteur, this change was not due to a qualitative superiority of bottom-fermented beers, as such, over top-fermented beers but to the brewing industry's need for 'more stability and uniformity, both in the production and the sale of goods'.22 It was a scientist's concise but remarkable apt identification, in economic terms, of the motive that impelled the continental brewers to industrialize. The transition from top fermentation to bottom fermentation, the beginnings of which in Northern Germany, Austria, Bohemia and Denmark may be traced to the 1840s, constituted the actual path towards 'industrial maturity in brewing'23 on the continent of Europe. It was through the success of this change-over that brewing in certain continental countries became industrialized and was effectively in a position to catch up with and compete with the brewing industry in England. What it amounted to, in reality, was the transformation of small-scale brewing into large-scale brewing that had occurred in England much earlier, between 1700 and 1830. Here it was brought into being by developments that included porter brewing, the adoption of the thermometer and hydrometer, the use of steam power and the mechanization of most brewing operations.24 Integral to the production of bottom-fermentation beers is the maintenance of low temperatures during the process of fermentation and their storage. During the early 1870s this was still achieved by means of natural ice, but then the large lager breweries on the continent began to experience difficulties in securing the necessary amounts of ice because of the uncertainty of climate and the problem of finding enough space for storage. To indicate the size of the problem: within two decades between 1846 and 1869, the amount of natural ice required by the Spaten brewery rose dramatically from about 10 kg to about 72 kg per 1 hi of sold beer. At present it is not possible to give a considered answer to the question whether the brewing industry in England was scientifically, technically and economically in a backward or progressive state in the period under review.

130

Yeast Cultivation and English Brewing, 1883-1913

In general, the relations between scientific, technical and economic factors are complex.25 But whatever the answer, it is fair to say that economic considerations must have weighed greatly if not finally with the English brewer against his changing to bottom-fermentation brewing when he was capable of maintaining the quality and regularity of his product by sticking to one of the traditional top-fermentation systems, which did not call for such low temperatures. So far as cellar temperature is concerned, about 1213°C were demanded in an English top-fermentation brewery and 5-6°C in a continental bottom-fermentation plant in the 1870s. While with the former an evenness in the desired range could be managed without ice, this did not apply to the latter. Under these circumstances it is not difficult to perceive why English brewers preferred to persevere with practised top-fermentation systems rather than to interest themselves in unfamiliar bottom-fermentation operations. To quote Pasteur: The fact that English brewers have not as yet adopted 'low fermentation' may be accounted for, in a great measure, by the difficulty of enlarging existing breweries in cities like London, to the extent required by the new method of manufacture. Even in the event of public taste demanding a 'low beer', English brewers will hesitate a long time before converting their breweries. Such conversion would impose upon them expenses and difficulties of a very serious nature. If ever such a change should take place, it will probably be inaugurated out of London. It is, however, worthy of remark that English brewers, without adopting 'low fermentation', have introduced considerable improvements in brewing, especially in the management of the temperature during fermentation; this must be preserved within narrow and exact limits, for fear of injury to the product. It might easily be shown that these improvements have resulted from the liability of the beer to contract diseases, although this fact may not have been recognized by the brewers who have introduced them.26 For bottom-fermentation brewing, the solution to temperature control lay with the adoption of the artificial refrigeration system, theoretically devised and experimentally tested by Carl (von) Linde during the 1870s. At first, brewers of lager beer were very hesitant to install an artificial refrigeration system into the localities containing vats for the fermentation and storage of beer. Finally, shortage of space forced some of the large breweries to try out the Linde system with great success in the early 1880s, which fostered its further spread. Conclusion Artificial refrigeration and pure yeast culture vastly increased the safety of bottom fermentation. Together with the steam engine, these three scientifically-based techniques effected the completion of industrialization of brewing in Germany and presumably in other continental countries within the three decades between 1883 and 1913. (It was the installation of

MikuldS Teich

131

artificial refrigeration which changed the scale of the use of steam power in lager breweries.) We know virtually nothing about the demand for artificial refrigeration in the brewing industry in England during the period under consideration. It may be presumed that it was not part of the normal order of things in a top-fermentation brewery, which in this respect was cheaper to run.27 As to Hansen's method of culturing yeast, English brewers had not switched to it for a combination of economic, scientific, technical and physiologico-psychological reasons during the period under review. Although there were endeavours, especially on the part of certain Danish specialists, to suggest procedures expected to recommend themselves to English brewers, nothing tangible materialized. Particularly active in this direction was Alfred Jorgensen, the Director of the Laboratory for the Physiology and Technology of Fermentation at Copenhagen, whose publications translated into English appeared in several editions. Thus in 1913 a new English edition of his work The Practical Management of Pure Yeast was published and reviewed by Julian L. Baker. Thirty years after Hansen's method was adopted by a good many breweries on the European continent and after it spread to America, Asia and Australia, Baker described the situation in England as follows: The employment of pure races of yeast in our English breweries may be said to be practically non-existent. This is no doubt due to the exaggerated claims which some followers of the late Professor Hansen put forward. The caution of our brewers was shaken by the failures of the early trials of the pure yeast system and the recovery has yet to take place. But this observation may be contrasted with the subsequent opinion which Baker offered pursuing the theme: Although the conditions affecting stability, brightness, flavour, etc., of Continental bottom-fermentation beers differ so greatly from the English top-fermentation beers, the reviewer from his own experience is still convinced that the general principles laid down by Hansen will in the end obtain. Unfortunately an English brewery is not the place for experiment; still, much can be done by the rational use of pure yeast cultures in mild ale and porter fermentations.28 In fact, it took two more decades before renewed attempts were made to develop culture yeast (possessing a highly attenuative character) though not from a single cell but from a number of individuals.29 Clearly, on the main point the proposed procedure differed from the method as originally conceived of by Hansen. Notes 1. Where I specifically do not acknowledge references regarding German and general brewing history, as here and elsewhere in the text, I draw on my Bier, Wissenschaft und Wirtschqft in Deutschland 1800-1914, to be completed shortly.

132

Yeast Cultivation and English Brewing, 1883-1913

2. See letter by J.C. Jacobsen to Gabriel Sedlmayr Jr., the owner of Spaten brewery, dated 7 May 1884. It was published in Die Hefereinzucht in der Entwicklungsgeschichte der Brauerei, Berlin, 1931, pp. 72-4. Consult also for further quotations. 3. J. Pedersen, The Carlsberg Foundation, Copenhagen, 1956, pp. 10-11; H. Holter, 'An outline of the history of the two departments' in H. Holter and K.M. Moller (eds), The Carlsberg Laboratory 1876/1976, Copenhagen, 1976, p. 24. 4. Quoted by B. Trolle, 'The origins of the laboratory and its background in the Carlsberg Laboratory' in The Carlsberg Laboratory, as note 3, pp. 17-18. 5. F. Sedlmayr, Die Geschichte der Spatenbrauerei unter Gabriel Sedlmayr dem Alteren unddem Jiingeren 1807-1874 sowie Beitrage zur bayerischen Braugeschichte dieser £eit, Nuremberg, 1951,11, pp. 130,55. 6. H.T. Brown, 'Reminiscences of fifty years' experience of the application of scientific method for brewing practice' in Journal of the Institute of Brewing, Vol. 22, 1916, pp. 267-348. The reference to seasonal brewing occurs on p. 284. 7. E.C. Hansen, Practical Studies in Fermentation Being Contributions to the Life History of Micro-Organisms, London and New York, 1896. It constitutes the English edition revised by the author of his Untersuchungen aus der Praxis der Garungsindustrie, first published in 2 parts, in 1888 and 1892. I have drawn on the English volume in this section. 8. Hansen's italics, see Practical Studies in Fermentation, as note 7, p. 155. 9. This point is also made in private correspondence to me by Nils Roll-Hansen, who worked on Johannsen's genotype theory and its relation to plant breeding and the study of evolution. See his 'The genotype theory of Wilhelm Johannsen and its relation to plant breeding and the study of evolution' in Centaurus, Vol. 22, 1978/9, pp. 201-35. He writes among others (11 June 1982): 'Johannsen certainly perceived an analogy between his own "pure line" and Hansen's "pure cultures". But how much of the inspiration came that way is difficult to judge.' 10. E. Fischer and H. Thierfelder, 'Verhalten der verschiedenen Zucker gegen reine Hefe' in Berichte der Deutschen chemischen Gesellschaft, Vol. 27, 1894, pp. 2031-7. 11. E. Fischer, 'Einfluss der Configuration auf die Wirkung der Enzyme' in Berichte der Deutschen chemischen Gesellschaft, Vol. 27, 1894, pp. 2985-93. 12. This and the following passages are from letters in 'E.C. Hansen Correspondence' in The Royal Library (Manuscript Division), Copenhagen. 13. R.E. Moritz and G.H. Morris, A Text-Book of the Science of Brewing, London and New York, 1891, pp. 336-7. How well the volume was thought of can be judged from the fact that at the instigation of the influential brewing research institution Versuchs-und Lehranstaltfiir Brauerei in Berlin it was translated into German by one of its prominent associates W. Windisch and published under the title Handbuch der Brauwissenschaft, Berlin, 1893. 14. Brown, 'Reminiscences', as note 6, p. 317; G.H. Morris, 'Some experiences in the use of pure cultivated yeast' in Journal of the Institute of Brewing, Vol. 6, 1900, p. 335. 15. Morris, as note 14, pp. 333-49. For discussion, see pp. 349-53. The two quotations are taken from pp. 347, 345. 16. F. Schonfeld, Die Herstellung obergahriger Biere, Berlin, 1902, p. 140. 17. J.L. Baker, The Brewing Industry, London, 1905, p. 114. 18. A.C. Chapman, Brewing, Cambridge, 1912, p. 89. 19. Brown, 'Reminiscences', as note 6, pp. 316-18. 20. For a brief informative treatment of a hundred years of yeast culture in Britain, see N.S. Curtis, 'A century of yeast culture' in Brewers' Guardian Centenary Issue, 1971, pp. 95-100.

MikuldS Teich

133

21. L. Pasteur, Studies on Fermentation: The Diseases of Beer, Their Causes, and the Means ofPreventing Them, London, 1879, p. vii. This is the English translation of Etudes sur la biere, published in 1876. 22. Ibid., p. 13. 23. P. Mathias, The Brewing Industry in England, Cambridge, 1959, p. 213. 24. Ibid., p. xxviii and Chap. III. 25. For an account denying a 'revolutionary' impact of science upon the British brewing industry during the second half of the nineteenth century, see E.M. Sigsworth, 'Science and the brewing industry 1850-1900' in The Economic History Review, Vol. 17, 1965, pp. 536-50. For a different perspective of the consequence of scientific activities on brewing from Pasteur to the mid-1930s, see H.L. Hind, 'Pasteur to 1936 An account of the development of science in brewing' in Journal of the Society of Chemical Industry, Vol. 56 1937, pp. 125-33. 26. Pasteur, Studies on Fermentation, as note 21, p. 14. 27. H.S. Corran, A History of Brewing, Newton Abbot, 1975. This is a generally informative account but not so on this topic. Cf. pp. 196-203, 227. 28. J.L. Baker, 'The practical management of pure yeast. By Alfred Jorgensen' in Journal of the Institute of Brewing, Vol. 19, 1913, pp. 515-16 (Review). 29. B.M. Brown, 'The production of pure yeast' in Journal of the Institute of Brewing, Vol. 40, 1934, pp. 9-13. For discussion, pp. 13-16.

E d u c a t i o n t h e

a n d

T e c h n o l o g y

I n d u s t r i a l GEORGE

i n

R e v o l u t i o n

TIMMONS

Every advanced industrialized country invests heavily in education because its economic future depends not only on a fairly high level of general education throughout the entire population but also on an extensive and deep spread of technical skills of all sorts. Educational provision and development cannot be left to chance because it is too important, and this has probably always been the case. In discussing economic growth, R. M. Hartwell1 quotes in a footnote S. Kuznets who argues that 'the sustained growth of population and product was made possible by the increasing stock of tested knowledge' and 'one might define modern economic growth as the spread of production . . . based upon the increased application of science, that is, an organized system of tested knowledge'. This implies that both science and education were important to the Industrial Revolution itself and that the 'uneducated empiricism' of a few brilliant amateurs cannot explain fully the profound technological developments which occurred in the eighteenth and nineteenth centuries. One thing is certain: the rapid rise in the economic strength of countries such as Germany and Japan depended, in part at least, on the conscious development of their education systems. Furthermore, as early as 1851, it was becoming obvious that, despite its initial success, British industry would face severe difficulties in dealing with the gathering threat of competition from foreign rivals and certain men, such as Lyon Playfair and Prince Albert, recognized that education, and in particular technical and science education, was a key factor in the developing crisis. Though efforts were made to improve and extend technical and science education from that time onwards, the improvements were of such a kind and carried out in such a way that even now our education system does not serve, as well as it should, our needs as an industrial nation. It is almost as though the true significance of education has never been realized in Britain because, whether nineteenth-century elementary schools are contrasted with public schools, or twentieth-century secondary moderns are compared with grammar schools, social control rather than industrial growth seems to have been more important. And yet if Britain was the first country in which industrial take-off occurred, then the influence of education on technology should have been apparent earlier. Crouzet says that A study of this question [i.e why Britain took the lead in industrialization] which also took education into account would be very useful. In Britain, dissenters and Scotsmen played a major role in the industrial revolution because, thanks to the dissenting academies and the much

136

Education and Technology in the Industrial Revolution more advanced Scottish educational system they were the best educated section of the middle class. French business men (on the other hand) had either a very elementary or purely classical education.2

Hartwell devotes a section to education and he says that Modern theorists of economic growth argue that education to-day is important in determining the pace and pattern of growth; and historians tell us that education was important for the industrializing nations of the nineteenth century. It is reasonable to assume, therefore, that education was also important for the economic growth of Britain in the eighteenth century, before and during the industrial revolution.3 However, what is found is disagreement on the part of economic historians, historians of science and historians of technology about the relationship between developments in science, technology, education and industry from 1600 to 1800. Authorities such as Musson and Robinson4 accept that something like a scientific revolution had taken place before the Industrial Revolution had got under way and that both the knowledge which it made available and the methods which it promulgated were important to technical developments and thus to industrial production. On the other hand, Eric Ashby5 maintains that the work of 'cultivators of science' . . . was not regarded as having much bearing on education and still less on technology, and A.R. Hall6 says that literacy and learning had little to do with technologyIt could be that one of the reasons why such disagreements exist is that science and technological innovation were not contrasted in any truly distinctive way before (say) 1825. R. A. Buchanan7 says that at the time when science and technology begin to become recognizably modern, they were virtually indistinguishable. They were practised by the same sorts of men with almost indentical objectives. In some cases they were united in the same personality as in that of James Watt. In Britain, they were products of the new society—merchants, Quaker businessmen, non-conformist ministers, Latitudinarian churchmen, Whigs and Tories, captains of East Indiamen, civil servants, military and naval officers, and aristocrats and academics with practical occupations. Of none of them would it have been possible to have said that they were 'scientists' or 'technologists' in the modern professional sense. In such circumstances, where science and technology are all of a piece, it is difficult to assign to science, in any precise way, the role it actually played in the early stages of industrialization, especially because those who introduced technical changes were not always aware of the importance of science to what they were doing. Unfortunately, this also means that the part played by education was not only obscured at that time but is still difficult to delineate now. Yet education must have played a part, both directly and indirectly, in the diffusion of scientific knowledge and it was certainly

George Timmons

137

available in a wide variety of forms. Furthermore, that willingness to change which encouraged technical innovation, and which is so discernible in the second half of the eighteenth century, must have derived some of its impetus from education. However, because it is difficult to see direct effects of education on technology, it is necessary first to make clear the influence of science on technology and then the influence of education on science and, where possible, on technology. Musson and Robinson8 say that Dr John Wallis remarked as early as 1697: 'Mathematics at that time, with us, were scarce looked upon as Academical Studies, but rather Mechanical; as the business of Traders, Merchants, Seamen, Carpenters, Surveyors of Lands or the like.' They also show that there were many instances of collaboration between the professors at Gresham College and naval officers and administrators as well as shipbuilders, at a very early stage. They maintain that John Moxon's works were not only meant for members of the Royal Society but also for literate craftsmen and that the scientific spirit in his work on geography, navigation, astronomy, mathematics and architecture, on the one hand, and 'Smithing, Founding, Drawing, Joynery, Turning, Engraving and Printing', on the other, was percolating down to craftsmen. J. A. Chaldecott9 shows in an extract from a letter from Wedgwood to Bentley that the potter*s chemical knowledge was used to improve his industrial techniques and that he understood the importance of mathematics to the shaping of mortars and pestles. But perhaps the most striking example lies in the influence of Black's research into latent heat on Watt's improvements to the steam engine. This is still a matter of controversy, but towards the end of his life Watt wrote in a letter to Brewster: 'Although Dr. Black's theory of latent heat did not suggest my improvements on the steam engine, yet the knowledge upon various subjects which he was pleased to communicate to me, and the correct modes of reasoning and making experiments of which he set me the example, certainly conduced very much to facilitate the progress of my inventions.'10 Here, then, is an example of an academic affecting the way an engineer thought about problems rather than merely handing on useful data to him. Moreover, Robison, another of Watt's associates in Scotland, makes a claim not only for the influence of science, but also for education, on Watt's work. He says: T had had the advantage of a more regular education: this frequently enabled me to direct or confirm Mr. Watt's speculations, and put into a systematic form the random suggestions of his inquisitive and inventive mind.' 11 However, we should recognize that this was a two-way process: Wedgwood not only sought Priestley's opinions, he also offered him advice,12 and D.S.L. Cardwell13 says that 'the doctrines of energy and laws of thermodynamics carry with them unmistakeable signs of their technological origins', i.e. Watt influenced Black. That the relationship between science and technology was important was understood at the time. Engineers like John Grundy and Charles Labelye, though very experienced in practical engineering, stressed the importance of mathematics and philosophy and also insisted that their work was both a science and an art. In the same way, the importance of chemistry to

138

Education and Technology in the Industrial Revolution

practical affairs can be illustrated by the work of Shaw and Hawksbee who in 1731 gave 'A Course of Chemical Experiments with a view to Practical Philosophy, Arts, Trades and Business'. These were published in 1734 and the authors said 'Our Endeavour has been to improve the useful Arts, by means of a more Philosophical Chemistry and at the same time shew the Method of conducting enquiries, so that they may terminate in useful Discoveries'.14 Mathias points out that, when looking at this problem of how extensive was the diffusion of science and its influence on technology, a distinction ought to be drawn between the intention of men such as Shaw and Hawksbee and their direct effect on practical affairs.15 But links can be found. For instance, Robert Dossie acted as a consultant chemist, wrote articles useful to manufacturers, was a member of the Society of Arts but, most important, had direct links with the toy industry where his discoveries in metallurgy and in certain finishing processes were aids to the capturing of markets from the French. His brother-in-law, a Manchester dyer, was also affected by his work. What is more, the first real firm of industrial chemists in this country, Godfrey and Cooke, was founded at the very beginning of the eighteenth century by Godfrey Hanckwitz, who was a consultant analyst for the Royal Society. Anyway, the intentions of these men were important even when they had little direct influence because their intentions show that they wished to make scientific principles available to manufacturers, entrepreneurs and artisans. They also contributed to the ethos in which the Industrial Revolution took place. Moreover, many of their ideas, findings and their organization of data (e.g. Beigh ton's table of steam engine powers and proportions) were of direct use and were used. For that reason, they should have been taken seriously both at the time and later. Unfortunately just how vital science was to developments in technology was grasped by too few people at that time. Desaguliers expressed the hope that engine makers would acquire mathematical skill and 'not attempt impossibilities',16 and Beighton, writing in the Ladies Diary in 1721, drew attention to the dangers of guesswork in engineering, particularly when it succeeded.17 This suggests that both men were well aware of the very necessary connection between science and industry but that they were concerned that too few people were prepared to cultivate it. The fact that the Royal Society became less interested in the practical applications of natural philosophy as the eighteenth century wore on is also indicative of a lack of appreciation of the interconnectedness of science and technology. On the other hand, the Royal Society of Arts was established to encourage just such a connection. This seems to be the crux of the problem: there were forces working counter to each other at the time, but because the changes which were taking place were so rapid and far-reaching, the true value of science, and therefore education, was not always recognized. Yet even if we take an industry such as cotton where early mechanization was not so influenced by science, Crompton being the only one of a host of inventors to have any great depth of theoretical knowledge, we still find the links. In a letter to his father-in-law, dated 30 October 1784, James Watt wrote:

George Timmons

139

As to Mr. Arkwright, he is to say no worse one of the most self sufficient ignorant men I have ever met with. Yet . . . whosoever invented the Spinning, Arkwright certainly had the merit of performing the most difficult part, which was the making it useful. Some years ago he applied to us at two different times for our advice which we took the trouble to give him, in one or more big letters, which he never had the j manners to answer.18 However, perhaps the ill manners of Arkwright are significant. G.F. French19 said, in 1859, Trained to a servile handicraft, and without a shilling of capital, the position from which he raised his fortunes had not one of the advantages enjoyed by Crompton; but to compensate for this he possessed an indomitable energy of purpose which no obstacle could successfully oppose, a bronzed assurance that enabled him unabashed to meet and to thrust aside either circumstances or men when they stood in his way, an unscrupulous hand to grasp and appropriate the ideas and immatured inventions oj others, a rude health that enabled him to work or travel when others slept, and an undaunted spirit for speculation, prepared to accept success or failure without any visible effect on his mind or his temper. So many entrepreneurs, with their insatiable urge to make money, forgot or found it inconvenient to recognize the debt which they owed to the 'philosophers'. This was obviously not the case with men like Josiah Wedgwood and Matthew Boulton in the Black Country, Peter Ewart and George Lee in Manchester, or Benjamin Gott in Yorkshire. But how many more were there of whom we know nothing or very little? How many were like Joseph Rogerson? He worked so hard in his own slubbing mill that he had little time for anything else. For those first four years of this mill of ours running I have seen us begin almost every morning sometimes at 5 and 6 o'clock in the morning and also seen the fires put at Night when we gave over, which was at 8, 9 and 10 o'clock at night; and we have generally run later in Winter than in Summer, in Winter frequently all night. I have had to go home for Breakfast—Dinner—and Drinking and I generally had got back before they had got theirs, time we allotted Slubbers for Dinner was 1 hour.20 There was no time left for reading or attending lectures or engaging in discussion with other mill owners in such a programme. There must have been many hard-working men like Rogerson who never bothered their heads with theory, and their very success hid the true nature of the changes taking place and made it appear as though science and education stood for little. This was not the case, despite the eminence of uneducated men like Stephenson and Bramah (who prove to be exceptions anyway), because science had already penetrated technology—most crucially in the production and use of power. As we saw earlier, even an industry which appears at

140

Education and Technology in the Industrial Revolution

first sight to have been transformed by ingenious amateurs (i.e. textiles) became a modern industry because of the way power was applied to it: its early development depended as much on Smeaton's improvements to the efficiency of the waterwheel as on the machinations of Arkwright, and Cardwell21 says that Smeaton's approach was scientific and based on Galilean mechanics—as was also the building of factories using iron beams to bear the weight of the new machines. Moreover, the industry's greatest expansion came when it adopted the steam engine, which became efficient enough only when scientific principles influenced its construction. Such penetration of technology by science was likely to occur only if the dissemination of science was wide enough, and that dissemination depended on education as much as on anything else. Kelly,22 writing about the period before 'mechanics institutes were introduced, said that 'for the humbler occupations, even in the new machine age, illiteracy was still no bar, but for the supervisory grades, it was necessary to have men who could read instructions, could follow drawings and had at least a smattering of scientific knowledge'. However, it is doubtful whether such men would have been available, even in the restricted numbers in which they were found, if the level of literacy in Britain had not been higher than has sometimes been suggested. There is now a considerable amount of evidence to show that basic literacy increased between 1750 and 1850, the crucial period of industrial development. Various studies made of the marriage registers, which all parishes had to keep after Hardwick's Marriage Act of 1753, suggest that about 50 per cent of those married in the 1750s could have been literate and that this had risen to abut 70 per cent by 1850. Because the evidence is based on signatures, this does not necessarily mean that the participants were functionally literate but, on the other hand, inability to sign does not necessarily imply inability to read. It is clear that the level of literacy was fairly high in the mid-eighteenth century and that it improved. There is also evidence of a more indirect kind. For example, V. Neuburg's study of popular education in the eighteenth century23 shows that the number of chap-books printed increased rapidly. These were not very advanced forms of literature but they gave 'the lower orders' what they wanted to read: lively tales of romance and violence. Moreover, the number of newspapers increased, as did the number of bookshops and booksellers. But perhaps what is more important is that the desire to read seems to have increased as the century wore on. The best evidence for this is the popularity of the Sunday schools when they began to open in large numbers towards the end of the century. Despite their methods and their reading matter, children flocked to them. Frank Smith wrote in 1930: Far more surprising is the evidence of the Sunday Schools as to what children read in the early days. One illustration will suffice. The catalogue of a large Sunday School in Newcastle-on-Tyne in 1815 contains 339 volumes, the bulk of which have titles like the following: The Religious Tradesman, Mrs. Rowe's Devout Exercises, Solomon's Temple Spiritualized, the Art of Divine Contentment, Mrs. Rowe's

George Timmons

141

Friendship in Death, Essay on the Usefulness of Oriental Learning, Milk For Babes (a catechism in verse), Precious Remedies for Satan's Devices, Sighs from Hell. Yet the report of 1816 states that the library 'continues to be resorted to by the children with increasing avidity'. It is by such illustration that we can best realise the passion for reading which the Sunday Schools met and fostered.24 Some of these schools had to open during the week and for adults as well as children. The Sunday schools were really only a variation on the much older Charity schools which, though past their best as early as 1730, had provided for the education of the lower orders throughout the eighteenth century. M.G. Jones, the principal authority on these schools, probably over-estimated the number of children who attended them but even so their contribution to literacy was considerable. Added to these were the schools of industry which, though similar to the Charity schools, were less successful and more ephemeral. There were also large numbers of Dame schools. It is very difficult to assess their contribution because it is impossible to discover how many there were or how effective they were, but if references to them in literature are considered,25 then they were many and some taught reading and even writing. They cannot be dismissed as mere 'baby-minding' institutions though that is what many of them were with the dames often as unlettered as their charges. There were also large numbers of'adventure' or 'common day' schools, though again what has been said of the Dame schools would also be true of them and Joseph Lancaster referred to the teachers in them as 'the refuse' of the other schools. But again there are so many references to them in the writings of the period, and so many references which imply that their work could be good, that they cannot be dismissed as totally ineffective. Obviously, the Charity schools and the Sunday schools were the most important institutions of education for the poor, not so much because of the numbers they taught but because it was from them that the nineteenth-century elementary system grew: Joseph Lancaster's first school was really only a Sunday school which opened on weekdays. Unfortunately, this stigmatized elementary education at its very inception with the stamp of inferiority. Daniel Defoe claimed that the original Charity schools were meant to ensure the continuance of'the great law of subordination'.26 This was to have important consequences later. However, at this point, suffice it to say that the population of Britain at the beginning of the Industrial Revolution was literate enough to assist in that process. Besides these basic forms of elementary education, there were in Britain in the eighteenth century forms of what we would now call secondary and higher education which were sufficiently well developed to cater for the needs of increasing industrialization. The picture of almost unrelieved decline in education which used to be painted of this period is now seen as largely inaccurate. Though the nine great schools were places avoided by conscientious parents because of their barbarism, and though many other endowed grammar schools suffered declining fortunes, some even disappearing completely or surviving only as sinecures,27 the total picture was far

142

Education and Technology in the Industrial Revolution

from desperate. Nicholas Hans 28 shows that the nine great schools educated 22 per cent of the intellectual elite of the eighteenth century and that Eton and Winchester, in particular, made an important contribution. Moreover, all of them introduced some science into the curriculum if only in the form of optional lectures by scientists from outside. Science was not seen as inappropriate to the education of the elite until well into the nineteenth century and it even had a place in the education of the future heirs to the throne: Hurde, the tutor to future George IV, included mathematics and natural philosophy in the course. On the future William IV, M.L. Clarke29 says that 'in 1780 he was studying astronomy . . . in the same year began optics and in the following mechanics'. The universities of Oxford and Cambridge had not sunk quite so low as is sometimes suggested. Hans points out that it was only after 1760 that they reached their lowest state. Before that, the chairs of science were rarely sinecures: the professors did lecture. Moreover, most students did some work and the group of aristocratic rowdies which brought down such opprobium was in fact small. On the other hand, most students did not take much interest in science, though Daniel Waterland's Advice to a Young Student was published in 1740 at Cambridge and then at Oxford in 1745 and seems to have been used: it assumed that mathematics and natural philosphy would be studied and included textbooks in these subjects in its bibliography. Christopher Wordsworth also claimed that the best tutors used Waterland.30 Nevertheless, the influence of the English univesities on developments in science and technology must have been rather slight and appointments to science chairs after 1760 were part of university politics. Watson, appointed professor of chemistry at Cambridge, knew nothing of the subject but was exceptional in that he did take the trouble to make himself proficient as a lecturer in it. Furthermore, McKendrick31 maintains that Watson's chemical essays 'did a great deal to bring to the notice of a wider public the new techniques of careful qualitative experiment that we find in the work of Black, Cavendish and Lavoisier' and that many of Watson's own experiments were concerned with the application of science to manufacturing. Watson also proposed the setting up of an Academy of Applied Chemistry because he did understand that his subject could have a direct effect on the technologies of several eighteenth-century industries. Watson may have been an exception in England, but in Scotland such attitudes as his to the industrial applications of science were not out of place, for the universities there helped to provide an ethos in which inventors such as Watt could flourish. A.L. Donovan32 says that Watt and Black 'had the good fortune to be born into a culture which did not elevate the philosopher and demean the mechanician . . . [where] the universities were committed to the development of those systems of political and social thought and those areas of specialised knowledge upon which the future growth and guidance of the nation depended'. Watt was only one of many Scotsmen who made his fortune in England. Furthermore, English parents who desired a 'useful' education for their sons sometimes sent them to Scottish universities, or perhaps abroad to Leyden. When we turn to the endowed grammar schools, we find again that their

George Timmons

143

plight was not as desperate as once thought. Their fortunes were affected by the changing value of their endowments, but what seems to have been more important to them individually was the calibre of those appointed as masters in them. Those situated in the more important towns usually prospered because of the demands put upon them by the merchant class and because of the quality of the masters appointed to them. Mathematics and science were to be found in curricula of many of these schools. The Royal Mathematical School, established as a department of Christ's Hospital School during the reign of Charles II, prospered throughout the eighteenth century and exerted considerable influence, especially on the navy. One of its former pupils, James Jurin, was later master at Newcastle Grammar School to which he introduced mathematics and science. The grammar schools at Dartmouth and Rochester, as might be expected, also introduced mathematics as well as navigation into their curricula in the eighteenth century. However, the grammar school with the most impressive list of scientific alumni at that time was Manchester Grammar School. The headmaster during the important period 1764 to 1804 was Charles Lawson, a mathematician who employed a former pupil, Henry Clarke, to teach science. Clarke's influence was wide for he also taught later in Leeds, Salford and Liverpool before becoming Professor of History, Geography and Experimental Science at the Royal Military College.33 Most of the pupils from Manchester Grammar School went into industry or commerce and many were active later in the societies which were so important to the spread of scientific ideas and to those attitudes which encouraged innovation. A similar, though not so outstanding a role, was played by Hull Grammar School, where the curriculum was modified to suit the needs of the changing city. This happened in quite large numbers of grammar schools and R.S. Tomson34 says that 'the significance of the changes is difficult to gauge but of one thing there can be no doubt: the eighteenth century grammar schools were innovating, probably in response to public demands in education'. Though it is difficult to judge how much direct influence the grammar schools had on technical change, it can be said that they were not the obscurantist and reactionary institutions they were once believed to be and they were certainly not barriers to change nor to the dissemination of new knowledge. However, a much more direct link between education and technical innovation can be seen in the academies and private schools of the eighteenth century. Firstly, Parker35 and then McLachlan36 claimed that the dissenting academies were the first institutions to introduce what might be termed a modern curriculum with subjects such as mathematics, natural philosophy, modern languages and commerce, but Hans maintains that they both over-estimated the importance of these schools. Calvinism, which was one of the motivating forces behind them, cculd have a narrowing effect on their educational philosophy just as easily as a progressive one. Some dissenting academies were little more than theological colleges. Nevertheless, their contribution to the forging of links between science education and technology was impressive. T.S. Ashton37 says that

144

Education and Technology in the Industrial Revolution ... they were nurseries of scientific thought. Several of them were well equipped with 'philosophical instruments' and offered facilities for experiment: their teachers included men of the quality of Joseph Priestley and John Dalton; and from them proceeded a stream of future industrialists, among whom were John Roebuck (who was trained at Northampton before proceeding to Edinburgh and Leyden), Matthew Boulton, John Wilkinson, Benjamin Gott and—of a later generationJoseph Whitworth.

The Warrington Academy, though a late development, and tragically short-lived (1757-83), provides the best example of how advanced a dissenting academy could be. It also, through its masters, pupils and parents of pupils, shows how science (which Ashley-Smith says was always as up to date as possible)38 and the worlds of manufacture and commerce were linked. Musson and Robinson39 illustrate how it had strong links with men of practical bent through the Manchester Literary and Philosophical Society (it also had links with the Birmingham Lunar Society). Indeed, when it closed down, largely because of its own internal religious difficulties, it scientific apparatus was purchased for the Manchester Academy by the Society. A similar connection is claimed40 for the Northampton Academy and the local philosophical society because Doddridge, one of the most important eighteenth-century academy masters, and Clarke, his usher, were both members and exerted considerable influence on both the economic and educational life of Northampton. This seems to be true of so many of the academies and their masters: the name of Caleb Rotherham of the Kendal Academy crops up in all sorts of connections, as either a teacher, friend or adviser, of industrialists. Yet Hans says that the private academies were probably more important and, according to his sets of figures, they certainly educated more scientists (by his definition) than did the dissenting academies. He refers to the private academies as the first comprehensive schools and this they were, certainly in their curricula. 'The subjects taught could be divided into four main groups: (1) Literary, (2) Mathematics and Science, (3) Vocational and Technical and (4) Accomplishments and Physical Training.' He lists the subjects in each section but then goes on 'in actual practice the pupils were divided into groups: (a) those who desired to enter one of the Universities, (b) those preparing for the Navy or the Mercantile Marine, (c) for the Army, (d) for business and Law clerks and (z) for some technical profession'.41 Not all schools had all streams and some had only one, but all the academies did teach mathematics and at least half of them included natural philosophy in one form or another. Just how many academies existed at any one time it is difficult to say and it is not always clear for how long some of them survived or to what standard they taught, but Hans has gathered together enough evidence to show that they were important institutions. Every local newspaper carried advertisements for them throughout the eighteenth century but especially during the last two decades. Some must have given an education of doubtful value, e.g. R. Wright's school in Pontefract advertised an almost encyclopaedic list of

George Timmons

145

useful subjects and followed with a note that 'all sorts of Artificer's Work may be survey'd, for either Master or Workmen, at reasonable rates; and the most irregular Parcel of Land may be survey'd, or Gentlemen taught to Survey it by a new and exact Method founded on the Doctrine of Fluxions'.42 Other schools, such as Hackney, had a long tradition and an aristocratic clientele. However, the most important were those which, like the Warrington Academy, had some connection with the literary and philosophical societies, for these can be identified as having an influence on technology similar to that exerted by the societies themselves. Mention has already been made of the Manchester Academy, but there was also the College of Arts and Sciences which during its brief existence even made provision for lectures in applied chemistry to operatives and artisans.43 The Askesian Society in London also had links with schools and some of its members were teachers.44 Almost invariably, whether Birmingham, Liverpool, Sheffield or Newcastle are considered, such links existed. Schools such as these, and those set up by itinerant lecturers,45 offered real technical education and they depended for their existence on the usefulness of their courses: Mr Pitt, advertising his school in the Manchester Mercury in 1779, listed his practical apparatus as including 'optical instruments, cranes, pile drivers, sawing engines, pumps and steam engines'.46 When formerly intinerant lecturers like Pitt established more permanent educational institutions, they continued to offer the same sort of education, i.e. useful knowledge. But they were also propagandists of science, because they disseminated not only useful knowledge but also new attitudes, the most important of which was that change was possible and profitable and more likely to succeed if based on organized knowledge. It is not surprising that the lecturers had firm links with the societies: they were members of them or invited by them to lecture. They could have direct impact. Musson and Robinson say that 'men of chemical knowledge quickly put it to practical test in the bleaching and dyeing trades'.47 Between them, the societies, the lecturers and the schools helped to create the conditions in which the changes could take place: they provided useful knowledge in an organized way. Thomas Henry of Manchester was someone bitten by enthusiasm for the practical application of science, and his biography48 gives some idea of how various influences could come together. After attending Wrexham Grammar School, he became an apothecary in Knutsford and so developed an interest in chemistry, but it was from the time that he developed connections with both the Unitarians and the Lunar Society that his magnesia factory prospered. Later, as a member of the Manchester Literary and Philosophical Society, he advocated the study of science: 'A general knowledge of all will tend to open and enlarge his understanding, . . . while the study of something in particular (such as mechanics, hydrostatics, hydraulics or chemistry) may supply him with a kind of information which he may turn to good account.'49 He and Barnes were instrumental in the setting up of the College of Arts and Sciences which, though it was short-lived, at least shows that men like Henry saw the importance of education. The same was true of the Aiken brothers who were

146

Education and Technology in the Industrial Revolution

members of the Askesian Society: their grandfather had trained under Doddridge, when his school was at Kibworth, their father at Warrington, and they themselves at the Barbauld school, with one of them going on to Hackney. Inkster50 says that for them cby 1796, science was a way of life'. He also describes them and their society as 'the new men'. More often than not, such men advocated education because they understood that itinerant lecturers and societies could do only so much. In fact, numerous schools were established in London at the end of the century and they had links with the societies such as the Askesian. It was as though each informal organization for the diffusion of science strove to establish some sort of more permanent institution. This also occurred elsewhere, e.g. the Sunday Society in Birmingham led to the establishment of a Sunday school where Thomas Clark taught science to the 'Cast-iron Philosophers' from the Eagle St Foundry.51 Such was also the original purpose behind the Royal Institution because a technical school and a repository of mechanical inventions were also intended to form part of it.52 It was also, of course, why the Mechanics Institutes were set up a little later. Others, besides those who set up schools, advocated the establishment of permanent institutions because they understood the importance of education: even Adam Smith, who called for elementary education as a means of social control lest the factory system should demoralize the working classes completely, also proclaimed the value of geometry and mechanics in primary schools.53 One important industry which, except indirectly through the use of pumping engines and safety devices, remained primitive was mining. However, there were calls for mineralogy schools in the eighteenth century and William Phillips54 castigated the 'total ignorance of almost everything relating to the sciences of geology and mineralogy and above all chemistry in the conductors of mines and their agents'. On the other hand, education was also seen as dangerous: Eusebius, writing in the Gentleman's Magazine in 1797, said that ' . . . the man, whose mind is not illuminated by one ray of science, can discharge his duties in the most sordid employment without the smallest views of raising himself to a higher station His ignorance is a balm . . . we may therefore conclude that the Sunday School is far from being a wise, useful or prudential institution.'55 And William Henry, in accounting for the failure of the College of Arts and Science, said that there was ' . . . a superstitious dread of the tendency of science to unfit men for the ordinary details of business'.56 This last point is important because the fear was justified! Inkster57 highlights the ferment of ideas which occurred in the capital towards the end of the eighteenth century, with the establishment of societies and schools, with the presenting of public lectures of all sorts, and with the publication of large numbers of scientific and technical books, all contributing to the creation of an atmosphere in which invention, backed by science, was deliberately sought after, because of its practical value. Unnecessary distinctions between science and technology were not made, and this proved beneficial. However, he goes on to show how, by the 1820s, the study of science had become stratified. Members of the Askesian Society were drawn

George Timmons

147

into the 'academic' orbit of the Royal Society and the Royal Institution. Indeed, it was transformed into the Geological Society which was to operate in isolation from the one industry on which it should have made an impact, i.e. mining. In this way, the 'new men' were made 'unfit for the ordinary details of business'; so was Thomas Henry's grandson. However, by that time other social factors, not unaffected by the French Revolution and the wars with France, weakened the alliance between science and technology with unfortunate consequences for education so that the gains made in the eighteenth century were lost in the nineteenth. Though education did develop, it did so only slowly along stratified lines and neither the elementary schools nor the revived public schools paid much attention to science. Hartwell58 says of the changes which occurred in the eighteenth century ' . . . the advances were made on a broad front—in textiles, basic metallurgy, mining, transport, agriculture, and power production—suggesting that the revolution in technology was not the product of any simple intellectual or industrial stimulus, but rather the result of a growing awareness of the potentialities generally of technical progress'. That growing awareness owed a great deal to education because it was science-based, and dissenting and private academies, some grammar schools and Scottish schools, as well as itinerant lecturers and the philosophical societies, were important not only to the dissemination of scientific data but also to the spread of those new attitudes which stressed the possibility of change. The Lunar Society of Birmingham consisted of highly-educated men, most of them examples of what eighteenth-century schools and universities could produce, and in their work we can see the interplay of science, technology and education. Their achievements were enormous: Factories were supplied with power, and freed from concentration along river banks by the steam engine invented by Watt, developed and distributed by the industrial vision of Boulton. Problems of transportation were eased by the turnpikes and then the canals which Wedgwood, Darwin, Boulton, Small, Galton and Watt encouraged and helped to finance; roads were further improved and carriage design advanced by the studies of Richard Lovell Edgeworth, who also attempted, in his mechanical telegraphy, to speed the communication of information. These solutions required a commercial system in which credit was available for capital expansion and in which there was a dependable flow of currency; the Galtons organised a country bank, Boulton supported the organisation of others, and Boulton's coining press supplied the coinage that was needed to pay increasing numbers of workers.59 It was obvious to the members of the Lunar Society that education was important: it formed an integral part of their discussions. The same was true of the Manchester Literary and Philosophical Society, of the Askesian Society and all the others. Furthermore, had the forms of education which existed in the late eighteenth century continued to prosper in the nineteenth,

148

Education and Technology in the Industrial Revolution

then the situation which so concerned Playfair and Prince Albert need never have arisen. Notes 9.

1. R.M. Hart well, The Industrial Revolution and Economic Growth, 1971, pp. 178-

2. F. Crouzet, 'England and France in the Eighteenth Century: A Comparative Analysis of Two Economic Growths' in M.R. Williams (ed.), Revolutions, 1971, p. 150. 3. R.M. Hartwell, as note 1, p. 227. 4. A. E. Musson and E. Robinson, Science and Technology in the Industrial Revolution, 1969. 5. E. Ashby, Technology and the Academics, 1958, p. 50. 6. A.R. Hall, 'The Historical Relations of Science and Technology', Inaugural Lecture (1963) in P. Mathias, The Transformation of England, 1979, p. 46. 7. R. A. Buchanan, 'The Promethean Revolution' in History of Technology, first annual volume, 1976, p. 80. 8. Musson and Robinson, as note 4, p. 13. 9. J. A. Chaldecott, 'Wedgwood's Ceramic Wares for Chemical Use' in Ambix, Vol. 27, pp. 186-9. 10. D. Fleming, 'Latent Heat and the Invention of the Watt Steam Engine' in D. Mayr (ed.), Philosophers and Machines, 1976, p. 123. 11. A. L. Donovan, 'Towards a Social History of Technological Ideas' in G. Bugliarello and D.B. Doner (eds), The History and Philosophy of Technology, 1979, p. 24. 12. J. A. Chaldecott, as note 9, p. 193. 13. D.S.L. Cardwell, 'Problems of the Data-base' in Bugliarello and Doner, as note 11, p. 16. 14. In Musson and Robinson, as note 4, p. 52. 15. P. Mathias, The Transformation of England, 1979, pp. 48-58. 16. In Musson and Robinson, as note 4, p. 38. 17. As note 16, p. 49. 18. E. Schofield, The Lunar Society of Birmingham, 1963, p. 350. 19. G.F. French, 'The Life and Times of Samuel Crompton' (1859) in R.H. Campbell and R.G. Wilson, Entrepreneurship in Britain, 1750-1939, 1975, p. 46. 20. M.D. Marshall, English People in the Eighteenth Century, 1956, p. 223. 21. D.S.L. Cardwell, as note 13, pp. 12 and 15. 22. T. Kelly, 'The Origins of the Mechanics Institutes' in British Journal of Educational Studies, No. 1, 1952, p. 18. 23. V. Neuburg, Popular Education in the Eighteenth Century, 1971, pp. 115-25. 24. F. Smith, A History of English Elementary Education, 1930, pp. 66-7. 25. J.H. Higginson, 'Dame Schools' in British Journal of Educational Studies, Vol. XXII, No. 2, 1974, pp. 166-81. 26. D. Defoe, 'Everybody's Business is Nobody's Business'^ (1725) in M.G. Jones, The Charity School Movement, 1938, p. 4. 27. N. Carlisle, A Concise Description of the Endowed Grammar Schools in England and Wales, 1818, Preface p. xxxv. 28. N. Hans, New Trends in Education in the Eighteenth Century, 1951, p. 19. 29. M.L. Clarke, 'The Education of Royalty in the Eighteenth Century' in British Journal of Educational Studies, Vol. XXVI, No. 1, 1978, p. 84.

George Timmons

149

30. C. Wordsworth, Scholae Academicae, 1877, p. 11. 31. N. McKendrick, 'Science in the Industrial Revolution' in M. Teich and R. Young, Changing Perspectives in the History of Science, 1973, pp. 101-2. 32. A.L. Donovan, as note 11, p. 29. 33. N. Hans, as note 28, pp. 95-7. 34. R.S. Tomson, 'The English Grammar School Curriculum in the Eighteenth Century' in British Journal of Educational Studies, Vol. XIX, No. 1, 1971, p. 39. 35. I. Parker, Dissenting Academies in England, 1914. 36. M. McLachlan, English Education under the Test Acts, 1931. 37. T.S. Ashton, The Industrial Revolution, 1948, pp. 20-1. 38. J.W. Ashley-Smith, The Birth of Modern Education, 1954, p. 246. 39. Musson and Robinson, as note 4, p. 90. 40. D. Harding, 'Mathematics and Science Education in Eighteenth Century Northamptonshire' in History of Education, Vol. I, 1972, p. 146. 41. N. Hans, as note 28, pp. 64-5. 42. D.W. Sylvester, Educational Documents, 1800-1816, 1970, p. 252. 43. T. Kelly, as note 22, p. 21. 44. I. Inkster, 'The Askesian Society of London, 1796-1807' in Annals of Science, Vol.34, 1977, pp. 22-5. 45. Musson and Robinson, 'Science and Industry in the late Eighteenth Century' in Economic History Review, Vol. XIII, No. 2, 1960, p. 231. 46. As note 45, p. 234. 47. As note 45, p. 224. 48. W.V. Farrar, K.R. Farrar and E.L. Scott, 'The Henrys of Manchester' in Ambix, Vol. XX, 1973, pp. 183-208. 49. Musson and Robinson, as note 45, p. 224. 50. Inkster, as note 44, p. 24. 51. T. Kelly, as note 22, pp. 21-2. 52. W.G.G. Armytage, 'Some Sources for the History of Technical Education, Pt. 2 in British Journal of Educational Studies, Vol. V, 1957, p. 162. 53. In T. Kelly, as note 22, p. 18. 54. Quoted by R. Porter, 'The Industrial Revolution and the Rise of the Science of Geology' in Teich and Young, as note 31, p. 333. 55. J.M. Goldstrom (ed.), Elementary Education, 1780-1900, 1972, pp. 22-3. 56. Musson and Robinson, as note 45, p. 225. 57. Inkster, as note 44, pp. 5-31. 58. R.M. Hartwell, 'The Causes of the Industrial Revolution' in Economic History Review, Second Series, Vol. XVIII, No. 2, p. 175. 59. E. Schofield, as note 18, p. 438.

T h e

C o n t r i b u t o r s

W. ADDIS is a lecturer in materials science and structures in the Department of Construction Management at the University of Reading and is currently completing a Ph.D. on theory and structural engineering. HANS-JOACHIM BRAUN is Professor of Modern, Social, Economic and Technological History at the University of the Federal Armed Forces (Hochschule der Bundeswehr) in Hamburg. W. BERNARD CARLSON is an assistant editor on the Thomas A. Edison Papers at Rutgers University, and is currently completing a dissertation on Elihu Thomson, the electrical inventor. THOMAS DAY is a Senior Lecturer at the Scott Sutherland School of Architecture, Robert Gordon's Institute of Technology, Aberdeen, where he teaches theory and design of structure and the history of civil engineering. G. HOLLISTER-SHORT was formerly Principal Lecturer in History at Shoreditch College and is Honorary Lecturer in History of Technology in the Department of Humanities, Imperial College, London. R.H.J. SELLIN is Senior Lecturer in Civil Engineering in the University of Bristol. MIKULAS TEICH is Senior Research Fellow and Librarian of Robinson College, Cambridge. GEORGE TIMMONS is a Lecturer in Education at the University of Warwick and is currently engaged in a study of British education in the eighteenth century.

C o n t e n t s

o f

F o r m e r

V o l u m e s

First Annual V o l u m e , 1976 D.S.L. CARDWELL and RICHARD L. HILLS, Thermodynamics and Practical Engineering in the Nineteenth Century. JACQUES HEYMAN, Couplet's Engineering Memoirs, 1726-33. NORMAN A.F. SMITH, Attitudes to Roman Engineering and the Question of the Inverted Siphon. R.A. BUCHANAN, The Promethean Revolution: Science, Technology and History. M. DAUMAS, The History of Technology: its Aims, its Limits, its Methods. KEITH DAWSON, Electromagnetic Telegraphy: Early Ideas, Proposals and Apparatus. MARIE BOAS HALL, The Strange Case of Aluminium. G. HOLLISTER-SHORT, Leads and Lags in Late Seventeenth-century English Technology. Second Annual V o l u m e , 1977 EMORY L. KEMP, Samuel Brown: Britain's Pioneer Suspension Bridge Builder. DONALD R. HILL, The Banu Musa and their 'Book of Ingenious Devices'. J.F. CAVE, A Note on Roman Metal Turning. J. A. GARCIA-DIEGO, Old Dams in Extremadura. G. HOLLISTER-SHORT, The Vocabulary of Technology. RICHARD L. HILLS, Museums, History and Working Machines. DENIS SMITH, The Use of Models in Nineteenth-century British Suspension Bridge Design. NORMAN A.F. SMITH, The Origins of the Water Turbine and the Invention of its Name.

152

Contents of Former Volumes T h i r d Annual V o l u m e , 1978

JACK SIMMONS, Technology in History. R. A. BUCHANAN, History of Technology in the Teaching of History. P.B. MORICE, The Role of History in a Civil Engineering Course. JOYCE BROWN, Sir Proby Cautley (1802-71), a Pioneer of Indian Irrigation. A. RUPERT HALL, On knowing, and knowing how to . . . FRANK D. PRAGER, Vitruvius and the Elevated Aqueducts. JAMES A. RUFFNER, Two Problems in Fuel Technology. JOHN C. SCOTT, The Historical Development of Theories of WaveCalming using Oil. Fourth Annual V o l u m e , 1979 P.S. BARDELL, Some Aspects of the History ofJournal Bearings and their Lubrication. K.R. FAIRCLOUGH, The Waltham Pound Lock. ROBERT FRIEDEL, Parkesine and Celluloid: The Failure and Success of the First Modern Plastic. J.G.JAMES, Iron Arched Bridge Designs in Pre-Revolutionary France. L.J.JONES, The Early History of Mechanical Harvesting. G. HOLLISTER-SHORT, The Sector and Chain: An Historical Enquiry. Fifth Annual V o l u m e , 1980 THOMAS P. HUGHES, The Order of the Technological World. THORKILD SCHI0LER, Bronze Roman Piston Pumps. STILLMAN DRAKE, Measurement in Galileo's Science. L.J. JONES, John Ridley and the South Australian 'Stripper'. D.G. TUCKER, Emile Lamm's Self-Propelled Tramcars 1870-72 and the Evolution of the Fireless Locomotive.

Contents of Former Volumes

153

S.R. BROADBRIDGE, British Industry in 1767: Extracts from a Travel Journal ofJoseph Banks. RICHARD L. HILLS, Water, Stampers and Paper in the Auvergne: A Medieval Tradition. Sixth Annual V o l u m e , 1981 MARJORIE NICE BOYER, Moving Ahead with the Fifteenth Century: New Ideas in Bridge Construction at Orleans. ANDRfi WEGENER SLEES W YK, Hand-Cranking in Egyptian Antiquity. CHARLES SUSSKIND, The Invention of Computed Tomography. RICHARD L. HILLS, Early Locomotive Building near Manchester. L.L. COATSWORTH, B.I. KRONBERG and M.C. USSELMAN, The Artefact as Historical Document. Part 1: The Fine Platinum Wires of W.H. Wollaston. A. RUPERT HALL and N.C. RUSSELL, What about the Fulling-Mill? MICHAEL FORES, Technik: Or Mumford Reconsidered. Seventh Annual V o l u m e , 1982 MARJORIE NICE BOYER, Water Mills: a Problem for the Bridges and Boats of Medieval France. WM. DAVID COMPTON, Internal-combustion Engines and their Fuel: a Preliminary Exploration of Technological Interplay. F. T. EVANS, Wood Since the Industrial Revolution: a Strategic Retreat? MICHAEL FORES, Francis Bacon and the Myth of Industrial Science. D. G. TUCKER, The Purpose and Principles of Research in an Electrical Manufacturing Business of Moderate Size, as Stated by J. A. Crabtree in 1930. ROMAN MALINOWSKI, Ancient Mortars and Concretes: Aspects of their Durability. V. FOLEY, W. SOEDEL, J. TURNER and B. WILHOITE, The Origin of Gearing.

Natural K n o w l e d g e in Social

The Thomas

Journals

Context

of

A r c h e r H i r s t ,

FRS

B y W i l l i a m H . B r o c k and R o y M . M a c L e o d For more than forty-five years, from 1845 to 1892, Thomas Archer Hirst, mathematician, teacher and educational reformer, kept a diary almost daily. These diaries represent an incomparable social document of the world of Victorian science, and one of the most extensive portraits available of the scientific, intellectual and cultural background to the Victorian era. They reveal scientific rivalries, hidden motives, political reform, and the background to administrative reforms in the patronage of science at a critical period in its professionalization. For the microfiche publication, the diaries have been divided into nine chapters, each preceded by an editorial introduction. The journal pages, reproduced in full on thefiches,are interspersed with pages of notes that try to answer questions the user may ask: they identify personalities mentioned, verify dates, give bibliographical data for titles cited, and provide general historical background material. A booklet accompanying the fiches contains a full table of contents, a biographical essay on Hirst, a note on the history of the journals, a bibliography of Hirst's writings, and a list of the illustrations reproduced on the fiches. 1980, 80fiches(COSATI 60 frame format), box container, printed booklet (ISBN 0 7201 0373 8), £132.00 the set 'Anyone interested in the development of the sciences in the second half of the nineteenth century, or in any aspect of the scientific activities in London, or even in the general intellectual or literary history of this period, willfindthese journals a rich and varied source ofinformation.' MICROFORM REVIEW

M a n s e l l 6 All Saints Street, London N1 9RL, England