The Science of Bombing: Operational Research in RAF Bomber Command 080209936X, 9780802099365

Citation preview

THE SCIENCE OF BOMBING: OPERATIONAL RESEARCH IN RAF BOMBER COMMAND

After suffering devastating losses in the early stages of the Second World War, the United Kingdom’s Royal Air Force established an Operational Research Section within bomber command in order to drastically improve the efficiency of bombing missions targeting Germany. In The Science of Bombing, Randall Wakelam explores the work of civilian scientists who found critical solutions to the navigational and target-finding problems and crippling losses that initially afflicted the RAF. Drawing on previously unexamined files that reassess the efficacy of strategic bombing from tactical and technical perspectives, Wakelam reveals both the important role that scientific research and advice played in operational planning and the remarkable intellectual flexibility that existed at Bomber Command. A fascinating glimpse into military strategy and decision making, The Science of Bombing will find a wide audience among those interested in air power history as well as military strategists, air force personnel, and aviation historians. randall t. wakelam is the director of research and symposia at the Canadian Forces College and an assistant professor of Defence Studies with the Royal Military College.

This page intentionally left blank

RANDALL T. WAKELAM

The Science of Bombing Operational Research in RAF Bomber Command

UNIVERSITY OF TORONTO PRESS Toronto Buffalo London

© University of Toronto Press Incorporated 2009 Toronto Buffalo London www.utpublishing.com Printed in Canada ISBN 978-08020-9936-5 (cloth) ISBN 978-08020-9329-6 (paper)

Printed on acid-free paper

Library and Archives Canada Cataloguing in Publication Wakelam, Randall T. (Randall Thomas), 1953– The science of bombing : operational research in RAF Bomber Command / Randall T. Wakelam. Includes bibliographical references and index. ISBN 978-0-8020-9936-5 (bound) ISBN 978-0-8020-9629-6 (pbk.) 1. Great Britain. Royal Air Force. Bomber Command. Operational Research Section – History. 2. Bombing, Aerial – Great Britain – History. 3. Bombing, Aerial – Germany – History. 4. Harris, Arthur, Sir, 1892– 1984. 5. World War, 1939–1945 – Aerial operations, British. 6. Operations research – Great Britain -- History. 7. Military scientists – Great Britain – History. I. Title. UG735.G7W34 2009

940.54c4941

C2009-902334-2

University of Toronto Press acknowledges the financial assistance to its publishing program of the Canada Council for the Arts and the Ontario Arts Council. University of Toronto Press acknowledges the financial support for its publishing activities of the Government of Canada through the Book Publishing Industry Development Program (BPIDP). This book has been published with the help of a grant from the Canadian Federation for the Humanities and Social Sciences, through the Aid to Scholarly Publications Program, using funds provided by the Social Sciences and Humanities Research Council of Canada.

Contents

Preface

vii

Introduction: Commanders, Headquarters, and Decision Making 3 1 Strategic Bombing: When Theory and Practice Do Not Match 12 2 Operational Research: Finding a Solution through Science 24 3 Boffins at Bomber Command: September 1941 34 4 Sorting Out Process and Producing Results: September 1941– February 1942 48 5 Arthur Harris and a New Beginning: February–December 1942 64 6 Concentration and Other Curatives: February–December 1942 80 7 Putting Concepts to the Test: January–February 1943 101 8 New Concepts, Old Targets – The Ruhr: March–June 1943 116 9 Hamburg and New Science: July–October 1943 139 10 Science versus Attrition – Berlin: November 1943–March 1944 157 11 Debates Both Political and Technical: April 1944–May 1945 183 12 Round-the-Clock Bombing: April 1944–May 1945 205 13 Conclusions 226 Appendix 1: Britain and Northwest Europe: Selected Bomber Command Targets 237 Appendix 2: Glossary of Operational Teminology 239 Appendix 3: Air Ministry and Bomber Command Headquarters Organization Charts 249 Appendix 4: Air Ministry and Bomber Command Appointments 253 Appendix 5: Air Staff Directive to Bomber Command, 13 July 1940 257 Appendix 6: Principal Bomber Command Aircraft 261

vi

Contents

Appendix 7: Bomber Command ORS Senior Personnel Appendix 8: Bombing Techniques 271 Notes

285

Bibliography Index

327

333

Illustration section follows page 182

269

Preface

Many people have helped me, not just with this book but with my graduate and professional education writ large. Foremost, I would like to express my thanks for the infinite patience and deft guidance of my doctoral supervisor, Professor Terry Copp. Taking on an ‘adult learner’ can be a significant engagement and Professor Copp was faced with a number of unique challenges in getting me through seven years of on-again, off-again study and writing, compounded by my often disruptive military career. Through all of this, he expressed his continued confidence that I would both survive the passage and become a credible historian in the process! He has been both enormously supportive and an exemplary role model; for him history is truly a vocation. His ethic is indicative of the directors and faculty of the Tri-University PhD in History at Wilfrid Laurier, Waterloo, and Guelph. They have offered and continue to offer an exceptional opportunity for students seeking a distinctive program which draws on the richness of three history departments. I have particularly appreciated the confidence of Professor Joyce Lorimer in getting me started and the efforts of Professors John English, Barry Gough, Geoff Hayes, and Lynne Taylor and, more recently, Roger Sarty in guiding me along the way. Several of my professional colleagues have also been great supporters. Three ‘bosses’ in particular, Brigadiers Ed Gosden and Robin Gagnon as well as Rear Admiral David Morse, worked their magic to get me the time needed to produce seminar papers along the way. Military historians Drs Jack English and Allan English wisely encouraged me to publish. Ms Catherine Murphy, Head of the Information Resource Centre at the Canadian Forces College (the Canadian military’s senior professional school) in Toronto, repeatedly went well beyond any reasonable level

viii

Preface

of support in finding for me a wealth of materials on a wide range of unique and often obscure topics. Her encyclopedic knowledge of the literature of military art and science has been a boon to me as it has to thousands of staff and students at the College over the past two decades. Her domain, ‘the room with the books,’ is an inviting and intellectually stimulating retreat at that busy institution. Turning to the support I have received in the writing of the book I would like to start with Ms Murphy, who was able to produce a working copy of the Bomber Command material in the Ronnie Shephard Collection Archives held by Wilfrid Laurier University. I would also like to thank Dr Steve Harris, Senior Historian at the Directorate of History and Heritage at National Defence Headquarters in Ottawa. As Canada’s resident expert on Bomber Command, he not only made available to me the Canadian copies of the Public Record Office files which he had brought back to Canada for his own research, but he also asked a number of hard questions which helped to shape my thinking. I would also like to thank Mr Sebastian Cox, Head of the RAF Air Historical Branch, for reviewing some of my early writing and for giving me access to the Branch’s records. While I was working with those documents he was also the consummate host, standing me a drink at lunch and giving me full use of his personal coffee supplies – not a small matter when one is jet-lagged. I would be very remiss were I not to make mention of the staff at the U.K. Public Records Office who went far out of their way to help me see many more files than would normally be the case. I must pay special thanks to Janet Lacroix of the Canadian Forces Joint Imagery Centre and Mike Bechthold of the Laurier Centre for Military Strategic and Disarmament Studies at Wilfrid Laurier University. Janet found some remarkable photos to help tell the story while Mike very expertly revamped a number of largely unusable primary diagrams and charts and drew or redrew all of the material in appendix 8 from scratch. Both of these visual forms are critical to getting a sense of the challenges which faced the scientists and the flyers. Finally, I would also like to thank a number of people from the editorial staff at University of Toronto Press without whose efforts the manuscript would never have become this book. Len Husband has had the challenging role of guiding a first-time author through the labyrinth of publishing. His frank and sound advice and constant good humour have been very much appreciated. The toughest work has gone to Frances Mundy and Judy Williams, who have had to slog through the manuscript sentence by sentence. They too have been enormous help – and have boosted my morale with their compliments about the story.

Preface

ix

In the end the story which I set out in the following pages is my own interpretation of the events which the protagonists recorded in their memos, reports and files. I sincerely hope that I have represented them fairly and accurately; if there are errors they are my responsibility and burden. It must also be stated that family have played the major part in the genesis of this volume. My interest in things military and historical stems from my parents, Donald and Sharon Wakelam, who encouraged me to research and teach even when in high school. It has taken me a while to get to this point, but here I am. Finally, though, I want to say that this book could not have been completed without the selfless support of my wife and family. Norma, Christian, Karen, and Kevin have given up untold hours of husband and father and have patiently endured my frustration when the project had to take a back seat or, worse, be put on hold when life got in the way. Despite this they have always encouraged me to keep on, even refusing to let me ‘bail out’ when I could see no way forward. To them I offer my unqualified thanks and love. RTW Kingston, 2008

This page intentionally left blank

THE SCIENCE OF BOMBING

This page intentionally left blank

Introduction: Commanders, Headquarters, and Decision Making

This book examines the contribution of RAF Bomber Command’s Operational Research Section to the strategic bombing offensive. The central focus of the study is the large and largely unrecognized contribution to the success of Bomber Command made by scientists of the ORS. The issues described in the following pages are based on a number of important but in the main neglected original sources, including the unpublished history of the ORS written by the scientists themselves, the reports produced by the Section, and in particular the relevant documents in the Bomber Command Headquarters files. A number of the reports produced by the Section have been used by historians previously to assist in constructing a narrative of the bomber offensive, but this book offers the first analysis of the role of the Section within Bomber Command and by extension provides an examination of how that headquarters directed the strategic bombing campaign against Germany. Additionally, the following pages contribute to an understanding of the way in which the function of command was exercised in the Royal Air Force and Bomber Command. Given the number of books on campaigns and leaders, one might expect to find useful studies of the way in which generals, admirals, and air marshals have carried out their duties. Instead we find a literature focused on ‘great men’ and ‘great events,’ with remarkably little attention paid to the interaction between staff officers and their commanders, never mind the scientific advisers that served the army, navy and air force during the war. As in all history, this study contains an autobiographical thread which should be disclosed. This investigation has been undertaken by someone who has had a career of some three decades in the air force including over three thousand flying hours, experience in command and staff ap-

4 The Science of Bombing

pointments, and over a decade providing education programs for midlevel staff officers and senior commanders. My approach to this topic was therefore influenced by my professional experience, which provided me with a degree of scepticism about the value of operational research. To be specific: not once, whether in flying units or in the national headquarters, did I see the active participation of the air operational research staff in providing advice on how best to organize for specific roles and missions, what equipment to purchase, or how to optimize the air force and its policies and procedures for the most efficient use of limited resources. Perhaps it was the relative stability of the Cold War which led to a disuse of operational research, perhaps it was the relative smallness of the Canadian air force which rendered any OR staff so small as to be non-effective, or perhaps it was the apparent inclination of the scientists of the era to immerse themselves in mathematical modelling. In any case, before I undertook this research, my opinion of the worth of operational research was not high. This study has also been framed in a very real sense by my first encounter with some of the key primary sources – the partial records of the Operational Research Section of RAF Bomber Command held in the Ronnie Shephard Military Operational Research Archive at the Laurier Centre for Military Strategic and Disarmament Studies. Amounting to two banker’s boxes of reports and files, the material is most intriguing. The ORS manuscript history, also part of the Shephard collection, describes the work of a fairly large group of scientists and their technical staff, covering all aspects of Bomber Command operations. As will be seen in detail in the chapters which follow, their work took them into areas of military technology and process for which they had no previous training. Despite this, their story indicates that once they had ‘won their spurs’ (perhaps ‘wings’ would be the more appropriate term) the conclusions and advice they provided were accepted on all occasions. These claims, on the basis of my own experiences working in flying headquarters, seemed beyond belief, and yet there were enough references to ‘brilliant’ young scientists in Sir Arthur Harris’s writing and other accounts of the period to suggest that the truth might well be closer to the ORS account of events than I was initially prepared to accept. The more I read of the manuscript the more incredulous I became: how could career aviators turn such fundamental questions as navigation and aircraft protection over to people who by and large had never flown in a bomber, let alone at night on a raid over hostile territory? Finding the answers to these questions seemed important to me as an aviator, as a

Introduction 5

military educator, and as a historian. There was clearly something missing in the story of Bomber Command. I have decided to concentrate on issues of fundamental importance to bomber operations: offensive effectiveness, mainly a question of navigation and target finding; and defensive effectiveness through measures of self-protection designed to reduce losses. The reason for choosing these two issues is fairly straightforward. First, crews whose navigation was faulty could lose their way and fail to find the target; their bombs became a waste. Moreover, separated from the bomber stream they could become isolated and more easily fall prey to enemy attack. Finally, they could fail to find their way back to England. Even aircraft that did reach the target area and return home safely, but which could not identify the aiming point, were also likely to have wasted their bombs. On the opposite side of the balance sheet, losses due to enemy action meant that aircraft and crews would have to be replaced from a pool of finite resources. Resolution of these issues would render a raid more efficient; this in turn would make the overall work of the Command more effective. This work focuses on questions of tactics and technique, rather than strategic policies and ethics of area bombing. The scientists, as shall be seen, were not given a mandate to investigate these issues and did not do so. Some readers may find the lack of discussion of the direction of the strategic bombing campaign and of the morality of area bombing somewhat surprising, but these issues have already been profoundly considered from many points of view in a literature that is now quite vast. The arguments in that literature tend to be two: first that the commanders should have known that strategic bombing could not break and was not breaking German morale or degrading their war economy; and second, that whether it was achieving these ends or not, it was immoral. There also enters a personality aspect to the controversy wherein, according to received wisdom, Arthur Harris’s bullish support of area bombing during his three years in command made him particularly culpable of sending out his crews to a pointless and morally indefensible death. As elsewhere, there are many historical interpretations to facts, and the construct just presented fails to recognize that Harris was not the originator of the concept of strategic bombing, or, in the specific case, the policy of area bombing. The former was the product of the thinking of several men immediately after the First World War: Italian Guilio Douhet, American Billy Mitchell, and from Britain Hugh Trenchard, the father of the RAF. The notion of carrying the attack to the enemy was

6 The Science of Bombing

firmly rooted in military thinking, most significantly in the ideas of nineteenth century German military philosopher Karl von Clausewitz, arguably the most important military thinker in the West, who reasoned that warfare was intended to defeat the will of the enemy. Indeed official RAF doctrine, found in Royal Air Force War Manual AP 1300, published first in 1928, stressed that an enemy was defeated once its government and citizens had lost the will to continue. That said, area bombing was not acceptable: ‘all air bombardment aims to hit a particular target,’ an ‘exact target,’ and these were listed in the various war plans developed in the 1930s.1 The actual decision to switch to area bombing was not Harris’s at all. It was not even an air force decision, but rather was made at the strategic level with the endorsement of the prime minister and came to Bomber Command in July 1941 (eight months before Harris took over the Command) via a Chiefs of Staff Committee decision. If we accept for a moment that strategic bombing was not Harris’s concept and that area bombing had been adopted before he came to Bomber Command, then there are still the questions of legality and morality. Dealing with legality first: because there were no laws governing air warfare at the time, scholars have based their arguments on naval law, which proscribes attack against ‘undefended’ communities. In this case a Germany defended by hundreds of night fighters and tens of dozens of anti-aircraft batteries was arguably well defended. Moreover, attacks on arms depots and plants were allowed under naval law.2 Turning to the question of morality, historians have argued that, in view of the clear immorality of the Nazi regime, the use of strategic bombing to knock out Hitler was relatively moral. If there was no other tool available for bringing offensive action against Germany for an extended period, then the use of strategic bombing, imprecise though it was until 1944, was better than doing nothing. This brings us to the question of the effectiveness of the campaign. As will become all too apparent in the coming pages, attacks against Germany, or any target, were never guaranteed to work and were in many cases abject failures. But the overall impact of those raids, which by 1944 had become largely successful and precise, was to divert significant resources to the air defence of Germany and away from offensive operations.3 The other element of the critics’ argument is Harris himself, who is typified as an uncompromising proponent of area bombing. There is little disagreement that Harris was a determined leader, but, as suggested above, he was not the originator of strategic bombing, and his was not the decision to use area bombing against Germany. Significantly, Har-

Introduction 7

ris had not always been a Bomber man; he started off as a fighter pilot and subsequently commanded transport and flying boat units. While he was fully committed to the bombing campaign, when directed to use his forces for anti-shipping mining operations and in support of the D-Day armies he approached these missions with equal resolution, albeit not without initial debate as to the viability of such undertakings. There is one other facet of Harris that is often criticized: he is seen as a ruthless commander who sent his men to their deaths without much thought. But, as will be seen, this was not the case. What Harris did do was direct the operations and staff activities of his force as effectively as possible, day in and day out for over three years, to ensure that when he did send his crews against the enemy they would be as effective, and safe, as possible.4 As a personal observation, I do not believe that Harris had any unethical or criminal desire to kill German civilians. Rather it seems to me that he was attempting to follow strategic directives which called for the dislocation (i.e. destruction) of Germany’s economic and morale capacities to wage war. Clearly it would have been more effective and efficient to do so with a precision bombing capability, one which would theoretically have permitted the achievement of decisive results sooner. Before 1939, the bomber commanders had believed they had such a capability; however, it had become very apparent that the best they could do, in 1941, was to conduct area bombing. At that point they could either sit idly attempting to sort out their failings, or make best use of the resources they had while fixing problems as they went along. The first option was obviously not going to keep Britain in the war and so they employed the second alternative. And, as this study will show, they worked hard every day to resolve the problems which were stopping them from applying the very concepts which they had intended to use from the outset. In the interim, they were obliged to live by the axiom of Lord Kitchener, voiced during the First World War: ‘We must make war as we must; not as we would like.’5 In the course of this research, the question began to focus on the extent to which the analyses, conclusions, and recommendations of the scientists affected the technical and tactical decisions of the commander and his key staff. Put simply, did the ORS have a significant influence on command decisions or did it not? While initially I had no evidence of either possibility, my first scan of the manuscript history shaped my preliminary conclusion that there had been a positive correlation between the scientists’ work and Harris’s major decisions.

8 The Science of Bombing

In order satisfactorily to prove this conclusion I have felt it necessary to demonstrate a link between work done by the ORS and either decisions made by Harris and key subordinates or changes to procedures instituted by the staff on the basis of ORS recommendations. My desire was to find evidence of a sort of continuum which started with a problem and led to the adoption of an ORS developed solution. Visually this chain might look like the following:

> 1. Problem 2. Problem defined by passed to Commander ORS or key staff

3. ORS develop research plan and conduct research

4. ORS submit conclusions and recommendations

5. Commander or key staff accept and implement solution

There were two broad approaches possible for the assessment of the primary sources: thematic or chronological. I felt the former would not portray the complexity of the day-to-day circumstances faced by Harris, his staff, the scientists, and the flyers. A thematic approach, too, would tend to miss the linkages between the various problems faced by the Service staff and the scientists. I felt that the thematic approach would simply not present an accurate picture of the fog and friction of war that even headquarters-bound aviators must experience. In the end, however, I decided to adopt a combination of the two approaches: while following a chronological path, the following pages present offensive and defensive matters separately within each time block. Chapter 1 provides a background of strategic bombing concepts, discusses the genesis of bombing within the Royal Air Force, and examines the problems encountered by Bomber Command during the first two years of the war. Chapter 2 presents an introduction to the science of operational research and looks at how OR came to be employed in other commands within the RAF. Chapter 3 examines the structure of Bomber Command Headquarters as well as the events and personalities involved in the formation of the OR Section. Chapter 4 turns to the actual work of the section; in its early days that work focused as much on developing analytical processes as it did on solving the problems confronting the Command. Chapter 5 reviews the arrival of Sir Arthur Harris in February 1942 and the Headquarters’ efforts in turning around the bombing offensive. The remaining chapters follow the successes and setbacks of the Command from

Introduction 9

mid-1942 until the end of the war, covering the major campaigns, technical developments, and policy shifts from the perspective of the scientists and the headquarters staff. Finally, a brief conclusion offers an evaluation of the achievements and impact on decision making of the Operational Research Section as well as some thoughts on commanders and headquarters. Each of the analytical chapters is built around a framework of main events and issues facing Harris and the Command. The framework itself has been drawn from such sources as the RAF official history, Harris’s personal memoirs and his official record of the Command’s activities, and the Canadian official history. These skeletons are then ‘fleshed out’ largely using primary sources which demonstrate the involvement of the ORS in resolving the problems besetting the Command. There is no lack of published material on the related issues of Harris, Bomber Command, and the strategic bombing campaign. Analyses of the campaign are almost legion. It has been approached in studies ranging from detailed technical histories of aircraft and equipment to philosophical works that question the morality of area bombing. Some studies examine individual raids, others consider battles (a series of raids against a specific target), while yet others review the entire strategic bombing campaign. Many weave together events and debates about policy, strategy, tactics, and equipment that that cut across all levels of government and the military. The official histories which have been used are the Royal Air Force’s The Strategic Air Offensive against Germany6 as well as volume 3 of the Royal Canadian Air Force official history. The latter, The Crucible of War, 1939– 1945,7 details the participation of the tens of thousands of Canadians who, like other Commonwealth personnel, flew with the RAF during the war. One additional official history is of fundamental importance. Published in 1963, The Origins and Development of Operational Research in the Royal Air Force deals with the impact of operational research across the RAF.8 Harris’s own thinking on the campaign is recorded in two works. The first, which became readily available to the public only a decade ago, is Despatch on War Operations 23rd February, 1942, to 8th May, 1945.9 In this official document Harris and his staff provided a brief description of the operations carried out by the Command while he was commander. More widely available is his 1947 (republished in 1990) Bomber Offensive.10 The key unpublished source used in this study which describes the Operational Research Section’s contribution is Operational Research in Bomber Command.11 This is the manuscript history that the Section produced, probably in late 1945, under the signature of the Section head, Dr Basil

10

The Science of Bombing

Dickins. Six hundred and twenty pages in length, the document presents the accomplishments of the Section in thematic order, dealing with various problems as they were investigated and resolved over the course of the ORS’s mandate. For the more mathematically minded reader, there is an excellent and lengthy chapter dealing with the processes of qualitative and quantitative analysis, as well as the development of the data collection techniques which permitted meaningful assessment of operations. Although scientists and technicians associated with the Section are identified in the annexes of the document, there is no discussion of personalities or mention of which individuals would have worked on particular issues. In addition to the manuscript, there are also the actual ORS reports, numbering in the hundreds. The RAF Air Historical Branch in London holds an almost complete set of each of the series of the Bomber Command reports, including associated diagrams, as well as ‘Interim Raid Reports’ and ‘Night Raid Reports’ that detail the planning, conduct, and outcomes of individual raids and are referred to in the manuscript history and in the British official history. The only limitation of these papers is that they lack any indication of their impact or even of who had seen them. Thus while the manuscript and the reports are sufficient to whet the curiosity they leave significant questions and some doubts as to the scientists’ role. The manuscript is, in a sense, an autobiographical document, and while the veracity of the story cannot be disproved, neither can it be proved. Missing from the picture is the actual interchange of ideas and the debate that must surely have gone on as Service officers and scientists attempted to define and then solve the many issues described in the manuscript. Moreover, the manuscript, by discussing the ORS’s work thematically, does not adequately show the complexity of the Section’s work, given that the scientists would have been dealing with multiple, and potentially contradictory, problems at the same time. In essence, the manuscript offers an altogether too tidy and positive source from which to work exclusively. Finding a source that would reveal what issues were discussed, and also how, why, and when, led to a review of personal papers, including the Harris papers, but these did not provide the anticipated detail on ORS work. As a result, the main evidence in support of this book was eventually found in the working files of the Command which are held at the UK Public Records Office. There are approximately four thousand Bomber Command files, and those dealing with aircraft and personnel losses, navigation and target-finding devices and procedures, and enemy countermeasures were reviewed in detail. These working files are generally excellent in that they contain relatively complete accounts of major issues, including minute

Introduction 11

sheets which record in detail the opinions of various commanders, staff officers, and scientists. As well there are copies of key documents which often show the handwritten amendments made by senior personnel. Taken together these documents provide an excellent sense of the sorts of questions that the ORS investigated and the reaction that their work got from the staff and commanders both in Bomber Command and at more senior and subordinate formations. In particular, the great value of many of these files was their ability to confirm the direct or indirect involvement of the ORS in resolving problems that the Service staff could not. Before we can begin to consider these sources in detail it is necessary first to review the RAF’s record as a strategic bombing force up to mid-1941. We will also briefly explore the concept of operational research and then to see how it had been applied to other problems facing the RAF in the late 1930s and during the early years of the war.

1 Strategic Bombing: When Theory and Practice Do Not Match

In 1939, Britain’s Royal Air Force had been in existence for just thirtyone years and Bomber Command for just three. The Command numbered some 200 aircraft of various vintages, almost all of them obsolete or obsolescent.1 By 1945, the strength of the command had grown to some 1609 modern frontline aircraft. In the opening year of the war, attack forces had amounted to a few dozen aircraft which often suffered grievous losses while striking little on or even near targets in Germany. By the end of the war, raids of several hundred aircraft, sometimes more than a thousand, could be mounted repeatedly; losses had fallen to an average of 1 per cent or less and the precision of the attacks, whether by day or night, in fair weather or through cloud, was regularly measured in a few hundred yards’ divergence from the aiming point.2 Bombing of strategic targets – industry, transportation systems, and energy supplies – had begun during the Great War when powered flight had given combatants a new weapon, but it was not technically advanced or tactically potent. Still, military planners found the concept well worth pursuing. Indeed the RAF was in part the child of a desire by Britain to carry the war to Germany just as the German Zeppelin attacks, which had begun in 1915, and the later Gotha bomber raids had brought ‘death and destruction’ to England. Many historians agree that the genesis of Britain’s bombing concepts came in the experiences of the German bombing of London during the Great War, in which the civilian population had reacted with terror and the British defences appeared ineffectual. These factors, coupled with the apparent successes achieved by the bombing work of the ‘Independent Force’ of British bombers in 1918, allowed its commander and future head of the RAF, Sir Hugh (later Lord) Trenchard, to conclude that the

Strategic Bombing 13

bomber was an effective offensive weapons system against which fighter aircraft would never amount to a credible defence. ‘In other words, in 1918, it was already part of Trenchard’s military thinking that the bomber deserved to be placed on the plinth as the ultimate weapon, although he had not amassed sufficient evidence to test any such assumption.’3 This concept, coupled with Trenchard’s dominating personality and the need for a role different from those of the army and Royal Navy for the newly created but politically weak Royal Air Force, formed officially on 1 April 1918, would coalesce to form an air doctrine of strategic bombing during the coming decades.4 During the 1920s and early 1930s, bombing remained a largely theoretical matter as there were few opportunities to carry out bombing trials. There were only a few small conflicts, in which limited ‘frightfulness bombing’ was conducted more as a show of force than to destroy major targets. At the same time the RAF, like the other services, suffered from a lack of resources as Britain went through a period of minimal defence spending. As the threat of a second European war started to become a reality in 1933–4, thoughts of restructuring took shape.5 In 1934 navy, army, and air force planners set about to define possible scenarios for hostilities. The work ended up taking two years, after which it was necessary to turn the broad concepts for possible operations into detailed plans that could be implemented on short notice. Many of these plans were joint in nature, involving forces from two or more of the services, but counterattacks against the Luftwaffe and the attack on Germany were the responsibility of the RAF, and more specifically Bomber Command, alone. There were three plans which would involve the Command: WA1, an attack on the Luftwaffe; WA4, an attack on rail and waterways systems; and WA5, an attack on war industry including oil refineries particularly in the Ruhr, Rhineland, and Saar areas of western Germany (see map at appendix 1). Of these, the first two were found to be problematic for a variety of reasons, while the third could be prosecuted, but only with difficulty owing to a lack of crews and aircraft. Also required was a shift in policy to include industrial sites as ‘military’ targets, since there was concern that attacks against non-military targets would trigger retaliatory bombing of British civilian populations.6 Paradoxically, while this strategic planning was going on, the Air Officer Commanding-in-Chief (AOCinC) of Bomber Command, Air Chief Marshal Sir Edgar Ludlow-Hewitt, expressed his concern that losses among his squadrons would be catastrophic, given the state of training and the capabilities of his aircraft.7 The Chief of the Air Staff (CAS), Air

14

The Science of Bombing

Chief Marshal Sir Cyril Newall, was also on record as being unsure that a nation could be knocked out from the air in any case.8 Ludlow-Hewitt had specific concerns about the actual abilities of his forces. He did not believe that his crews were prepared for war, citing, among other examples, the employment of vitally important mechanics as part-time air gunners.9 Based on these sorts of weaknesses, he argued that he would lose his medium bombers within three and a half weeks and his heavy squadrons in less than two months. He suggested that the use of fighter escorts would mitigate these difficulties, but Trenchard, the first head of the RAF and long-time proponent of long-range bombing, had vetoed such a possibility in 1923, and the doctrine of the RAF had not changed subsequently. Ludlow-Hewitt also recognized that he did not have the ability to take on all of German industry, plus communications and transportation systems, at one time and so proposed limiting his attacks to power stations and coking plants. He had seen what needed to be done, but the Trenchard legacy would not permit a shift to this more pragmatic implementation. Arthur Harris considered LudlowHewitt to have the best intellect he had ever seen, but a strong intellect did not carry the day.10 Indeed, while strategic priorities were in general sound, there appeared a lack of desire to fully acknowledge the impracticality of dealing with these objectives.11 Target sets like power and fuel facilities were well chosen, but at the same time it was accepted that attacks against targets in the heavily defended Ruhr valley, the industrial heart of Germany, would have to wait for the new and more capable aircraft still on the drawing boards. These aircraft would have better performance, increased bomb loads, additional defensive weapons, and new generation navigation systems. Additional complexities were added to the challenges of longrange bombing as the last months of peace passed. When staffs began discussing the possibility of sending bombers into Germany by night, the senior officers reminded themselves that most bombing was intended to be carried out by day, not least because precision bombing at night appeared problematic based on difficulties in identifying targets. Given all of the problems facing effective bombing operations, the RAF did not look unfavourably on the 21 June 1938 announcement by the prime minister that Britain would bomb only purely military targets, thereby avoiding the risk of casualties among German civilians, and, in the politicians’ minds, of retaliation in kind. By 1939, the only plan which seemed to be operationally possible was a program of leaflet-dropping to be conducted at night.12

Strategic Bombing 15

Even in that year the Air Staff at the Ministry had not fully realized and certainly not mastered the tactical and technical challenges associated with strategic bombing. There was insufficient production of new pilots and, until the previous year, no recognition of the need for a navigator in the crew. Little night flying was being done and that which was accomplished was not conducted under operational conditions. Navigation training and flying precision were poor. Yet, the Committee for the Scientific Survey of Air Offence, a group of scientists under Sir Henry Tizard, a researcher who had been intimately involved in RAF scientific investigations since the Great War and who also held the post of rector at Imperial College London from 1929 to 1942,13 recognized the potential of radio navigation aids in addressing these problems. Air Ministry bureaucracy precluded these researchers, however, from talking directly to Bomber Command or from having access to an actual operational unit in order to conduct experiments and trials, so that little progress was made by Tizard’s group.14 Gunnery for self-defence was based on the .303 calibre machine gun, which lacked stopping power and range, but was at least being mounted in many cases in power-operated turrets. There were also problems with the theory and practice of bomb dropping. High-level bombing seemed to be the best way to proceed, but it was virtually impossible to conduct valid trials owing to the lack of a suitable bombing range or testing organization. The need for post-attack photo reconnaissance and bomb damage assessment to confirm the damage done during an actual raid was also recognized, but adequate aircraft and cameras had yet to be designed or procured. Target identification and marking at night were seen as major problems, but again the lack of a testing range made progress slow and ineffective before 1939. On balance, relatively little work had been done in proving the theories to which proponents of strategic bombing had so long subscribed or in finding effective means for implementing these concepts. As the official history states bluntly, ‘thus, when war came in 1939 Bomber Command was not trained or equipped to penetrate into enemy territory by day or to find its target areas, let alone its targets, by night.’15 There was no lack of hard data to point to these problems. Between 1937 and 1939 there were no fewer than 478 cases of aircraft being forced down when the pilots became lost, and this over the United Kingdom. As well, just before the war began, it was concluded that the average pilot could only get his aircraft to within fifty miles of the target using ‘dead reckoning’ navigation by day (‘dead reckoning and other terms are defined in the glossary at appendix 2). And yet there had been no

16

The Science of Bombing

air navigation office within the Air Ministry until 1938. But the anomaly between data and inaction had been noticed by the scientists, Tizard writing that ‘the fact is that no one seems very anxious to get our advice on these subjects ... and there seems to be no anxiety on the part of the Air Ministry that we should meet.’16 ‘The Air Staff simply did not appreciate the need for accuracy in navigation, but merely assumed that their bomber crews had such an ability.’17 In hindsight it seems clear that the magnitude of the problems was apparent even at the time. For example, Sir Jack Slessor, who worked in the plans staff at the Air Ministry during these years and who would himself rise to the post of Chief of the Air Staff in the early 1950s, admits to the many shortcomings of the staff during the pre-war years. The legitimate criticism of the Air Staff before the war is that we paid insufficient attention to the technique of bombing. ... there is no doubt that we did underestimate the technical difficulties of modern air bombardment, and might have been more farsighted in our efforts to develop the major weapon of air-power, the bomb. We attached insufficient importance to things which afterwards became commonplace; like bombing and navigational aids, signals equipment, D/F beacons and blind landing systems. We had no Bomber Development Unit until 1939. ... Crews had little or no experience of really long-range navigation in bad weather or training with live bombs as we now understand it. ...18

Slessor points out that many of the lessons of bombings’ effects learned from the Great War and from colonial air operations during the 1920s and 1930s had created false impressions, but that these were nonetheless the basis for many of the plans and decisions which had been developed. He makes no attempt to excuse these shortcomings, simply wanting to explain their naissance. Indeed, Slessor was head of the Plans branch in the Air Ministry from May 1937 until 1941. As such, he participated in and was to some extent responsible for the development of the major expansion plans for the RAF. In his memoirs, The Central Blue, he openly admits that senior aviators were working in somewhat of a vacuum, lacking expertise and experience. ‘But we really did not know anything about air warfare on major scale. In a memorandum of April 1938 urging the acceleration of Scheme K [one of several pre-war expansion plans] we wrote “the air staff are faced with a peculiar difficulty, in the absence of any real experience of air warfare between first class powers.”’19

Strategic Bombing 17

Given the fumbling going on in London, Ludlow-Hewitt was perhaps too honest in his assessment of the Command’s shortcomings, making it clear he feared that bombers could not operate safely over Germany and that navigational proficiency was below expectations, with more training needed. ‘It was this mixture of percipience, persistence and pessimism on Ludlow-Hewitt’s part which doubtless inspired the Chief of Air Staff and the Secretary of State with the belief that his post perhaps called for someone younger and more optimistic.’20 That younger person was Sir Charles Portal, who would spend just six months as commander of Bomber Command before moving to the top post in the RAF. Prior to this, Portal had had an impressive career. Born in 1893, he had joined the British Army in August 1914 as a dispatch rider. By the fall of that year he had been promoted to second lieutenant for his outstanding efforts and the following year, transferring to the RFC, he became an observer in No 3 Squadron, one of a number of units under the command of then Lt Col Hugh Trenchard. The following year he took pilot training and then in June 1917, after an interview with Trenchard, he was promoted to acting Major and given command of No 16 Squadron, again a reconnaissance unit, operating in support of the Canadian Corps. He later participated in night bombing operations. By the end of the war, Portal, now twenty-five years old, had been promoted once again and put in command of a training wing in the UK. He would continue to rise quickly in rank and to hold positions of responsibility until, in 1939, he was made the Air Member for Personnel, in charge of all personnel issues throughout the RAF.21 By 1939, the structure of the RAF and of the Air Ministry was generally set, although there would be modifications throughout the war. Under the Chief of the Air Staff there were a number of Air Officers Commanding-in-Chief (AOCinCs), each of whom was responsible for a command. There were also a number of overseas commands such as in the Middle and Far East. Until 1934, all units based in the UK had been part of a single organization called the ‘Air Defence of Great Britain,’ but as new squadrons came into being this structure was found to be increasingly unwieldy and so new commands were established in 1936.22 The three major home-based fighting commands were Fighter, Bomber, and Coastal, the latter providing air force support to the navy, principally for anti-submarine duties. The Air Ministry itself was a large organization responsible for all aspects of the air force that it administered (appendix 3 contains a number of organization charts). Command headquarters were generally set up such that the organiza-

18

The Science of Bombing

tion and conduct of flying operations was the purview of an ‘Air Staff’ (a term also used to describe personnel working at the Air Ministry), under the Senior Air Staff Officer (SASO), while all supporting activities were handled by the Administrative staff run by the Air Officer Administration (AOA). The philosophy of this approach will be examined in more detail in the following chapter, but the point should be made here that, beneath this major division, the actual structure, functions, and reporting channels within the two branches could and did vary within particular headquarters. As one officer at 5 Group Headquarters, one of Bomber Command’s subordinate formations, pointed out during the war: ‘There is no standard system in Bomber Command for the allocation of Air Staff duties at groups and this differs considerably from one Group to another. ... As a result [of shortages of qualified personnel] allocation [of work] is often governed by personalities rather than by the basic principles of sound organization.’23 Mastering the changing structure of the Bomber Command staff and the plethora of acronyms and abbreviations to describe rank and functions, as well as the comings and goings of incumbents, would surely have been be a significant challenge (these are detailed in appendix 4). These position titles were a mix of ranks and functions, a relatively confusing system; however, the constant throughout the war was the functions. As of 1941 the major appointments were the SASO, the senior member of the Air Staff and effectively the second in command of the Command; the Group/Captain Operations (G/C Ops), charged with organizing all aspects of day-to-day operations; various Operations staff sections; the Chief Signals Officer (CSO), responsible for all aspects of telecommunications and in particular the radar staff (under the Wing Commander Radio Direction Finding (W/C RDF)), who were charged with navigation devices; the Group Captain Training (G/C Trg) who coordinated all operational training for crews coming into the Command; the Senior Intelligence Officer (SIO); the Wing Commander Navigation (W/C Nav), whose mandate covered the standardization of navigation practices; the Command Armament Officer (CArmO), who dealt with bomber defensive weapons as well as bombs; and the Wing Commander Photos (W/C Photos), who coordinated the technical and procedural aspects of bomb release photos. Added later were the Deputy SASO (D/ SASO), the Chief Intelligence Officer (CIO), and the Chief Engineering Officer (CEngO). By 1944, the Air Staff had been somewhat reorganized to include a Group Captain Plans (G/C Plans), who acted as the Air Staff second in command and was responsible for all planning for future activ-

Strategic Bombing 19

ities, as well as the flying control section (today’s air traffic control) and the air/sea rescue staff. By this time, too, the Group Captain Operations had been ‘up-ranked’ to Air Commodore Operations, this reflecting the much increased complexity of operations and the increased size of the Command as a whole.24 The Air Ministry did not control the day-to-day operations of the Commands, but rather issued directives that, in the case of Bomber Command, described the targets which were to be attacked. Detailed planning based on weather, distances, threats, and forces available was left to the staff at the Command headquarters. The first directive was issued in April 1940 shortly after the German invasion of Scandinavia, and Sir Charles Portal received four directives during that July.25 Directives were often viewed by those receiving them as overly specific; commanders did not appreciate being told what to attack and when and how to do it. For example, the directive of 13 July 1940 (reproduced at appendix 5) caused Portal to respond critically that the prescribed targets could not be adequately engaged due to isolation, distance, and crew ability. Moreover, he argued, attacking isolated targets meant that bombs which missed the aiming point would be wasted. Portal’s concerns were indicative of the debates at the strategic level that would be waged between the Command and the Ministry throughout the war. In this case, Portal felt that ‘in the Bomber Command we have the one directly offensive weapon in our whole armoury, the one means by which we can undermine the morale of a large part of the enemy people, shake their faith in the Nazi regime, and at the same time and with the very bombs dislocate the major part of their heavy industry, much of their chemical industry and a good part of their oil production.’26 Why, he wondered, should he be attacking the Luftwaffe and its infrastructure? Perhaps part of Portal’s frustration came from the fact that crews were not causing much damage, regardless of the target or the conditions encountered during the raid. By December 1939, the limited daylight attacks made by the Command against military targets had frequently led to loss rates of 50 per cent, ten times more than the Command could absorb and maintain the ability to continue. These early raids employed only twenty to thirty aircraft, and attacks took place at intervals of several days or even weeks. Still, by Christmas, crew casualty rates meant that day precision attacks had begun to look impossibly costly.27 By comparison, night operations during this period had experienced relatively few casualties. But this did not mean that these attacks were effective or efficient. ‘“The real constant battle” the [4] Group Commander pointed

20

The Science of Bombing

out, “is the weather. ... The constant struggle at night is to get light on the target,” and he foresaw “a never ending struggle to circumvent the law that we cannot see in the dark.”’28 Added to these problems, of course, was the possibility of the Germans mounting an effective defence as well as the as yet unrecognized difficulty with navigation.29 All of these, as shall be seen, were to plague the bombers, to a greater or lesser degree, throughout the war. While it perhaps seems implausible by today’s standards, the RAF official history records that ‘despite the long-range attacks which had been planned from 1937 onwards, little or nothing was done to tackle the problem of air navigation, and at the outbreak of the war Bomber Command was not far removed in its technique from the days of open cockpit flying when navigation was simply a question of instinct and map reading.’30 At the end of 1939 aids for night navigation consisted of maps, astronomical sextants, and direction-finding radios, the latter of limited accuracy. Moreover, this equipment required considerable training. Overall, ‘the standard of navigation training did not justify the hope that such skill would be forthcoming. ... In fact, unless they could constantly pin-point their aircraft by direct observation, the night crews were seldom sure of their position.’31 The official history does indicate that there had been sporadic discussion about these problems and possible technical solutions had been suggested, but decisive action was not taken until the problem was recognized by Downing Street in 1941.32 As early as 11 October 1939, A/V/M Arthur Coningham of 4 Group had reported that crews were having problems locating their targets at night. Coningham, who would go on to much success in the Desert Air Force, observed that the difficulty of pinpointing the objective varied with illumination, cloud cover, searchlight dazzle, and attack height. He suggested that crews be given some freedom to bomb a target using an identified ground feature as a reference point, even if that point was not the actual target. Still concerned about casualties to civilians, the Air Ministry was not persuaded to accept this tactic as it would not be as accurate as if the actual target was identified. At this point in the war, precision attack was the only accepted policy. The debate over effective bombing practices continued through the winter and resulted in a number of trials. One such attempt, on 19 March 1940, was conducted under near ideal conditions with virtually all the attacking force reporting accurate bombing. The results, however, said otherwise, as reconnaissance photos indicated no obvious damage. Bomber Command’s conclusion, issued on 10 April 1940, was that ‘Our general opinion is that under war condi-

Strategic Bombing 21

tions the average crew of a night bomber could not be relied upon to identify and attack targets at night except under the very best conditions of visibility, even when the target is on the coast or on a large river like the Rhine.’ Worse, their statement continued, ‘if the target has no conspicuous aids to its location, very few inexperienced crews would be likely to find it under any conditions.’33 Just days later, despite these conclusions, the Air Ministry instructed Bomber Command to concentrate on night operations in an attempt to avoid the prohibitive casualties associated with the relatively more precise daylight attacks.34 Charles Portal had long been an advocate of strategic bombing, but when he became CAS on 20 October 1940 he was forced to face the prohibitive losses of unescorted day bombing and the navigation problems of night bombing. Portal’s solutions – night precision attacks where illumination permitted and area attacks when not – ‘met not only the tactical but the emotional need of the hour. It fitted with Churchill’s cruder concept of taking revenge for the damage done by the Luftwaffe. But above all it suggested a profitable way of fully employing the only British force then capable of striking directly at Germany.’35 Despite the poor results and Portal’s non-precision option, even at the end of 1940 the belief still existed that precision bombing could be conducted at night. While there were now eight months of evidence which tended to support the conclusions reached before the war that effective night bombing would be problematic, this evidence was viewed with a ‘strong element of doubt’ and any problems were seen as atypical.36 Calculations and conclusions were based, for the moment, on the premise that an average crew under moonlight conditions would be able to bomb within three hundred yards of the target, that figure being ‘exactly the margin within which, before the war, it had been hoped that the day bombing might be confined.’37 Many, but not all, of the senior leaders within the Command felt that problems were being worked out. Moreover, raid reports claimed outstanding navigation and bombing accuracy, and while these reports may seem frequently implausible to the modern observer they were in fact believed. The disappointing results recorded by the few bomb release cameras available to the Command were simply discounted.38 There appeared to be an almost complete lack of policy and procedure for the coordination and conduct of night bombing operations, with, in some cases, individual crews being left to decide how they would mount an operation.39 Only the analyses of confused operations, such as the raids on Mannheim in December 1940,40 done by Mr D.A.C. Dewdney, an oil-targeting expert assigned to Bomber

22

The Science of Bombing

Command since the summer of 1940, as well as repeated comments by A/V/M Coningham, began to attract attention and gain credibility in the early weeks of 1941. As was much later noted, there were simply too many insurmountable challenges to accurate navigation and bombing. ‘The “never ending struggle to circumvent the law that we cannot see in the dark” was now to begin, but the formidable legacy of the previous neglect could not be overcome in a day.’41 It was this inability to hit much of anything that had started the nation’s leaders along a path towards adopting an area bombing policy.42 By the fall of 1940 Portal had figured out the main problems with strategic bombing. He had first to build up a sufficient force and then be able to use it effectively even when conditions were less than ideal. He must also find ways to make bombing more accurate, but until solutions could be implemented, he recommended choosing objectives that could be more easily found and attacked. He recommended the development and fielding of electronic navigation aids and the use of expert crews who could accurately identify and mark targets for the remaining crews who followed behind. To make target identification easier and bombing more efficient, Portal, following Churchill’s lead, recommended that industrial areas of towns or industrial communities become the aiming points. Accuracy would be more easily achieved, and any bombs missing the intended industry would still do some worthwhile damage. Portal would later say that any civilian deaths came not as a primary objective, but rather as a by-product of using the bombing force as efficaciously as possible to destroy German war potential.43 On 31 July 1941 the Chiefs of Staff Committee issued a statement indicating their intention to pass to a ‘morale’ policy for bombing where attacks against population centres would eventually so demoralize the populace as to bring down the German state.44 This shift in policy still did not help quantify or resolve the problems facing crews each night. Professor A.F. Lindemann (later Lord Cherwell), a First World War aviation researcher who had been kept from flying because of his eyesight and later an Oxford physicist, had joined Churchill when the latter was named First Lord of the Admiralty in 1939. While at the Admiralty the scientist was asked to head a small statistical section and within a few weeks his work led him to challenge senior naval officers to back up their various tactical claims with data.45 Now, as the prime minister’s scientific adviser, he was able to suggest the same approach to clarifying the problems of Bomber Command. The CAS welcomed the proposal for a review of air photos.46

Strategic Bombing 23

It was Mr D.M. Butt of the War Cabinet Secretariat who set about to examine the available evidence of bombing accuracy. This data consisted of 650 bomb release photos taken between 2 June and 25 July. Butt reported that ‘of those aircraft attacking their target, only one in three got within five miles.’47 This was the average, but the results varied based on weather conditions; for example in full moonlight two attackers in five got within five miles of the target, but in thick haze that ratio dropped to one in fifteen. Target location was also a factor, the German industrial heart of the Ruhr valley being the worst locale in which to try to find the objective, with only one aircraft in ten coming within five miles. But these were values for crews that actually claimed to have attacked the target and only 66 per cent of the 6100 sorties reviewed fell into this sub-group. Moreover, Butt reminded those reading the report, a radius of five-miles would, except for Berlin, include considerable open area around the target community and thus not all bombs falling in the fivemile ring would necessarily have caused any damage at all.48 The report generated considerable debate. The senior airmen doubted its conclusions, but could not counter the facts or the prime minister’s concern.49 On 11 September, Portal minuted Churchill acknowledging the problems brought out by these conclusions and suggesting solutions. ‘Much more,’ he said, ‘must be done to improve the accuracy of our night bombing. I regard this as perhaps the greatest of the operational problems confronting us at the present time.’50 His list of curatives included the introduction of pathfinder crews and the development of suitable target-marking devices with which they could identify the target to following aircraft, the development of radio navigation systems, and the establishment of an operational research section at Bomber Command.51 That most scientific of the three services would now turn directly to the scientists to find the cure to their problems.

2 Operational Research: Finding a Solution through Science

Operational research was a relatively new concept, but it had already proved its worth across the RAF. Now, in an attempt to find ways of resolving its problems, Bomber Command chose to follow the experience of Fighter Command and Coastal Command in setting up an Operational Research Section within the headquarters. This section, staffed by scientists from a variety of backgrounds, would collect data and review the facts, much as Dewdney and Butt had done. At the same time their duties would include analysing the data and presenting the commander and his senior staff with conclusions and recommendations as to what courses of action might be available to make good the various problems confronting them. These scientists would be working in a newly defined research field – operational or operations research – which would subsequently develop into an important means by which public and private sector organizations are able to maximize their potential. Operational research has been described as the way in which ‘natural scientists study the consequences of decisions in complex situations not only in business, commerce, and government, but also in the military field. The research work in this general area has come to be called operational research and stems from pioneer studies carried out just before the Second World War by groups of scientists.’1 Researchers take into consideration all the factors, variables, and constraints so that it is possible to discern why things happen as they do and what can be done to change outputs and thereby solve problems. Operational research does not, however, deal with what have been called ‘trivial’ problems, those situations where all the facts are available to the decision maker. ‘[These sorts of problems may be difficult] for him to control as an executive, but arithmetically and structurally the system is

Operational Research 25

trivial and the research scientist is not concerned with decision-making problems of this nature or type.’2 OR has been called ‘a serious attempt to bring scientific methods to the aid of business problems.’3 According to this hypothesis, its origins lie in the expansion of enterprises to the point where their complex nature results in multi-faceted problems the solving of which requires specialized assistance at the central executive level of the whole organization. A method of resolving issues spanning various ‘compartments’ is needed. The formal definition of the field, developed in the 1960s by the Council of the United Kingdom Operational Research Society, covers all of these elements: ‘the attack of modern science on complex problems arising in the direction and management of large systems of men, machines, materials, and money in industry business, government, and defence.’4 The Council says ‘the distinctive approach is to develop a scientific model of the system, incorporating measurements of the factors such as chance and risk, in order to predict and compare the outcomes of alternative decisions, strategies, and controls. The purpose is to help management determine its policy and actions scientifically.’5 Arguably this very broadly based and comprehensive analysis, however, relates better to business or public policy issues than to the exigencies of war, when time pressures are extreme and information about the enemy’s intentions and evolving capabilities is particularly incomplete.6 An alternative conception of OR is much more straightforward and useful in this context: ‘[OR is] concerned with allocation and planning in complex situations involving scarce or limited resources.’7 This definition would seem to fit military circumstances very well, for the commander and senior staff are continually involved in managing scarce and costly resources in situations often shrouded by the fog of war. Implicit in these definitions is the relationship between the scientist and the executive or the commander – the decision maker. Indeed, presenters at a 1958 NATO conference described operational research as ‘the use of scientific method in providing executive departments with a quantitative basis for decisions regarding the operations under their control.’8 In solving problems scientists and executives do not work in isolation; rather, ‘the good planner, the good executive with imagination, has to be fully informed about ... capabilities and limitations. ... The good operational research worker will inevitably find himself drawn (given a not too hostile environment) into the planning.’ 9 Looking at it another way, one might say that there is little point in the scientist going off on his or her own to devise marvellous models and solutions which

26

The Science of Bombing

have no chance of implementation. This symbiotic relationship between scientist and executive, and between reality and possibility was commented on in 1947: ‘Past operations were studied to determine facts, theories were elaborated to explain the facts, and finally the facts and theories were used to make predictions about future operations.’ The value of a quantitative assessment of particular circumstances could be found in the opportunity, frequently, to confirm the commander’s intuitive conclusion with statistical analysis.10 While the foregoing definitions in part suggest that operational research drew on civilian science and later came to play a large role in business and government management, its origins lay to a great degree in the development of radar for the air defence of Britain in the 1930s. There the scientists who had developed the theories and processes of radar detection that worked well enough in controlled conditions found that the technology did not seem to be as effective in real, i.e. operational, settings. As a result, small groups of scientists called ‘radar operational research teams’ were established to examine the use of radar under actual conditions. Sometime thereafter the term ‘radar’ was dropped from the teams’ title. By the time war broke out, these teams had carried out a number of trials related to radar, but more importantly their value had been recognized by the Royal Air Force and the other services. And while these first scientists had been expert in the technology they helped operationalize, the whole notion of using pure science and scientific process to work out combat tactics was something new. ‘It was found that, once they were able to understand the technology of the systems they were studying, they could make significant improvements to the way in which these systems were controlled.’11 In other words, it was not a matter of being expert in the subject matter, but rather in seeing how the system was intended to work and making suggestions about how to make that intent possible. While it is one thing to understand the theorists’ definitions of operational research, it is important to review how the RAF defined the science as it was practised during the war. The British Air Ministry monograph The Origins of Operational Research in the Royal Air Force suggests that ‘of the many definitions [of operational research] offered, possibly the most concise is that it is “numerical thinking about operations, with the aim of formulating conclusions which, applied to operations, may give a profitable return for given expenditure of effort.”’12 Four branches of OR had evolved during the war. The first dealt with an assessment of weapons and equipment to determine how well a system functioned in

Operational Research 27

operational use; i.e. how often did it break down. The second looked at operations to see how well a piece of equipment worked for given tactics, or alternatively to what extent tactics dictated the form of weapon to be employed. A third branch was used to predict the outcome of operations either at the tactical or strategic level, particularly in terms of how these results might influence policy. Finally, a fourth type of work was conducted to measure the efficiency of the units which employed various equipment.13 The RAF history attributed the phrase ‘operational research’ to Mr A.P. Rowe, who was actively involved in various scientific committees during the pre-war years and who would become the head of the Ministry of Aircraft Production’s Telecommunications Research Establishment, an organization charged with developing new devices for navigation as well as communications. In 1938 he used the term when describing the work being done on radar by Mr E.C. Williams and a team who had been given ‘the problem of assessing and passing on rapidly the information that came through the Chain Home (C.H.) [radar] stations; in other words the process of plotting, filtering and telling.’ 14 This was in fact one of the ‘teams’ referred to above. Sir Robert Watson Watt, the pioneer of the technical aspects of radar, put forward several definitions of OR. In one case, he saw a role for OR in confirming tactical efficiency. One needed, he wrote, to examine quantitatively whether the user organization is getting from the operation of its equipment the best attainable contribution to its overall objective, what are the predominant factors governing the results attained, what changes in equipment or method can be reasonably expected to improve these results at a minimal cost in effort and in time, and the degree to which variations in the tactical objectives are likely to contribute to a more economical and timely attainment of the overall strategic objective.15

In another instance he said that radar was the genesis of this new field of research where scientists were moving beyond the study of natural phenomena to the analysis of problems as they occurred in the field. There is no previous work that fits the simple criteria for classification as Operational Research, to wit that it was an investigation carried out, by scientific method, on actual operations, current, recent or impending, at the request of those responsible for the initiation or conduct of the operations, and explicitly directed to the better, more effective and more economical conduct of similar operations in the future.16

28

The Science of Bombing

At its essence, Watson Watt felt that operational research sought ‘maximum effect from available resources.’17 Professor P.M.S. Blackett, one of the best-known OR scientists to come out of the British services and discussed in more detail below, wrote a short text on OR in October 1941. In ‘Scientists at the Operational Level’ he described the functions of an OR section.18 First, it was up to OR staffs to analyse data not gathered as part of the work of the scientific establishments. This included ‘the data provided by the operational staff such as signals, track charts, combat reports, meteorological information, etc.’ The data was best reviewed by scientific officers simply because the average military staff officer did not have the education or scientific expertise to effectively analyse the material. Second, the scientists were best placed to provide liaison with the scientific and test establishments. They knew both what problems were facing the operational units and which of the expert establishments were best able to solve these problems. But Blackett did stress the limits of the scientists’ authority and responsibility. ‘It was not the function of operational research sections to act in any other than an advisory capacity. The headquarters staff carried the responsibility for making decisions and could use or ignore the scientific analysis of any problem as it wished.’19 The origins of what became operational research can be traced back to the integration of scientists into the defence ministries and armed services that began in the 1920s. In the case of the RAF, the Directorate of Scientific Research came into being in 1924 under the aegis of the Air Member for Supply and Research. That, of the three Services, the RAF later seemed to most effectively exploit science can be attributed to four factors. First the air force was totally dependent on science for the operation of its major ‘systems’ – aircraft. Second, chance brought radar to the RAF, marrying capable scientists with open-minded officers like Air Chief Marshal Sir Hugh Dowding, Air Officer Commanding-in-Chief of Fighter Command. Third, the RAF as a new service was not bound by tradition. Finally, initial successes and a culture of open-mindedness attracted young and capable minds. The ‘boffins,’ as these young men and women came to be known, were to become very much a part of the RAF’s success.20 The first Director of Scientific Research was Mr H.E. Wimperis, whose mandate was to provide advice to the Secretary of State for Air. This was accomplished through an advisory body which included many of the scientists who would later play leading roles in the Second World War; among them were Professor Lindemann from Oxford and Mr T.H. (later

Operational Research 29

Sir Henry) Tizard of Imperial College. By the mid-1930s the Air Ministry had established a laboratory at Imperial College which looked at such technologies as aviation fuels and navigation devices. At the end of 1934, Wimperis suggested the formation of a committee to examine the use of scientific advances in the defence against air attack. This group became known as the ‘Committee for the Scientific Survey of Air Defence’ and included Tizard as chairman, as well as Professor Blackett, Mr Wimperis, and Mr Rowe as members. It was also Wimperis who engaged the services and the support of Watson Watt in early 1935. The latter was then the senior researcher in the Radio Department of the National Physical Laboratory. Watson Watt and other scientists worked on the technical aspects of radar.21 The first examples of operational research were related to work done in 1936 under Tizard’s direction which aimed to identify the most effective method for employing radar in the interception (as opposed to detection) of attacking aircraft. At that time he set up a mixed civilianmilitary team at Biggin Hill, including a young scientific officer, Dr Basil G. Dickins, described in more detail in the following chapter, to conduct experiments into this question. Dickins’s task was to develop a methodology that would optimize the use of data provided by the radar receivers in assisting controllers to effectively direct fighters against incoming enemy aircraft. In hindsight, the experiments were critically important for what was to come. ‘Firstly, they developed the technique that won the Battle of Britain and, secondly, they marked the beginning of an era of close cooperation between the serving officer and the scientist in the study of operational problems.’22 While there had previously been resistance to using scientists to deal with operational problems, by this time Fighter Command was in the midst of implementing a demanding new technology – one which could influence national security – and could not fail. Now, work continued in refining the processes of analysing the data coming in through the radar system itself and in using that refined information to direct the fighters towards the threat.23 Key players in these activities were Mr Rowe, Mr G.A. Roberts, and Mr Harold Larnder. The latter’s work would later be citied as an exemplar of the effective OR scientist. ‘He caused his staff, a number of whom were intellectually his superior, to work effectively together; he had a gift of anticipating what was to be required in six months’ time; and above all he got on well with service officers of all ranks, won their complete confidence and led his staff to work closely with them.’ Fighter Command’s AOCinC, Sir Hugh Dowding, later sent

30

The Science of Bombing

Larnder a note saying ‘“Thanks. This war will be won by science thoughtfully applied to operational needs.”’24 Various teams of scientists had been working at different locations but, as 1939 advanced, Dowding asked for a permanent section to be set up at his Headquarters at Stanmore in north London. What was called the Radar Research Section, and after February 1940, the Stanmore Research Section, was thus in place to provide assistance to the Commander of Fighter Command at the outbreak of war.25 The section was involved in the complete range of Fighter Command activities from the outset and the success of the radar system during the Battle of Britain in the summer of 1940 has been attributed to the careful analysis by these scientists of the response to every raid sent against Britain in the first months of the war. In particular, failures to make interceptions were looked at in detail.26 Scientists were always present in the control rooms and played a part not just by observing and gathering data, but also in providing advice and guidance to the air force controllers. Much of this work fell under the supervision of Mr Roberts, who had become the ‘second in command’ of the section. His work was instrumental; for example, he was the lead author of the 1940 edition of ‘Fighter Command Battle Orders,’ a volume that detailed the duties of personnel working in operations rooms.27 In 1941 he would move to Bomber Command and become the number two scientist of that ORS. The scientists also provided advice on the selection of personnel for employment in the highly stressful controller positions (not unlike current air traffic control work) and on ways to optimize the passage of information within the complex chain of control teams located in the radar sites and at Group and Command control centres. In doing so they helped to develop an appreciation of and solutions to the problem of system overload which became noticeable during the actual period of the Battle. Other areas of research included determining how to make best use of automatic ‘Identification Friend or Foe’ (IFF) systems and methods of night interception once the Luftwaffe abandoned its day attacks.28 Early operational research functions in Coastal Command were also linked to radar, in this case the effective employment of ASV (Air detection of Surface Vessel) radar. Once settled in, however, these scientists began looking at all aspects of the command’s activities. By mid-1941, when an ORS was formally established under Professor Blackett, there was sufficient background information to comment on what seemed to be the command’s Achilles heel – the very low rate of sinkings achieved

Operational Research 31

by attacks on U-boats. The scientists’ report was considered one of the ‘classics of operational research literature’ (and is cited in many postwar OR monographs, whether they deal with military matters or not). The scientists were able to point out that both tactics and weapons were flawed. It was much better, they said, to attack submarines on the surface or those that just dived, rather than going after those which were well submerged. As well, they concluded that the lethal radius of the 250-pound depth charge was only fifteen to twenty feet and so it was critical both to set the detonation for fifty rather than one hundred feet (since the submarine would be closer to the shallower depth having just dived) and also to increase the quantity and power of the explosives used. Implementation of these changes resulted in an increase in sinkings of near-surface subs from 2–3 per cent in 1941 to 40 per cent by 1944.29 This model success of OR came about when commanders were at a loss to identify the reasons behind their failure to sink submarines. It was an archetypal operational research problem and solution. It dealt not just with the sinking of submarines but also with the underlying and more important issue, which was to mitigate the threat to convoys which were themselves vital to the greater war effort.30 In other words, it is not necessarily the immediate effect which is most critical, but rather the relative effectiveness and more importantly the relatively efficient use of scarce resources in addressing the larger task. While the scientists had obviously scored a very real victory in solving the depth charge problem, it had not been without debate and there were other occasions when their solutions were not well received. Comments by Air Marshal Sir Jack Slessor, AOCinC of the Command in 1943 and a post-war CAS, show both sides of the command-science relationship. Describing the overall contribution of his ORS to the command, he was full of praise for these ‘indispensable’ scientists: ‘They proved beyond any doubt that the scientifically trained, analytical mind, applied to any problem, could produce valuable results; and they frequently surprised me by telling us, not only what we did not know, but what otherwise I should never have realized was something we ought to know about an operational or administrative problem. ... A good O.R.S. is a vital part of the fighting staff in the operational headquarters.’31 Slessor’s praise was not unqualified, however, and he also had strong reservations about accepting a purely scientific solution. In reviewing the scientists’ answer to the allocation of anti-submarine patrol aircraft to the Battle of Biscay, for example, he took exception to the recommendations provided by researchers working for the Admiralty, including by

32

The Science of Bombing

this time the much-respected Blackett. Slessor’s account of the circumstances points to the tensions between the classic heroic approach to command and a more management-like approach based on statistics. The object they had set themselves was ‘to prevent U-boats from operating in the Atlantic altogether,’ which itself I thought a bit ambitious. If the Air Ministry for instance had demanded, successfully, a fighter force capable of preventing any bomber from ever penetrating into our airspace, there would not have been much left for anything else. The Combined Chiefs of Staff had themselves defined the object as ‘the reduction of sinkings to a level where they do not constitute a serious limitation on our offensive capability,’ which goodness knows was a sufficiently difficult job.32

Slessor went on to argue that battles are not won by irrefutable and absolute mathematical models but rather by men and women led by commanders who must weigh the evidence, test it against their experience, and make hard decisions. In this case Slessor reported that the submarines were defeated, but with only a fraction of the 260 aircraft recommended by Blackett.33 Whether or not we can say that this was an example of a lack of trust in OR is difficult to say. There was a strong element of ‘I’m the practitioner and I know best’ in this, while from the scientists there was perhaps a belief in a purely ‘mathematical’ solution to the problem, one which simply could not be accepted. Slessor says that when the scientists’ plan was presented to the War Cabinet one minister said ‘C’est magnifique, mais ce n’est pas la guerre,’ suggesting that it was too rich a solution given everything else that was going on.34 Coastal Command’s most brilliant ORS chief had been Patrick Blackett. Born in 1897, Blackett had a brief naval career, participating in the Battle of Jutland. After the First World War he took his bachelor’s degree in physics from Cambridge in 1921 and subsequently advanced in academic circles until the outbreak of the war, when he began working in OR-related posts. In 1945 he returned to academic research, winning the Nobel Prize for Physics in 1948.35 Having been involved in operational research from 1940, he moved from the air force side of air defence to the army and set up the first OR establishment for the land force as an arm of Anti-Aircraft Command. Among other things, Blackett’s scientists studied the question of providing data to the gun batteries and found that the basic organization was flawed – there could be twice the number of guns in each battery without overloading the supporting radar opera-

Operational Research 33

tors.36 These experiences and his subsequent work in Coastal Command were behind his short précis written when he transferred yet again to the Admiralty at the end of 1941. That document, ‘Scientists at the Operational Level,’ has already been referred to. It became, he later wrote, widely quoted and appeared ‘to have some influence on the setting up of similar organizations.’37 In it he had concluded that too many of the best scientists were being misemployed developing new weapons and systems rather than making sure that those already in production were doing what had been intended; this latter employment of scientists would do much to support the operators of those weapons systems. More scientists, he felt, should be transferred to the fighting commands in order to get on with this more important task.38 By the time he set these ideas to paper at the end of 1941, just such a move was already underway in the rapidly expanding Bomber Command Operational Research Section.

3 Boffins at Bomber Command: September 1941

By the time the Butt Report had been published, action was in train to provide Bomber Command with its own operational research capability. A small number of scientists had already been examining bomber operations, and from them the nucleus of the Bomber Command ORS would be drawn. These men and women, and their uniformed counterparts, had then to work out exactly what needed to be researched, how the data would be gathered and interpreted, and how the conclusions of the research could be used to maximize return on the huge investment in national, and indeed Allied, resources that Bomber Command consumed. By the time Sir Arthur Harris took over as AOCinC in February 1942, much had been accomplished. Investigations into bombing issues had begun as early as 1937. In January of that year the Air Ministry set up a parallel committee to the Committee for the Scientific Survey of Air Defence, the former to deal with offensive operations.1 The membership of this new Committee for the Scientific Survey of Air Offence and that of the defence committee was the same; Tizard ran both groups and decided which issues were to be considered under which mandate. In essence, while the Defence Committee dealt with matters relating to Fighter Command and the radar system, the Offence Committee looked as issues such as “‘the scatter bombing of aircraft from above,” bombs, aircraft armament and aerial gunnery.’2 At the end of September 1939, the two committees were merged into the Committee for the Scientific Survey of Air Warfare.3 At the recommendation of the Committee, the first bombing study was conducted almost a year later. In July 1940, Mr A.E. Woodward-Nutt of the Directorate of Scientific Research (DSR) of the Ministry of Aircraft Production (MAP) attempted to determine ‘the causes of bomber losses

Boffins: September 1941

35

[while conducting] a critical examination of the phenomena reported by aircrew during operations.’4 The report was titled ‘Interception of British Bombers at Night (May–June 1940)’ and was well received. It was subsequently decided that Dr Basil G. Dickins, also working for the DSR, would take over this project and produce monthly reports; these were initially called ‘Phenomena connected with Enemy Night Tactics’ and later became ‘Report on Losses and Interceptions of Bomber Command Aircraft.’ These synopses included both statistics and ‘various conclusions which it was possible to draw from time to time.’5 When it became apparent that further analysis could be done on the data which would permit a better understanding of bombing operations, it was decided to attach a scientific officer permanently to Bomber Command. Miss K.M.M. Coggin was therefore posted to High Wycombe in August 1941.6 As described in the previous chapter, by the summer of 1941 the value of scientific work, in Bomber Command and the other commands, had been amply proven, and the Air Ministry ‘gave consideration’ to normalizing the establishments and mandates of the scientists who were largely still assigned either to the Telecommunications Research Establishment or in some cases the MAP.7 Now each command would have its own operational research staff reporting directly to the commander. As a result, additional staff were transferred to the DSR and then seconded to the Air Ministry. As it happened, in August Sir Richard Peirse, AOCinC Bomber Command, formally requested an OR section and one was established on 1 September. Given his familiarity with the issues, Dr Dickins was named as section head and those personnel who had been recently attached to the Command were assigned as the nucleus of his staff. The commander-in-chief set out the mandate of the new section in broad terms: ‘the general study of operations with a view to determining how the efficiency of operations in terms of bombs on target per aircraft lost could be increased. This objective remained the aim of the O.R.S. throughout the war, [although studies in maintenance efficiency were added to the mandate in later years.]’8 By the end of the month, there were seven scientists assigned to the OR section, but it was apparent that this number would be inadequate to deal with the scope of the research needed.9 Dr Basil G. Dickins, the head of the Bomber Command ORS throughout the war, had been born in 1908. The archives of Imperial College London indicate that he entered that college in 1927, after attending Mount View School and King Alfred School, Streatham, in south London. He studied physics, earning his BSc degree in 1929, and continued

36

The Science of Bombing

with postgraduate study on the topic of ‘Conduction of Heat through Gases,’ on completion of which he was awarded the Diploma of Imperial College (DIC) and the Associateship of the Royal College of Science (ARCS), in 1932.10 Soon after, he joined the Royal Aircraft Establishment (RAE), a research and development unit within the Air Ministry. After the war he went on to appointments of Director of Guided Weapons Research in 1956, Director General of Atomic Weapons in 1959, and Director General of Guided Weapons Research from 1964 to 1967.11 He passed away in 1996.12 After leaving Imperial College, Dickins had worked in and around defence aviation issues. Chronologically, the first reference to him is in conjunction with the development of radar. In the summer of 1936, Dickins was working as a mechanical engineer at the RAE at Farnborough when he was approached by A.P. Rowe and asked if he would like to work on an air defence problem. The nature of the problem, defined by Tizard but not revealed to Dickins, was to determine how to get intercepting fighters into an attack position against incoming bombers if given fifteen minutes’ warning (with subsequent updates every five minutes) as to the latter’s height, speed, and course. Working with two RAF officers, Dickins devised a set of experiments to try what was ‘almost entirely new,’ – dispatching fighters waiting on the ground towards the intruders based on this data. Although none of the team knew of the existence of radar, Dickins apparently asked Tizard why they could not have updated fixes every minute and was then told that the data could be updated almost continuously. As the experiments progressed successfully, Dickins was able to recommend the use of a vertical plexiglass plotting board, more convenient than the horizontal slate board then in use, based on what he had seen earlier at the Observer Corps Headquarters.13 One gets the sense that Dickins was seen by some of his seniors within the scientific world as an up-and-coming figure. He had been asked by some fairly important and expert people to work on a critical problem, had grasped the essence of it quickly, and had made contributions to bettering the associated processes. It was perhaps not surprising that in 1939, Dickins, now working at the Air Ministry, was added to the Maud Project, dealing with atomic weapons.14 Once again this was no small kudos in terms of the importance of the assignment. Dickins’s contributions were equally valued after his move to Bomber Command. ‘Dickins soon became very close to Harris and Saundby, who sent for him frequently and bombarded him with questions.’15 The creation of the Bomber Command ORS and the importance of Dickins

Boffins: September 1941

37

and his staff within the Command were recorded in the official history: ‘Henceforth the scientific and empirical appraisal of operational problems was increasingly extended to matters which, in the past, had been left to chance, guesswork, or the spasmodic investigations such as those carried out by Mr. Dewdney and Mr. Butt.’16 A rather different picture of Dickins has been painted by Sir (later Lord) Solly Zuckerman, himself a brilliant young scientist, one well regarded in both service and scientific circles, who, after the war, would go on to hold high office in NATO. In 1944, Zuckerman worked on operational planning for Overlord and was intimately involved in the air bombardment plan which called for Bomber Command participation over an extended period. This diversion from strategic bombing did not sit well with Harris. At a meeting on 15 February 1944, Harris brought Dickins to back up his belief that Zuckerman’s plan to attack the transportation network (the Transportation Plan) was unsound. According to Zuckerman, Dickins explained that he had been told that ‘400 aircraft would be expected simultaneously to bomb each target’ and had concluded that these plans would lead to grossly inaccurate area bombing. Zuckerman countered that the targets would require fewer than one hundred aircraft each and could be struck without stirring up dust clouds to hinder the bomb aimers’ view. With this new information, Dickins, claimed Zuckerman, agreed that the Command could handle most of the targets effectively. Zuckerman was not impressed by this volte face. This was the first, but unfortunately not the last time that it became apparent to me that not all scientists who during the war had been drawn into Service posts were as questioning and as independent in their judgements as they could have been. On occasion, they were constrained by assumptions which uncannily fitted their masters’ preconceived ideas. There was also a tendency to defer more to rank than the intellectual competence of the officers they served.17

This may well have been a general observation, but there can be little doubt that Zuckerman intended fully that it be applied to Dickins. It has also been noted that while Dickins attended daily planning sessions, he was not taken on as the AOCinC’s personal scientific adviser as had been the case in other commands. Dickins had access to Harris, but worked directly for the AOCinC’s deputy, Air Marshal Sir Robert Saundby. The conclusion inferred is that ‘Dickins was a highly competent civil service scientist, but did not have the stature or independence

38

The Science of Bombing

of mind of Blackett or Williams.’18 This same critique cites Freeman Dyson, a member of the Bomber Command ORS staff, who felt that the Section was ‘too timid to challenge any essential element of policy.’19 In Disturbing the Universe, a cathartic memoir, Dyson goes further and explains some of what he saw, did, and thought while a junior scientist in the Section.20 It is clear from his comments that he was unhappy with the area bombing policy, with Harris for prosecuting it, and with the ORS for not countering the policy based on research and findings. Dyson had been assigned to study the relationship between aircrew experience and losses. In 1944 he had determined that losses were no longer likely to drop off as crews gained experience. This should have sounded alarm bells for the leadership that something was wrong. From Dyson’s perspective this did not happen, and he implied that his seniors, Dickins included, were either too stupid to realize the situation or simply ignored the evidence. He indicated that it was not possible for him to relay these concerns directly to the crews and that official documents containing these conclusions were highly classified and not revealed to the flyers.21 He likened the crews to the infantry of the First World War, sent out as cannon fodder by uncaring commanders. One of the causes of these losses was the difficulty of getting through the escape hatches of the bombers. Dyson wrote that a colleague, Mike O’Loughlin, had identified this problem, but that his data, like that of losses generally, was largely suppressed. He stated that O’Loughlin had shown that the bail-out rate was worst in Lancaster bombers, otherwise the Command’s most successful aircraft (see appendix 6), but that only his friend was concerned that the Command was withdrawing other types of aircraft from front line service.22 Dyson offered no evidence to substantiate either of these claims, both of which are discussed later in this study. While he was correct that these issues were kept from the crews, he was incorrect, as shall be shown, in his suggestion that the data was ignored or accepted as a cost of fighting the war. Dyson laid the blame for these and other problems at the feet of Dickins, whom he categorized as ‘a career civil servant. His guiding principle was only to tell the commander in chief things that the commander in chief liked to hear.’ Dyson said that this led to promotion and knighthood after the war.23 While Dickins did rise to a Director General appointment by the 1960s, there is no evidence of a knighthood, although he was awarded a Commander of the Order of the British Empire.24 The work of the ORS would go on for four years at Bomber Command Headquarters, located at High Wycombe some thirty miles west of

Boffins: September 1941

39

central London. The headquarters has been described by Charles Carrington, the Army Liaison officer attached to the Command, as being ‘not in the least like a military or naval station, having no sign of martial pomp except a soldier-sentry clicking his heels outside the C-in-C’s office door.’25 Actually located outside of High Wycombe by about five miles, it was hidden away on a hilltop in the Chiltern beechwoods and consisted of three sites. Site One was the underground Operations Room accompanied by two brick office blocks where the staff worked in normal circumstances. Ten minutes’ walk away through the trees stood Site Two, centred on the officers’ mess, described as a country club complete with tennis court. There was a lack of formality in the mess, with no rules of seniority and with people coming and going as they liked or needed. When they dined, they sat where they liked. ‘We used none of those conventions traditionally associated with officers’ messes; I do not recall that we ever drank the “Loyal Toast” [to the Crown].’26 There was a third site located again about ten minutes away, but in another direction, which was a housekeeping area. There were found quarters for the other ranks, as well working areas and a special signalling apparatus needed for Bomber Command to be able to talk to senior and subordinate headquarters. The nature of the work in the headquarters did not typify a military culture, certainly not a combat atmosphere. Staff in large part worked regular office hours, although there was much additional ‘unpaid overtime.’ Junior personnel and NCOs dressed much like the officers in ‘natty uniforms and towny shoes, not battle dress.’ All expected leave in town once their work was over, and parades were unheard of.27 Not only was the working atmosphere different from what an army officer might be used to, but the organization was uniquely ‘air force’ as well. The British army for centuries had organized its headquarters into three departments: the General, or G, staff dealing with plans, operations, and intelligence; the Adjutant-General, or A, staff charged with personnel questions; and the Quartermaster, or Q, staff responsible for materiel and accommodations. In the RAF [these] staff duties were differently distributed: the ‘Air’ Staff dealt with everything that flew and the ‘Admin’ staff with everything that stayed on the ground. No question, but the flying men were the élite and the ‘Admin’ people their servants, yet no Air Staff decision could be taken without ‘Admin’ consideration. Every problem had its ‘Admin’ side which imposed limiting factors and vetoed what was impracticable. The ‘Air’ staff

40

The Science of Bombing made the plans and the ‘Admin’ staff provided the ways and means, the two were not interchangeable. The high command was drawn mainly from officers who were primarily airmen, often air aces, who had acquired some ‘Admin’ experience as experts in this or that technicality on the way up.28 (Italics in original)

What would have been called the Chief of Staff at army headquarters was known as the Senior Air Staff Officer (SASO for short, always known as ‘Sasso’); he dealt with plans, operations, intelligence, air training, and all that part of armament and signals that was airborne. His partner, and in last resort subordinate, was the Senior Administration Officer, or SAO, who dealt with everything else: personnel, equipment, armament and signals (on the ground), food and clothing, aircraft (on the ground), airfields, works, and buildings, as a universal provider.29 Although the ‘fly boys’ might have considered their administrative counterparts as lesser persons, these same underlings ‘carried the flying men on their backs.’30 Writing about headquarters organizations in the years immediately after the war, another observer commented that, like it or not, the majority of officers serving on the Air Staff were pilots and that they displayed a ‘slight air of superiority’ to others. The pecking order, when discussing flying issues, ran from current flyers to past flyers (usually seniors past their flying prime) and finally to non-flyers within the RAF (administrative officers, specialists, etc.).31 Each of the Bomber commanders had his own leadership style. Sir Charles Portal, who would move on to become Chief of the Air Staff in late 1940, sat quietly in the Operations Room and did his work without speaking to the staff or being interrupted. ‘It was his custom to sit in the Ops Room in the mornings, doing his own work and rarely speaking to anyone.’32 Portal’s successor, Sir Richard Peirse, seemed more social and would spend much time in the Operations Room after ‘morning prayers’ (the daily operations briefing), chatting amiably with the staff. Carrington observed that after the briefing, ‘Peirse would wander round the Ops Room having a word with each of the specialists and would sit on the corner of my table and pump me about Army intentions.’33 The key member of the Air Staff was Robert Saundby, who served as SASO, and subsequently Deputy CinC, from 21 November 1940 until the end of the war. Originally an infantry officer, he had become a flyer in 1916. In the inter-war period he had served with Harris on two occasions and had developed a keen commitment to strategic bombing. Saundby had been one of Harris’s flight commanders in Iraq in 1922, where the

Boffins: September 1941

41

latter went about converting the squadron’s transports into aircraft capable of bombing, calling them ‘the first of the post-war long range heavy bombers.’34 Harris thought much of his deputy and had unqualified faith in him. His ideas about bombing were absolutely sound, and naturally I relied enormously on his experience when I first became Commander-in-Chief, as I did throughout the war; he was with me at Bomber Command the whole time, and left at the same time I did. ... Saundby had less side than almost anybody I have known; out of office hours anyone could to talk to him and he would talk to anybody. He was liked by everybody in the Command, and his sociability in the mess where he lunched and also had a drink in the evening before driving to Springfield, was a great help to me, as it was quite impossible for me to see as much of my officers as I should have liked. ... He was one of the hardest workers I have known.35

When later in the war two attempts were made by the Air Ministry to post Saundby away to positions of higher responsibility, Harris would not stand in his deputy’s way, but when Saundby decided to stay on, Harris spoke forcefully to keep him.36 Saundby has been described as a big, jolly, social, and somewhat eccentric officer. ‘He never imposed his authority, never condescended, talking in the same easy colloquial style to all ranks.’ Moreover, while he was ‘the centre of attention’ socially, at the same time ‘he had a natural dignity and no one took liberties with him.’37 While a keen professional, Saundby was also a bit of a renaissance man, with a passion for flora and fauna and a delight for model railroading and fly fishing as well. His family home was far from High Wycombe, so he lived as a permanent house guest with Harris for the duration.38 As the AOCinC’s understudy, Saundby was also responsible for the day-to-day management of the headquarters, including the all-important morning briefing. This gathering would take place at precisely 9 a.m. when ‘the C-in-C appeared to select the target for tonight.’39 Once that was done it was up to the SASO to assemble the key staff in the office block above ground to discuss issues and problems. ‘The discussions were frank and critical: on technical errors in the [previous] night’s operations; improvements in signals; armaments loadings; navigation problems; and the continual necessity for better crew training.’40 Many of these issues would have been investigated by the ORS, but no previous writers have commented on the presence or integration of the

42

The Science of Bombing

scientists, and so one is left to imagine how exactly the RAF officers, the ‘Service’ staff as Dickins referred to them, related to the ORS. From the outset the relationship had official support, as it was the AOCinC himself, Sir Richard Peirse, who had asked for the ORS to be established. If Peirse, however, was keen to have scientific advice, such sentiments were not universally felt, as Slessor’s earlier remarks suggest. There was, apparently, a certain uneasiness regarding scientists among high-ranking officers. ‘The warriors, generals and marshals, were, in their hearts, a little afraid of these master-minds.’41 The problems which Peirse and his staff set before the scientists have been described as very, very broad and this would suggest that from the beginning there was much confidence and trust in the scientists’ ability to find explanations and solutions to the problems which had been plaguing the Command. Dickins described the close working relationship between the scientists and the uniformed staff of Bomber Command. It is important to stress that the Section functioned as an integral part of the Command and worked in the closest collaboration with the other branches of the Headquarters. ... It is pointed out that the results achieved were in many cases the outcome of teamwork between the Service and the scientific staffs. In fact, it is not too much to say that such success as was achieved by the Section was as much due to the receptiveness of the Service as to the efforts of the Scientists.42

An effective relationship was critical because the ORS needed access to all manner of material: ‘tactics, use of and requirements for navigational and bombing aids, weapons effectiveness, training and aircraft maintenance etc.’43 This specialized information came from other branches of the command; thus it was essential for the ORS ‘to hold a special place in the Command organisation and to work in the closest collaboration with the other branches.’44 The Scientists needed all available past information and the maximum forewarning of anticipated operations if they were to provide accurate and useful advice. Dickins was to write that these circumstances were ‘well appreciated’ by both Peirse and later by Harris, as well as by Saundby.45 From its inception the Commander-in-Chief and the Deputy Commanderin-Chief gave the O.R.S. their fullest encouragement and did much to facilitate the conduct of its investigations. Frank expressions of opinion and ad-

Boffins: September 1941

43

vice offered, even though not previously asked for, were always welcomed. This attitude did much to establish in the eyes of the command staff the correct position and functions of the O.R.S. Complete confidence in the O.R.S.’s ability to help was not, of course, established overnight but apart from a few minor exceptions the necessary collaboration and recognition was quickly forthcoming. The incursion of the scientist into the field of operations, was, after all, an innovation insofar as most of the service staff was concerned and in a large headquarters opposition in some quarters was to be expected. Most of it was short lived and the remainder removed by postings occurring in the normal course of events.46

The officers within the Operations branch, the central element of the Air Staff, and generally throughout the headquarters, fell into two categories. At the Group Captain level there were the ‘old seniors [who were] curiously, not air-minded, rather office-bound.’ Then there were the Wing Commanders. ‘Young-ish and growing younger as the war went on, [they] had a quick turnover for they were marked men destined for promotion, fresh off “ops,” newly decorated, hustled through Staff College, and anxious only to complete the tour of staff duty which would qualify them to get back on ops at a higher level. Boyish, open-air men, they resented their confinement “down the hole” and longed to get airborne.’47 One is left to wonder to what extent these young men, whose work experience was framed around finding solutions to problems and implementing them immediately, would get along, despite Dickins’s statement, with civilians seeking systemic solutions based on dry scientific process. Dickins wrote that the scientists were not interested in shortcuts or working outside of the normal staff channels. When a report had been drafted it would normally be reviewed by the applicable Service staff before onward transmission to the SASO or the AOCinC. Dickins also recorded that once the scientists had learned the ropes and developed a credible expertise in the practical and technical issues touching on bombing, ‘no proposal made by the O.R.S. was ever turned down by the command on the grounds of impracticability.’48 Given this central niche, Dickins would have preferred that his OR section report directly to the AOCinC instead of Saundby, though as section head he had direct access to the Commander.49 Other OR Sections did work for the Commander, but this was not the Bomber Command practice; here too the lack of standardization between headquarters can be seen.50 The broad lines of the research problems had been in the minds of

44

The Science of Bombing

the senior flyers for some time. For instance, in July 1941, shortly before the establishment of the Operational Research Section, the AOCinC had visited Hull after a Luftwaffe attack. He noted that very few of the hundred or more attackers had come close to the target and wondered: ‘Was our performance any better? [Peirse said] that he would be delighted to learn that 60% of ours reached the target area, but he feared it was not often more than 40 percent. Of this minority how many – how few(?) – scored a bull’s eye on the aiming point?’51 In addition to bombing accuracy, there remained the problem of minimizing losses, and a second primary focus of the ORS would be to analyse the causes of losses. This task would be based to a large extent on the analysis of crew reports, as well as data from intelligence staffs. ‘Intelligence reports gave, for each sortie, particulars as to the time and height of bombing, enemy aircraft seen and the time and place at which aircraft were seen to be destroyed, together with the details of intercepted wireless traffic.’52 Additionally, much coordination would be required with OR sections at Fighter Command as well as army Anti-aircraft Command OR scientists. It was from concerns over losses, the broad mandate Peirse had given to the scientists, and the more tangible realities described by Butt, that a research program was developed. In Dickins’s words: When the O.R.S. was first formed the Commander-in-Chief gave the branch a broad program of research, covering the problems which he regarded as of the greatest importance. The precise items of research necessary to meet this general directive were left to the Officer in Charge to decide, and in consultation with the senior members of his staff the detailed program was built up as the staff increased. From time to time, the Commander-in-Chief, his deputy, headquarters branches and external establishments would request specific investigations. These would be given priority, but normally the items for research originated in the section itself. A detailed research program was prepared occasionally and submitted to the Commander-in-Chief and the Senior Air Staff Officer for approval and guidance as to priorities. It was also forwarded to the Air Ministry.53

On 17 September Dickins submitted draft terms of reference as well as a letter intended for the Groups describing the main thrusts of the OR system to Saundby, and these were signed and promulgated the following day, alerting the Command to the official activation of the OR section.54 The letter to the Groups provided a broad picture of the section’s purpose.

Boffins: September 1941

45

A Branch known as the Operational Research Section has recently been established in this Headquarters for the purpose of analysing bombing operations with a view to determining weak points in the enemy defence system, to ascertain the cause of casualties so that steps can be taken to reduce them and to assess the effectiveness of bombing attacks. The section will also be investigating various radio problems relating to this Command.55

The letter went on to say that members of the ORS would be visiting stations and squadrons and that all assistance and access was to be provided. Air Staff Memorandum No. 32, ‘Operational Research Section,’ provided the Headquarters staff with a brief introduction to the scientists. The functions indicated were identical to those sent to the groups. Additionally it was stated that the Section will take over from the C.Arm.O. [Chief Armament Officer] the analysis of interceptions et cetera, ... and be available to all branches for consultation on scientific and technical matters. All branches are to collaborate closely with the Operational Research Section and should supply such information as may be required. In addition branches should forward papers likely to be of interest to the section as occasions arise.56

Initially the staff was organized by subsections to study four questions: bomber losses; successes in bombing ops; vulnerability; and radar and radio problems. The number of scientific officers grew steadily, along with the growth of the Command, until the Section reached its peak at fiftyfive scientists in August 1943. Dickins took care to point out that ‘very few of the officers posted to the brnch [sic] proved unsatisfactory or disliked the work, a state of affairs which was largely due to the care taken by the Ministry of Aircraft Production in selecting the staff.’57 By 1943 ten ‘laboratory assistants’ had been brought on strength and another ten would be added in 1944. Finally, approximately a dozen clerks supported the OR Section.58 The senior staff of the ORS are listed in appendix 7. A detailed review of the build-up of the Section shows that there was some degree of trial and error in defining work and assigning resources. Between September 1941 and early 1942 the Section grew by another subsection, set up to deal with day operations, but it soon became apparent that radar and radios were too closely associated with operations for them to be considered separately. Thus the organization was rearranged into three components: ORS 1 – Success of Night Operations; ORS 2 – Losses in Night Operations; and ORS 3 – Day Operations. This last

46

The Science of Bombing

group was disbanded in early 1943 and when day operations resumed in 1944 ORS 1 conducted the associated analyses. A different sort of ORS 3 was activated in 1943, this time to conduct the statistical work in support of ORS 2. Finally an ORS 4 was formed in mid-1942 with a mandate to look into unique issues which did not interact with the main work; as well, ORS 4 was subsequently the home for the production of the Bomber Command Quarterly Review and Bomber Command Raid Reports, the latter a series of extensive analyses of each night’s ‘battle.’ By mid-1942 it had become apparent that the blending of the radar section into general operations had been premature and this section was once again broken out on its own as ORS 5. Meanwhile the issue of manpower had taken on considerable significance and an ORS 6 was established to conduct ‘administrative research.’ Field detachments were added at the group level, and, in 1944, a Bombing Research Unit, initially called the Bombing Analysis Unit, was established to examine the effects of bombing on the continent.59 Bomber Command was not alone in feeling its way through what and how the scientists should be contributing; the section was part of a bigger transformation across the RAF. An informal meeting of ORS Heads took place on 10 October 1941 with Sir Henry Tizard leading the discussion. The session was held at the Ministry of Aircraft Production despite the fact that the command representatives, while officially carried on the rolls as MAP staff, were part of RAF organizations. Such was the confusing organizational chain that existed at that moment. The aim of the meeting was to launch the expanded OR effort effectively, an effort that was seen as having considerable importance for the Air Ministry. Also, in these circumstances staffs were at pains to ensure that this or other sessions were not misinterpreted as a means to usurp the control that the Commands now had of their scientists.60 In opening the gathering Tizard stated the purpose of the OR sections in terms of two broad tasks: ‘to give advice on the more effective use of equipment and personnel; and to give a considered opinion on the demand and priority for new equipment.’ Discussion turned to the question of production and distribution of individual OR reports. From their experiences Blackett and Larnder felt that the best results were achieved when the flyers authored reports. It meant that the scientists were not bogged down with administration and that someone who understood the issues was able to write up the material in a fashion that could be understood by practitioners. Distribution was best decided by the SASO, with reports falling into two categories: those which should have broad exposure across the air force and those of ‘housekeeping’ variety that

Boffins: September 1941

47

dealt with internal matters. When Dickins’s turn came, he was asked to describe the major issues facing Bomber Command. He listed them in priority: first, finding ways of reducing increasing losses during night operations; and then determining what types of targets to attack under various weather conditions. The members then turned to the question of sources of ORS staff. Blackett felt that there was a need for high-calibre thinkers ‘of Fellowship standard.’ These people did not have to be scientists, but they did need first class minds and it was considered that statisticians would be of great use. Dickins further pointed out that he had a requirement for researchers who understood both the vagaries of radio aids and the needs of his Command.61 He was already making the critical link between science, whether pure or applied, and the needs of the operators. It was no doubt based on this meeting that the next day Saundby warned Peirse of a possible move by the Air Ministry to take control of the work of the OR sections, coordinating the efforts of the scientists at Fighter, Coastal, and Bomber Commands.62 In truth, as the minutes of the first two formal meetings of the Committee were to show, there was probably some value in ensuring that there was a degree of integration between the growing OR sections, and in the coming weeks a tiny coordination office was set up at the Air Ministry. Among other things, the Operational Research Centre or ORC would be responsible for running the Operational Research Committee, it too referred to as the ORC, which would review the work of the sections ensuring that important developments were shared across the RAF. Additionally, the Committee soon formed a subcommittee to ensure the best distribution of scientific staff between the sections – scientists were in much demand and central oversight was not disputed.63 For now, in responding to Saundby, Peirse thought that any such proposal for take-over could be dealt with.64 At the end of October, the first meeting of the Operational Research Committee of the Air Ministry was held. Many of the same issues discussed on 10 October among the scientists were reviewed, and roles and responsibilities among the various organizations were worked out. The meeting was chaired by the Deputy CAS and attended by several senior military officers from the Ministry as well as the senior leadership from the commands. Saundby was the Bomber Command representative and, along with Dickins, made a number of valuable comments.65 In the end an effective professional scientific network was established, but the Command had retained its boffins. It would need them.

4 Sorting Out Process and Producing Results: September 1941–February 1942

With its mandate established and the work force growing, there was now sufficient focus and energy to allow the Section to develop the necessary research programs with which to respond to Bomber Command’s problems. The reliability of the Section’s conclusions and recommendations had everything to do with the validity of its collection and interpretation of data. Dickins underlined this relationship, saying that ‘one of the primary aims of the O.R.S. was to study the success of operations, in terms of the percentage of aircraft despatched which bombed the target, and make recommendations whereby this percentage could be improved.’1 He emphasized that success was governed by three factors: tactics, equipment, and training. The last two, he argued, were hard to alter quickly and therefore tactics became the focus of analysis. To do effective analysis the scientists would need data from both successful and unsuccessful (failed) operations, and obtaining this information was a significant task rendered more difficult by the fact that Bomber Command was involved in a major raid – the equivalent of a land or sea battle – several times a month. This meant that data from literally thousands of aircraft had to be collected, collated, and interpreted. This, Dickins wrote, was no simple task. If the actual raid as a whole can be reconstructed and studied, it will be found to differ more or less from the raid as planned; such differences either may be of the order expected (from past experience) by the planners, or they may be critically large so as to be significantly discrepant from the planners’ intentions. The problem is to reconstruct the raid, compare it with the plan, and (if possible) to account for major discrepancies between them.

Sorting Out Process: September 1941–February 1942 49 In reconstructing the raid, the chief difficulty is that essential data are both scanty and (when available) variously unreliable. Thus, the analyst is forced to accept a sample drawn from the raid (considered now as a population of data relating to aircraft tracks) and this sample may be inadequate in size and or biassed in nature.2

The availability of data had always been the Achilles heel of any effort to come to terms with the ineffectiveness of bombing. Attached temporarily to the headquarters earlier in 1941 from Watson Watt’s telecommunications staff at the Ministry, Mr A.O. Rankine had left a short summary of his impressions on departing from High Wycombe. His key recommendation underscored the need for a more scientific collection of data. He had been forced to use information gathered through the intelligence system and found that it was incomplete and often gathered and sorted in such fashion as to make its scientific value questionable. He was also perplexed by the ways in which aircrew might or might not use equipment available to them, citing the unproven but popular belief that switching off the aircraft’s IFF (Identification Friend or Foe) system would lead to an immediate dousing of enemy searchlights. At the same time he observed that crews were not using blind landing aids, devices which might actually be of benefit when visibility at the landing field was reduced. Overall he suggested that it would appear that the men called upon to use new radio methods may not have been of the right type, or at least that their training has been insufficient to create in them the necessary degree of confidence in the apparatus. I believe the same thing applies to ASV in Coastal aircraft, and other airborne adaptations of RDF. There is reason to fear that apparatus is sometimes, if not frequently, reported unserviceable when in fact the operator does not know how to use it. Equally it is possible in present circumstances for a device to be indeed unserviceable but not to be so recognised, with the result that it misleads. Both these deficiencies could be most surely removed by employing operators highly qualified in the technical sense. Alternatively, but not so certainly, a remedy could be found by keeping actual operations under close scientific scrutiny.3

As the ORS took on the analysis function through the fall of 1941 a process was adopted whereby a Sortie Raid Report was prepared by each aircraft that had taken part in a particular raid and that data was then

50

The Science of Bombing

passed from the Squadron to the Group, and finally to Bomber Command and the ORS. Originally proposed on 9 August 1941, the report and associated forms evolved over time. A trial was conducted by 5 Group in November and December and their form was adopted for use by all Groups as of 1 April 1942. For purposes of analysis the key data were the target attacked; bombs dropped on target; time, height, heading, and airspeed at bomb release; what was in the bomb site (or other method used for bombing); and bomb aimer’s and pilot’s reports. Based on this information, the ORS was capable of developing a time histogram of the raid, ‘known from its size as the “Table-cloth.”’ This sheet depicted the raid as it had occurred minute by minute with each half-inch square representing one sortie. Each of these ‘cells’ contained detailed information about the aircraft, its target, the aiming point, whether the bomb release photo had been plotted, and even the bomb load.4 Beginning in 1942, similar data was added to the tablecloth that represented Pathfinder operations, if part of the raid. Additionally a series of points were eventually added, which showed the type and duration of any target indicators (special bombs that would create a bright pyrotechnic fire on the ground to be used as a reference or aiming point) which the pathfinders had employed during the raid.5 (A condensed version of a histogram for the raid on Kassel 22/23 October 1943 is shown in appendix 8.) Dickins’s reference to the ‘aircraft’s photo’ should be explained at this point. Initially Bomber Command had supplemented reports provided by the crews with air photographs of the targets taken by reconnaissance aircraft as soon after the raid as possible. These photographs rarely showed any evidence of the bomb strikes reported by the crews and so cameras began to be installed in the bombers in order to confirm what was ‘in the bomb-sight’ at the time of bomb release. The use of night photography was still in its early days when the OR section was created. As Dickins pointed out, ‘with the introduction of night photography this [bomb-sight] information became available and it became possible to determine the causes of failures and to improve the effectiveness of operations.’6 Butt’s data had demonstrated the value of night photos, and from Sir Richard Peirse and others there then came a requirement for wholesale installation of cameras for the purpose of recording the dropping of weapons, rather than for assessing damage, a function which would have demanded a different and much more complex technology. Some objected, feeling that this was an invasive action challenging the veracity of crew reports. The issue was settled on 19 September when the Head-

Sorting Out Process: September 1941–February 1942 51

quarters promulgated a statement on the matter. ‘For your information, the purpose of night photography in this Command can be defined as follows: (i) To confirm the location of the bomber aircraft at the time of the attack. (ii) To pin point the bomb bursts. (iii) To provide general information.’7 Once the cameras were installed and the analysts made available to study the film, the Commanders would have the necessary information to know where their crews were and what they were doing.8 But effective interpretation of night photography did not come quickly. There were several technical problems to be sorted out, as well as the formidable task of collating the mass of graphic data that was collected after each raid. The challenge in gathering accurate data lay in ensuring that the camera photographed that point on the ground where the bombs were about to impact. However, there was a range of problems that could affect this accuracy and which had to be overcome. After experimenting with rearward tilting cameras, such as used in day operations, the scientists finally decided that a vertically mounted camera would offer the best results. A photoflash device would be dropped automatically, activated by the bomb release key, and, by using a variable time-setting based on bombing height, the photoflash would then illuminate that point on the ground where the centre of the bomb stick was due to strike. Because of time errors with the photoflash, it was necessary for the ‘bombing frame’ on the film to be exposed for eight seconds (during which time the bomber was required to continue in straight and level flight) in order to be reasonably sure that the flash would illuminate the appropriate area. This resulted in a fairly long piece of film to be interpreted, but this actually allowed the interpretation unit to confirm a number of details based on the streaks of light within the frame, these created by fixed light sources on the ground.9 It was also possible, particularly in the course of a large incendiary raid, for the actual target area frame to be ‘fogged’ out by fires already burning on the ground.10 Happily, in early 1942 N Section of the Central Interpretation Unit developed a method of plotting aircraft position relative to the target area by means of this very fogging.11 The rate of accurate plottings continued to grow as the war went on. In August 1942 a ‘significant proportion’ of plottings were possible, while in the spring of 1943 50 per cent of all clear weather bombing was plotted. At the end of the war fully 80 per cent of all night raid sorties could be plotted. The scientists were confident that ‘as a result of a careful investigation, it can be said that from the beginning of 1943 onwards operational photographs yielded an unbiased sample of the size adequate for quantitative

52

The Science of Bombing

analysis on most clear weather raids and on the fair proportion of moderate weather raids.’12 Even for poor weather raids it had proven feasible to develop and implement a method of plotting using cloud patterns, which made it possible to assess the concentration of the raid in space. Better yet, if at some point there was a break in the clouds exposing the ground, then the interpreter could compare this data with the actual aiming point and target area. The analysis of greatest value came to the ORS from the RE 8 section of the Ministry of Home Warfare,13 which portrayed the data in terms of crater plots, these determined from post-raid daylight photo reconnaissance. The scientists felt that the crater plot was a more valuable and accurate tool as it showed the actual location of bomb impacts. ‘Also, ideally, the crater plot shows all the live bombs dropped, whereas at best the plot of photographs recorded the position of only a sample of the attacking aircraft.’14 There were nonetheless a number of problems in the assembly of data from an analysis of crater plots. First, if the reconnaissance aircraft had photographed the actual target area, but the raid had been scattered or badly off target, then the real results of bombing in terms of concentration would not be known. Second, even with a well-concentrated and accurate raid, it was not possible to ascertain the actual number of bombs which had landed in the target area, as it was estimated that only 80 per cent of the crater plots could be seen or discerned. If, for example, bombing had taken place in wooded country or heavily built-up areas it was particularly difficult to identify each individual crater. Significantly, ‘the most important point was that more craters tend to be missed in the central parts of the bomb distribution where the density was greatest and as a result the spread of bombs tended to be over-estimated.’15 In other words, if several bombs went into the same crater it would not be apparent that this had happened and analysis would then suggest that the actual density near or on the target had been less than was actually the case. One other problem was that crater plotting did not demonstrate how the raid had actually evolved over time. It simply showed the results, and so if there were tactical issues that might be the focus of further research these would not be aided by a review of bomb craters.16 It was possible to gather evidence from two other sources of night photo data: fire mosaics (the actual photographic mosaic of fires on the ground) and marker bomb (later called ‘target indicator’) plots. In the first case, N Section was able in some instances to follow the development of the incendiary pattern over time. The particular value of these

Sorting Out Process: September 1941–February 1942 53

fire mosaics was as a qualitative tool for quickly determining where the actual, rather than the planned, aiming point was rather than waiting for the plottings of light tracks or the quantitative analyses which could and did take weeks. Of even greater value in the scientists’ minds were the plots of target indicators provided by N Section beginning in early 1943. ‘It would be difficult to overemphasize the importance of these target indicator plots in raid analysis. Not only did they provide crucial evidence on how well the Pathfinders had centred the attack, but they also were invaluable in assessing the accuracy of various radar devices used in aiming the target indicators.’17 Thus, taken together the scientists had a fairly extensive range of data from which to analyse the effectiveness of bombing tactics and other issues.18 To do this they applied two somewhat divergent processes. Qualitative analysis permitted the rapid estimation of tactical successes or failures and the results were published as soon as possible in the B series of ORS reports which were distributed within the Command headquarters, as well as to the Groups that had been involved in the particular operation. Describing qualitative analysis, Dickins provided an intriguing discussion of the mental processes: This is a type of analysis about which is very difficult to give a clear description. The methods are not mathematical but more akin to those of the law courts, or of the detective of fiction. A great number of strands of evidence are mentally assessed for reliability (mainly from their coherence and mutual consistency) and then woven together to form what is, in effect, a reconstruction of the raid. With rare exceptions, one is not faced by deliberately misleading statements in the Sortie Raid Reports, but cases are frequent in which the reporter is genuinely mistaken about what he saw or where he bombed. In the case of a raid which has fallen below expectation (and these are the raids mainly selected for study), the time when and reasons why it failed will usually emerge as analysis proceeds. For example, it might be that a small group of aircraft claimed to have bombed a cluster of red target indicators at a time when it is known (from Pathfinder Force raid reports) that none were burning; one of these aircraft may have been plotted away from the main concentration and their descriptions of the target area may make it probable that the others were part of the same diversion.19

Because these analyses often dealt with success, they were generally well received. In those cases where the conclusions were critical of a particu-

54

The Science of Bombing

lar operation, the draft report was first sent to Pathfinder Force for review prior to publication. Quantitative results, on the other hand, appeared to be ‘somewhat academic and unpractical’ in nature. The product of quantitative analysis was a set of numerical conclusions, descriptive in nature, and not perhaps suited for publication. But quantitative conclusions could still allow the synthesis of more complex thinking: The fact is, however, that raid parameters, which are the end product of the process of quantitative analysis, become in turn the raw data for further researches. By collecting the parameters for whole groups of comparable raids, it became possible to characterize the group as a whole in a reasonably exact way, and thence to compare one group with another. Thus, the way was open to assessing and comparing different bombing techniques, to furnish precise information on the accuracy of Bomber Command attacks under given conditions, and to estimating the weight of attack required in future operations.20

Bombing accuracy could be reported in three ways. The first of these was to indicate the percentage of bombs falling within three miles of the aiming point, and this method was used between August 1942 and January 1944. Introducing the second method, Point Density, Dickins noted that ‘the obvious disadvantage of expressing raid success in terms of proportion of aircraft bombing within a selected distance of the aiming point is that parameters relative to one radius are not comparable to those relative to another, and that one cannot be calculated from another owing to the varying systematic errors.’21 Finally, beginning in November 1943, it became possible to describe bombing accuracy in terms of a ‘percentage within the target area.’ ‘This is a parameter of a totally different kind and, since it is relative to the size and shape of the target, it measures potential effectiveness, rather than accuracy.’ Since the target area was not generally a theoretical ‘bullseye’ of fixed radius, but rather a geographic section of a town or city, this new process was helpful in permitting the decision makers to know what percentage of munitions were hitting the target itself.22 Few of these sampling and analysis protocols were available to the scientists in the fall of 1941, but were only gradually developed, in some cases, over the period of more than a year. It was not surprising, therefore, that one of the first conclusions reached by the Section focused on

Sorting Out Process: September 1941–February 1942 55

scientific process. Dickins felt, and experience showed, that the analysts did not, at the outset, have adequate data with which to make conclusions or recommendations about operational matters. The final step in getting the analysis into the hands of the decision makers came in the form of a written report, which would normally include both conclusions and recommendations. The draft would first go to the branch or branches concerned and then on to the SASO or even the CinC for approval and further action if deemed necessary.23 These reports were organized into four different series. The General series (these did not have a ‘G’ prefix and were usually titled either a Report or Memorandum), numbered from 1 to 141, and were given wide circulation as they dealt with matters of general interest. The ‘S’ series, numbered 1 to 243, were distributed within Bomber Command, and also sent to those external agencies immediately concerned with the subject. The ‘B’ series, numbered 101 to 237, were ORS reports which had not necessarily been approved by the Command; ‘Interim Raid Reports’ were included in this series. Finally, the ‘M’ or Memorandum series (always containing the prefix M and not to be confused with G series memoranda), numbered 1 to 160, reflected various scientists’ views and conclusions which had not necessarily been approved by the OR section itself. Scientific process aside, there was little doubt that, as the members of the ORS began to assemble at High Wycombe in the late summer of 1941, their first and most pressing task was to come to grips with the challenge of measuring the accuracy, and by extension efficiency, of bombing operations. Even as they set to work to develop the necessary techniques for this task they were called upon immediately to help solve the problems that had been brought to the fore in previous weeks. Not only would they need to look at navigation, but they were also asked to address the very real problems of training and visual target identification, all for the purpose of getting more bombs onto the targets. Only days after the activation of the Section, Peirse asked Dickins to sort out the discrepancy between crews’ reports of success in the attack of a synthetic rubber plant at Huls on 6/7 September and the total absence of damage in subsequent day reconnaissance photos. ORS Report S 3 ‘Investigation of raid on Huls – night of 6/7th September 1941,’ dated 10 September 1941,24 reviewed the bomb release photos from seven of forty-five attackers and showed that the closest that any of these particular aircraft had come to the aiming point was six miles. While able to provide this sobering interpretation of photographic data, the scientists were not encouraged by the ‘scanty

56

The Science of Bombing

nature of the evidence’ and felt more investigation into attack results and weather conditions was warranted.25 Better days followed. Later in the month the Section was able to look at a series of three raids against the Baltic port of Stettin for which the scientists were able to build their conclusions using a variety of inputs: night photographs, information supplied by night photographic plotting forms, additional special interrogation reports, and navigators’ log books. One of the three raids, that of the night of 30 September, had been particularly successful, with 80 per cent of the photographs indicating bombs falling within two miles of the aiming point. It was concluded that the difference between this and the other two much less successful raids had to do with the condition of the moon, and it was apparent that moonliight had been the chief reason for that raid’s good results. In terms of scientific process itself, these investigations continued to demonstrate ‘the inherent unreliability of crews’ reports and the danger of basing any conclusions as to the success of the attack on these alone.’26 The essential theme of these analyses emphasized ‘that the difficulties involved in navigating to the target area, as great as they were, were relatively small compared to difficulty of identifying the target itself, especially on dark nights.’ Moreover, the analysis allowed the scientists to clearly delineate the various problems which the crews faced.27 The boffins were quick to see the importance of target recognition and minuted the CinC on 20 November to report their progress. While recognizing that bad weather had been a problem during the period of their observations, the data showed that, of 209 reports submitted, just 79 claimed to have identified the target. The scientists had only 39 corresponding successful photos; of these, only nine were within five miles of the target, with another six in the five- to ten-mile ring. Mitigating the dismal results was the fact that the weather had been bad. Seeking additional information, Dickins asked for Peirse’s support in getting station intelligence officers to ensure that forms were filled out correctly; he also asked that crews be required to explain what might have prevented them from identifying the target. Saundby and Peirse endorsed the plan for further research.28 As part of that research Dickins convened an informal gathering of several scientists and some of the navigation and ops staffs of the Headquarters during the first week of December. Sir Henry Tizard and Mr Capon of DSR were invited to attend. The discussion ranged widely in an effort to zero in on possible means for identifying the target. The use of flares, homing on an airborne target-finding aircraft, and enhanc-

Sorting Out Process: September 1941–February 1942 57

ing vision with night scopes and goggles were all discussed. The most bizarre proposal came from the Royal Aircraft Establishment, suggesting in-flight processing of photos which could then be used to confirm the aircraft’s position relative to the target. No conclusions were drawn, but the discussion would certainly have allowed the OR Section to share their thoughts with a wide range of interested and expert thinkers.29 Indeed, Dickins recorded, among the recommendations made in [the conference’s] report (ORS Memo M88) was one to the effect that consideration should be given to the ‘formation of special squadrons to initiate raids and to raise fires,’ which appears to be the first reference to the possibility of a special Pathfinder Force, such as was eventually instituted in August of the following year. Another recommendation stressed the need for the ‘use of reconnaissance flares by a number of aircraft in cooperation,’ a technique which formed the basis of nearly all the target marking methods subsequently developed.30

Meanwhile the report had now been seen by the Groups, and 5 Group in particular thought to remark upon the validity of the conclusions: it is observed that the general conclusion drawn in the penultimate sentence of paragraph 6 compares very closely with the experience of all good captains and navigators. The essence of good map reading is check and double check where possible. It is suggested that this conclusion should be reported to Flying Training Command and all Schools where map reading is taught, for it has a significance which cannot be too strongly impressed upon all inexperienced crews.31

In February 1942 the scientists were ready with a follow-up report that they first circulated to the navigation staff. In the scientists’ opinion: ‘Consideration of the conclusions reached in the report suggest (a) More training in map reading and practical experience of its use at operational heights during training is required, and (b) That a careful study of the target area by crews is essential.’32 Dickins asked for a review of the paper and of the proposed covering letter, which, in addition to restating these points, requested the full cooperation of crews and of intelligence officers who gathered the data on behalf of the boffins.33 Wing Commander Nav responded within a day saying that the report ‘is a most interesting document which deserves wide circulation.’ He recommended that a shorter more straightforward version be sent to

58

The Science of Bombing

the Groups and stations and that it should be accompanied with examples to illustrate what had been happening. ‘If you agree I would also like to publish an abbreviated report in the monthly Navigation Bulletin which as you know is sent to all training schools both in the U.K. and the Dominions, as I consider that this brings the training aspect to the fore.’ ‘I would welcome something from the O.R.S. each month, if this is possible, especially examples illustrating definite causes of weakness.’34 The scientists developed the summary over the next few days and by the end of February, Saundby had approved the package for distribution.35 Here was a clear instance of the ORS getting at the essence of what was troubling the operators. They had gathered and analysed data and presented recommendations to the decision makers for improving the effectiveness and efficiency of the Command. The feedback was clear and positive; it was obvious that the practitioners appreciated this second look at what was and was not happening and were ready to implement logical and supportable recommendations. December had also seen the conclusion of a larger trial which attempted to gain a better understanding of what crews actually thought they were seeing at night. The night photographic plotting form was modified; additional questions were asked about what was supposedly seen in the area being photographed. Crews were now asked to indicate what place they believed they had photographed; whether or not they had recognized what they had photographed; and by what means they thought that they had been able to make this recognition. When the results came in, the Section was able to conclude that there was a ‘big difference in the reliability of various ground features.’ Lakes, rivers, and other inland bodies of water, while very popular with the aviators, had proved to be very unreliable, while coastal features including docks were relatively ‘trustworthy.’36 Dickins minuted the CinC at the end of the first week of December reminding Peirse that ‘before this branch was established a report analysing night photographs was prepared by a member of Lord Cherwell’s office.’37 The OR staff had now conducted a similar study38 and were able to state that overall there had been no significant change in accuracy in the August to October period. Nevertheless, the data pointed out the existence of a link between success and such factors as target location, defences, and weather. Dickins expressed hope that some degree of precision could be achieved in meteorological forecasting given that raids conducted in good weather had a 50 per cent success rate compared to 20 per cent for those taking place in cloudy conditions. He also

Sorting Out Process: September 1941–February 1942 59

drew attention to the results of raids on moonless nights, which were almost complete failures, compared to the 50 per cent success achieved in full moonlight. In forwarding the report to Peirse, Saundby quickly saw the value of the analysis. The relevant paragraphs ‘clearly show the effect of moonlight, visibility, and the enemy defences on the numbers finding the target.’39 There remained, as this data began to be filtered and collated, two issues that continued to form the nexus of the challenge of getting bombs onto the target: navigating to the target area, and finding the aiming point. Scientists and flyers realized that both could be addressed to some extent by the operational use of T.R. 1335, or Gee as it had been dubbed, a system which provided crews with a series of radio lattices along which to navigate with considerable precision. Gee was a navigation device wherein a series of three ground transmitter signals would be received by an aircraft and displayed on a cathode ray tube as two position lines called Gee coordinates. Using these and a special map which had all Gee lines overprinted on it, experienced navigators could pinpoint an aircraft’s position in less than one minute. While limited in range by the earth’s curvature, the system had a range of 3–400 miles and an accuracy of 0.5 to 5 miles. Gee thus had considerable potential and during the fall of 1941 the ORS was well along in its thinking about how to optimize the use of the system. In two related reports the scientists recommended that research be conducted in order to determine the best tactical employment of the device.40 This optimization was deemed critical since it was commonly assumed that the RAF might get six months’ use out of the system before the Germans would figure out how to jam it. Dickins minuted the SASO on 2 December saying, ‘I do not know to what extent the special operational use to which Gee might be put has been considered, but it has occurred to us that a Memorandum on the subject might be useful.’41 Dickins went on to suggest that a number of experiments should be carried out and that a small committee be formed to oversee the work and consider the results. Discussions between the Group Captain Operations, the Chief Signals Officer, and Wing Commander Navigation, who at this juncture were the senior staff members in their respective departments, were suggested as a starting point for this work. The response from A/Cdre T.M. Williams, the Deputy SASO, was positive and immediate. ‘This is the first time that I have seen a detailed paper on the possible operational employment of “Gee.” It is very useful and I am sure will form the basis for drawing up a detailed programme of experiments.’42

60

The Science of Bombing

The actual document, ORS Memorandum 23 ‘Operational Use of GEE,’ dated 1 December 1941, reviewed the capabilities of Gee and then went on to suggest some ten possible uses of the device including the primary functions of navigation and attack during various weather conditions. Mr J.A. Jukes of Dickins’s staff proposed trials in which the system would be tested at ranges similar to those over Germany. As well, he called for work to determine the best method for bomb release with respect to lattice lines. There was, he said, a critical need to develop crew procedures for bomb release, suggesting that the navigator operate the Gee set while the second pilot (still part of the crew in most aircraft types at this time) occupied the bomb aimer’s position and actually released the bombs. He also identified the need to collect data and to reassess any preliminary decisions once actual operations had been conducted. In sum, he said that while it was clear that Gee had the potential to render operations more effective, while also cutting losses and permitting blind bombing, there was much that had to be learned.43 In sending the proposal to Saundby, Dickins added some additional thoughts: ‘It is clear, however, that these new tactics will have to be carefully developed and it will be necessary also to undertake tactical and operational experiments to determine in detail the best methods of using the equipment.’44 Later in the month Jukes provided additional material for review and decision. In this instance it had to do with whether ‘lattice charts,’ the grid of Gee signals printed on special maps to which the navigators would refer in confirming their position, should be used in the air. There was concern lest these charts fall into enemy hands as the result of crashes, but Jukes thought that not providing them to the crews would be equally detrimental. On 23 December, he produced a short study: ‘Some rough notes on results of 2 flights to investigate the possible use of T.R. 1335 without lattice charts.’ The experiments had revealed that the technique was not viable and Jukes predicted that there would be much confusion and delay in and around the target area if navigators were not provided with the lattice charts. Worse, it was not possible to determine position over the target area to less than two miles’ precision. He concluded that ‘without the use of lattice charts it is clear that the great advantages which will arise due to the possibility of accurate timing and routeing, and the consequent concentration over the target will be lost.’45 A report issued by W/C Navigation on the same date, ‘T.R. 1335 Development Unit – Interim report for Period 16.12.’ 41 to 20.12.’ 41,’ indicated many of the same problems.46 It seems clear from the like thinking of these two documents that Jukes had quickly developed an understanding of

Sorting Out Process: September 1941–February 1942 61

the operational problems confronting navigators, while it is equally reasonable to assume that some of Jukes’s thoughts were incorporated into the Navigation report. In the first days of the new year Dickins provided a detailed minute to Saundby which laid out the scientists’ concept for the upcoming Gee trials. The matter had been discussed with the Navigation and Signals staffs before the preparation of a draft plan. To this he now added the need for a Telecommunications Research Establishment (TRE) expert and asked Saundby to approve this temporary attachment. Anticipating the requirement to analyse the results of future raids using Gee, he also asked for approval to start collaborative work with the Air Staff to ensure that the necessary protocols be in place from the outset. Finally, he sought the SASO’s support for an increase in OR staff based on the volume of data that would have to be assessed once Gee operations began.47 Saundby was slow to reply, but gave his full support by the end of the month.48 It can be seen that the head of the ORS had now become au fait with how best to get things done. Dickins had defined his needs, determined how best they could be satisfied within the various Headquarters structures, and sought approval to make the necessary formal requests and changes. It was clear that he had learned the system and was working it with relative efficacy. As a result of these reports and proposals, the AOCinC called a meeting on 17 January 1942 with the ORS staff and Air Ministry officers to discuss how best to employ the new device. Two scenarios were considered: Gee aircraft operating alone; that is, conducting blind bombing in poor weather, and Gee aircraft leading non-Gee bombers. In this latter tactic Gee aircraft could assist non-Gee crews in one of three ways: by acting as ‘fire raisers’ through the use of incendiaries; by dropping flares to permit visual identification by following aircraft; or by leading formation attacks. Night formation flying was not considered to be practical so this last option was dropped. The first two, and particularly the flare and ‘follower’ idea, had potential. The ORS was ‘entrusted with the task of drawing up detailed plans for such experiments.’49 The subsequent plan, ORS S 30 ‘The Operational Use of Gee. III The Use of Flares in Conjunction with GEE,’ proposed the modalities that might prove effective in the use of flare-equipped Gee aircraft to guide nonequipped followers to the target. The document also proposed experiments that could be conducted over the Isle of Man.50 Five days later, the proposal had been revamped into a directive, signed by Saundby, tasking 3 Group with conducting the trial ‘Experimental Use of Flares

62

The Science of Bombing

by Tr. 1335 Aircraft in order to assist Aircraft without Tr. 1335 to locate the Target.’51 It was at this time that A/V/M Slessor, then AOC 5 Group, wrote to the AOCinC about the scientists’ contributions. He commented on the problems of finding the target and how the introduction of Gee would be of assistance. He could see the logic of assigning the new device to a few squadrons and using them to help the others find the target. He closed his letter with a comment on ORS work. ‘These analyses by the Operational Research Section are most interesting and valuable, and show clearly that we have been unduly complacent in the past as to the results we have been getting.’ He added a caution. ‘I am sure you are alive to the danger of taking them too much at their face value. They are liable to get into the hands of people who do not realize that war – and especially air war – is not merely a matter of material effect measurable in terms of Gaussian law.’52 As he would when at Coastal Command, Slessor had captured the essence of the relationship between scientists and commanders. The analysis was invaluable but had to be appraised in relation to the greater corpus of professional experience, not to say the responsibility and accountability, that came with high command. Experimental attacks, codenamed ‘Crackers I and II,’ were conducted on 13 and 20 February over the Isle of Man and Wales respectively. While there were some technical difficulties during the first of the two exercises the proposed ‘Shaker’ technique was found to be effective.53 In their first post-trial report, issued on 15 February, 3 Group described relative success. In addition to these observations they confirmed the viability of the attack profile that would use the Gee aircraft to first illuminate and then mark the target using flares and incendiary bombs respectively.54 The non-Gee main force would thus be able to bomb the ‘conflagration’ which marked the target. The ‘Shaker’ attack protocol would thus consist of three waves: Gee-equipped flare-dropping ‘Illuminators,’ incendiary-dropping ‘Target markers,’ and finally the ‘Followers’ bombing with high explosives.55 (Bombing techniques are illustrated in appendix 8.) The proposal, which had been developed by 3 Group and BC staff including Dr Dickins and other members of the ORS, concluded by stating that, should the proposed method be found acceptable, ‘it is assumed that ... your Headquarters in conjunction with the O.R.S. will issue a suitable plan.’56 While it is apparent from these remarks that AOC 3 Group had not yet grasped the fact that the ORS was part of Bomber Command Headquarters, at the same time it is clear that he understood the value of having the scientists review and confirm the various schemes

Sorting Out Process: September 1941–February 1942 63

put forward by the operators. Despite these positive recommendations, as a result of the problems that had occurred on the first attempt Saundby and A/V/M J.E.A. Baldwin, now the acting Commander,57 decided the following day that the trial should be repeated. The results of the first Gee raids will be discussed in the following chapter. Linked to the Shaker evolution, the OR section had also examined the question of a permanent target-finding force which could identify the aiming point for the main body of an attacking force. ‘“The formation of specialist squadrons to initiate raids was recommended by the O.R.S. as early as December 1941.”’58 It was not until the spring of the following year, however, that the topic began to generate long and heated discussions. These will be considered in more detail in the following chapter, but it is significant that the ORS had seen this possibility within only a few months of its activation. Indeed, in existence for just six months by the end of February 1942, the ORS had nevertheless accomplished much. The scientists had begun the difficult and at the same time critical task of assessing bombing tactics and techniques in an attempt to find the best solutions to problems which had plagued Service leaders for over two years. They had identified the sources of data that they needed and were familiar with the analysis techniques that they had brought from standard scientific practice. Their ranks were carefully filled to ensure they had the best minds and personalities to work with the flyers and they knew well enough to establish contacts both in the headquarters and across the command. Moreover, the boffins had earned the confidence, or at least respect, of those for whom they worked. Indeed, the AOCinC had, within days of the Section’s activation, sought Dickins’s recommendations, and later the ORS, having proposed procedures that might best utilize Gee’s potential, was tasked to take on important and groundbreaking tactical experimentation for seeing this aid into service. There was little doubt that a young Basil Dickins and his staff were providing invaluable assistance for the two most powerful men in the organization. Though both Peirse and Saundby, as well as Baldwin, seemed approachable and happy with the scientists’ contributions, this dynamic was likely to change with the arrival of the formidable Sir Arthur ‘Bomber’ Harris.

5 Arthur Harris and a New Beginning: February–December 1942

When Sir Arthur Harris took over the Command on 25 February 1942 the headquarters staff had had about six months to get used to the notion of scientists taking on those technical and tactical problems that were affecting the efficiency of the bombing campaign. But Harris was well known as a hard-nosed and rigid commander. How would he adapt to these scientists within the heart of his organization? Arthur Travers Harris had come to the RAF in a fashion somewhat typical for men of his time. He had joined the army in 1914 and, after spending many months on foot chasing German forces through central Africa, had applied for the Royal Flying Corps. After winning his wings in January 1916, Harris was asked if he could fly in the dark. His flippant answer, ‘“I replied that I couldn’t fly in daylight, so maybe I could in the dark!,”’ earned him a posting to 39 Squadron, which was part of the antiZeppelin defence of London. Here he taught himself to fly and land at night. After a frightening first flight (two nights after the station CO had killed himself attempting a night takeoff), Harris ‘“used to go up in the evenings and fly around and around, landing time and time again as it grew progressively darker, until I’d got this thing at night weighed up.”’ He used this self-developed training plan soon afterwards when put in charge of a flight at Hornchurch, and this need for intensive training then stuck with him throughout his career. Harris was also inquisitive and decided to figure out why a particular type of aircraft, the DH 5, was prone to spinning. He got into a bad spin and out again and in so doing also displayed a determination to know all that he could about the technical side of the business. Harris ended the war as a Major commanding 191 Squadron, a night flying operational training unit (OTU), and finally 44 Squadron, a UK-based night fighter unit.

Arthur Harris: February–December 1942 65

In 1921 Harris took command of 31 Squadron in India, where he worked hard to keep the unit flying in ‘“appalling conditions.”’ He made no friends among the senior army commanders for whom the unit worked and was about ready to quit the air force when he was convinced to move to Mesopotamia, where Sir John Salmond was to command the RAF organization.1 Harris took command of 45 Squadron, which was equipped with heavy twin-engined Vickers Vernon transports, and soon converted these into a dual transport-bomber configuration by adding a bomb aimer’s position and bomb racks. His flight commanders were Robert Saundby and Ralph Cochrane, two officers who would figure prominently in Bomber Command. Harris immediately started training his crews for precision bombing, winning a competition against the light bomber squadrons stationed in the country. He added night bombing to the squadron’s expertise, this only after he had developed a crude but effective target indicator pyrotechnic which was to be dropped by his best crew.2 From these experiments it was clear not only that Harris believed in bombing, but also that he was already defining night bombing tactics and techniques needed to locate the target. Harris returned to England at the end of 1924 and took command of 58 Squadron then being formed as a bomber unit with Vickers Virginias. He worked the squadron hard, emphasizing both day and night bombing, as well as long-range non-stop navigation training. For his efforts he was awarded the Officer level of the OBE. In the early 1930s Harris was again in command, this time at 210 Squadron, operating flying boats. It was there that he met Donald Bennett, who would become his Pathfinder Force commander. From 1934 to 1937 Harris served as Deputy Director of Plans in the Air Ministry. During that time he worked closely with the Air Ministry Bombing Committee, which defined the bombing policies for attacks against German industry. He was also involved in the definition of requirements for the heavy bombers which would ultimately take shape as the Stirling, Halifax, and Manchester, the latter being the precursor to the Lancaster. In 1937, with a promotion to Air Commodore, Harris took command of 4 Group within Bomber Command. While frustrated by a lack of bombing ranges, he was able to institute a focused program of night and long-distance flying. During this period, Harris’s superior, Sir Edgar Ludlow-Hewitt, sounded cautions over the efficacy of his command, stating “that the rapid expansion of the numbers of aircraft had been at the expense of crew training and navigation aids.”3 There is little doubt that Harris shared these worries and that they would colour his policies during his wartime leadership of the Command. Har-

66

The Science of Bombing

ris spent the final months before the war as AOC Palestine and Transjordan, and in September 1939 returned home to take command of 5 Group, Bomber Command. In November 1940 he went back to the Air Ministry as Deputy CAS. During his year at 5 Group Harris had seen at first hand the frequent inability of aircraft to navigate effectively to their targets as well as the repeated devastating losses in day attacks that made the increasingly irresistible case for night operations and area bombing.4 These were Harris’s experiences and accomplishments, but they do not reveal much of the man’s character. On his own departure from Bomber Command in 1940, Ludlow-Hewitt had provided Charles Portal with extensive notes on Harris: I have had considerable experience of Air Vice-Marshal Harris’s work and am impressed with the rare qualities of personality which arise from an exceptionally active and stimulating mentality. He has an exceptionally alert, creative and enterprising mind balanced by long practical experience together with energy and force of character to give his ideas practical shape and realisation. He has rendered great service in respect of improvements in the technical equipment of the aircraft in his command, and also in the creation and organisation of novel methods of dealing with the extremely difficult problems of crew training. His particular talent lies mainly along practical lines, but his ideas are inspired by an unusually well developed imagination, and it is the combination which makes him a very valuable officer in any task where creative ideas and the energy to put them into effect are required. As a commander his exceptional air experience gives him a sympathetic insight into the attitude of mind of the crews, which is well understood and appreciated by them.5

Harris, however, is often referred to as ‘Butcher Harris’ – mythologized as a cold, distant Bomber Baron who would dully throw his crews, attrition style, against the German defences. Typically, critics believe that Harris did not have a quick intellect: ‘A good case can be made that he was slow to grasp the possibilities and limitations of the new generation of radar technology.’6 Further, Harris’s staff have been seen as toadies: ‘There appears to have been a chronic lack of open, critical debate. ... [with] too many weak men and sycophants around the throne.’7 Freeman Dyson of Dickins’s staff, too, had little good to say about Harris. It is clear from his comments that he was unhappy with the area bombing policy and with Harris for prosecuting it. While finding much fault with his own superiors, Dyson believed that Harris was the main

Arthur Harris: February–December 1942 67

problem. He was ‘a commander in chief who accepted no criticism either from above or from below, never admitted his mistakes, and appeared to be indifferent to the slaughter of his own airmen as he was to the slaughter of German civilians.’ Dyson wrote that both the ORS and Harris’s uniformed staff were simply incapable of countering his character. Harris, he concluded, was ‘a typical example of a prescientific military man. He was brutal and unimaginative, but at least he was human and he was willing to take responsibility for the evil that he did.’8 Perhaps more revealing and certainly less negative are the comments of Charles Carrington, the Command’s Army Liaison Officer. Formerly a regimental peer of Saundby and between the wars well connected in academe and publishing, he was not nominally under Harris’s direct control. ‘Bert’ Harris ... was the most dominating personality with whom I became acquainted in the Second World War. My early meetings with him were unlucky, but, as soon as a real occasion arose for me to divert his notice for a moment to my affairs, I found him attentive and businesslike; and, when my turn came for an invitation to meet his family at home, his direct simplicity and sincerity made him a good host. While he never played for popularity, like some commanders I might mention, never wasted words or time on mere civilities, he instantly made his presence felt. As a horse knows by instinct when his rider holds the reins with a firm hand, so Bomber Command knew it had a master. The whole machine tautened up, seemed to move into a higher gear, and this though he rarely visited the squadrons and scorned to give pep talks.9 No one doubted that he was a master of his trade and had been so since the first year of the RAF’s existence when he was a young pioneer of nightfighting. With his power of concentration on the aim, while excluding the irrelevant, he retained a rugged common-sense which was displayed in flat statements about unpalatable facts; he enjoyed shocking the pedants. As I came to know him better, though I must not claim to have penetrated deeper than his outer circle of acquaintance, I realized that he was not unco-operative, not hostile to the interests I represented. When committed to a combined operation with Army or Navy, even if he opposed its inception, he gave his full support; he never shirked; he never compromised with half-measures.10

While Harris could be abrasive, he had ‘an intensively [sic] active and

68

The Science of Bombing

fertile mind eager to leave no stone unturned in the effort to prepare ... for the concentrated operations that lay ahead.’11 The final portrait was that ‘of an operational commander with great driving force, constantly engaging his mind on how to rectify problems and do things better, and possessing deep knowledge of his business.’12 Harris himself saw a different facet of the job of commander. Writing after the war, he was very specific about the responsibilities of command and the strains that commanding put on an individual. I wonder if the frightful mental strain of commanding a large air force in war can ever be realized except by the very few who experienced it. While a naval commander may at the most be required to conduct a major action once or twice in the whole course of the war, and an army commander is engaged in one battle say once in six months or, in exceptional circumstances, as often as once a month, the commander of a bomber force has to commit the whole of it every twenty-four hours; even on those occasions when the weather forces him to cancel a projected operation, he has to lay on the whole plan for committing the force. Every one of those operations is a major battle, and as much depends on the outcome, success is as vital and disaster as grave, as on any other occasion when the whole of a force engages the enemy. In addition, there is a continuous and fearful apprehension about what the weather may do, especially in the climate of NorthWest Europe. ... The whole of the responsibility, the final responsibility, for deciding whether or not to operate falls fair and square on the shoulders of the Commander-in-Chief, and falls on them every twenty-four hours. For all he knows he may lose the whole of a very large proportion of the force by weather alone, to say nothing of enemy action. It is best to leave to the imagination what such a daily strain amounts to when continued over the period of years.13

When Harris arrived at High Wycombe he was no stranger to the Command or its problems. He took over as fortunes were about at their nadir but with some initiatives in place for hopefully better days, or rather nights. Harris described the nature of the challenges which he now faced in the opening pages of his Despatch on War Operations: ‘These were, primarily, lack of suitable aircraft in sufficient number, absence of effective navigational aids, and a serious deficiency of trained crews. Moreover, the handicaps were not only due to a lack of material and trained personnel, as there were technical and tactical problems affecting the employment of the force which could only be surmounted by intensive research, continual experiment, and unshakable resolution.’14

Arthur Harris: February–December 1942 69

On 14 February 1942, just days before Harris’s arrival, Bomber Command had received a new directive from the Air Ministry. The command was to ‘focus attacks on the morale of the enemy population, and, in particular, of industrial workers.’ Targets in the Ruhr were to be the primary objective: Essen, Cologne, Duisburg, and Düsseldorf were specifically listed, and targets within those communities were included in a detailed annex. Cities in other regions of the country, but beyond range of Gee, the new radio-beam navigation equipment, were also identified as ‘alternative’ objectives. The weight of bombs deemed necessary to achieve ‘decisive damage’ was also provided. Specific direction was given that Gee would be used to the full extent possible.15 Through the spring and summer three additional directives would be issued, each of them adding to the list of targets that the Command was expected to attack.16 Taken together, these formed a tall order for an organization that had been struggling to see in the dark. Part of the solution to the Command’s chronic difficulties, and to meeting the demands of these directives, lay in finding a way to employ Gee effectively. The Shaker technique, described in the previous chapter, had been developed to exploit Gee’s capabilities and now the tactic was employed for the first time, just days after Harris took command, in a highly successful attack against the Renault works at Billancourt outside Paris. The raiders used the Shaker concept, but did not employ Geeequipped aircraft. The attack was deemed a total success, achieving a concentration in time and space that had not previously been seen. The entire raid had been completed in just over one hour and fifty minutes, with an average concentration of 121 aircraft per hour.17 Yet the attack on the Renault works had benefited from optimal weather and weak defences, conditions that could not be expected in the Ruhr, against which Bomber Command had been directed to concentrate its main effort. During March and April the Command made eight attacks against Essen, all with relatively poor results despite the use of the full Shaker protocol. Two hundred and twelve photos were taken (amazingly few given that over one thousand aircraft had participated in the raids), but of these only twenty-two could be plotted within five miles of the city. A variety of problems were reported, the most serious and frequent of which was the failure of the flare droppers to use Gee to determine their release point.18 These sorts of problems had been under review at the time of Harris’s arrival. ORS Report S 31 ‘Statistical Evidence of the Visual Location of Targets by Night’ had been issued to the Groups on 7 March. The purpose of the study was to determine the crews’ ability to see the ground

70

The Science of Bombing

from altitude, and 1162 photos had been analysed. Factors examined included cloud cover, haze, and moonlight (i.e. illumination). The boffins found that haze had a significant effect on the ability to find the target, regardless of the amount of moonlight available; even moderate haze on dark nights would cause considerable difficulties. Several conclusions were reached based on various combinations of factors. The best results could be achieved, not surprisingly, on bright nights with good visibility (moonlight and little cloud or haze), and these conditions would in theory allow about 50 per cent of the aircraft to bomb on target. On dark cloudless nights the maximum altitude for effective target identification was thirteen thousand feet. This figure was linked to the ability of the existing flares to provide illumination. As a corollary it was seen that flares did not work well in moderate or heavy haze. Finally, in the worst case, with poor visibility due to thick haze or much cloud, the success rate fell to about a 10 per cent correlation between presumed and photographic position.19 Number 5 Group, it is worth noting, gave a cool reception to this report.20 As had happened previously, A/V/M Slessor and his staff concluded that the report had nothing new to offer and wondered at the validity of the conclusions in view of the meagre data (something about which the scientists had long complained). Slessor directed that the report not be sent out to the squadrons.21 Harris later commented on Gee’s growing pains. He described the attack against Essen on the night of 8 March 1942 by 211 aircraft in weather conditions that were good except for industrial haze in the Ruhr. The hope was that the Krupp factory would be put out of action; however, photographic evidence showed that there was no damage whatsoever. Subsequent investigation revealed that, even though most of the flares illuminating the target had been dropped in the correct location, a majority of the main force arrived after the flares had gone out, and these tardy bombers had dropped their incendiaries over a wide area. Worse, many of these fell short of the target and duped the followers into bombing short. During the following three months, eleven more attacks were made against Essen using variations of these tactics. ‘The bombing,’ Harris reported, ‘was about a third as accurate as the results of the exercises carried out over Britain. For every hundred aircraft attacking Essen no more than between five and ten sticks of bombs could be expected, by the use of Gee alone, to fall on the built-up area of Essen and only two or three sticks on Krupps itself.’ But Harris did not blame the navigation device, acknowledging that there were meteorological and geographic challenges to finding the targets even when Gee put the crews in the

Arthur Harris: February–December 1942 71

right area.22 He concluded that Gee was not a blind bombing device; rather, it was a good device for getting aircraft into the vicinity of the target from where visual identification could be made. The problem in the Ruhr, he thought, was simply that it was impossible at that time to make positive visual identification. Harris felt that Gee had advantages; in fact, he said, it was ‘indispensable’ for allowing the force to concentrate in time and space thus providing protection from fighters and flak. Gee, moreover, was a marvellous tool for getting aircraft back to home base. Overall, he concluded, he needed two things to allow Gee to work better. The first was a marker bomb or what would come to be called a target indicator or TI, and the second, a new member of the crew, somebody who would act as a bomb aimer, whose functions would not be added to those of the navigator, but who rather would be a specialist trained specifically to sight and bomb the target.23 As Harris had said, there were some notable successes during these months. Cologne was attacked on 13 March with considerable effect. Most of the illuminators reached the objective and dropped their flares accurately and on time, some blindly. This allowed about three-quarters of the incendiary force to do its work and ultimately 50 per cent of the Followers bombed the target area.24 It was estimated that without Gee perhaps only 10 per cent of the force would have found the target. During the same months successful attacks were made against Lubeck (28 March) and Rostock (23–7 April) on the Baltic coast. Both were beyond Gee range but each was destroyed by effective attacks.25 The results from March and April showed that while improvements had been achieved (attacks were about 40 per cent effective versus 26 per cent for the preceding three months), the results were not all that had been hoped for. It seemed the best results had been achieved when crews were still able to bomb visually: ‘It was possible that the crews were not yet making the best use of Gee, it was possible that the Shaker technique was not yet sufficiently developed and refined, and, finally, it was possible that Gee itself was fundamentally incapable of producing sufficiently accurate results to make blind bombing possible.’ Gee was new to almost all crews, who were likely having difficulty using it. Worse, most crews did not have bomb aimers and so the navigator had to leave the Gee set to move into the nose and identify the target. More training and technical modifications to make operating the set easier were offered as solutions, as was the notion of concentrating the force: ‘Also a greater concentration of aircraft in the target area, by reducing the effectiveness of the defences, gives the bombers an uninterrupted run up to the aiming point.’26

72

The Science of Bombing

The OR section had continued collecting data and now, at the end of April, produced a third analysis of night photos.27 In this report they hoped to focus some attention on the success of reaching the actual target, as opposed to the target area. They noted that of 4061 sorties, 2444 had claimed attack. Of these there were 431 photos available for analysis, representing 18 per cent of those claiming attack. This number itself was double the sample available during the previous period. From this data it was possible to conclude that the ‘average success’ was up and that in good weather with moonlight fully 70 per cent of sorties claiming attack could be expected to bomb within five miles of the target; however, in good weather on dark nights that figure dropped to 50 per cent. Despite the encouraging numbers, the scientists still felt that of the more than 4000 sorties only some 550 had bombed within five miles of their respective targets. At 5 Group the Air staff passed a précis of these improvements to the senior leadership, but did not recommend circulating the paper. ‘At the moment we are considering the results achieved by only a comparatively small proportion of the effort. In these circumstances I do not think that this paper is very suitable for issue to stations. It requires very careful reading to obtain a true picture.’ Perhaps, it was thought, when all aircraft had cameras the data would ‘provide a most accurate estimation of results.’ The Group’s SASO was similarly unhappy. He felt that the scientists’ assumption that the photos provided a valid representation of the results of the entire Command was unsound and, in fact, overly optimistic. Photos were, in his view, taken more often by the more effective crews with the weaker ones not pressing to the target. The Group’s new AOC, A/V/M A.C. Coryton, agreed with his staff and directed that the report not be distributed.28 Whether the scientists had been too optimistic or not, Harris was profoundly disappointed by the results achieved. On 22 May he wrote to the AOCs predicting that continued lack of success would surely play into the hands of those who would want to destroy, first, an independent bombing force and, second, an independent RAF. Service politics aside, Harris denounced the ‘fearful waste of effort,’ declaring that if the air crews ‘do not on every operational flight make some worthwhile contribution to the aim of destroying valuable objectives then the whole of the effort that has been put into training and mounting them is being thrown away.’29 In early June frustration with the poor results against Essen led Harris to have the ORS take another look at what should be possible. The

Arthur Harris: February–December 1942 73

resulting study used the available evidence to conclude that an effective pure blind bombing attack by one hundred sorties could drop at least five and perhaps as many as twenty tons on the Krupp munitions works; an additional twelve tons would fall on the town.30 Dickins reminded Harris that these projections were better than anything so far managed and could be achieved under any weather conditions. In this case, he said, on a night when Shaker could not be used, a blind bombing attack would be worth the attempt. If Harris agreed, Dickins also suggested sending the ORS study down to the squadrons so that the flyers would have a sense of what was being done by the scientists.31 Harris was convinced with testing the blind bombing stratagem: ‘Let’s try one.’32 There is no indication in the records whether such a blind bombing raid was conducted, but within the month Dickins had more conclusions about Gee. On 11 July 1942 he minuted Saundby to report new findings.33 There were three main causes for Gee problems: the technical limitations of the system, errors in setting up the coordinates, and errors in reading the pulse alignment. Steps were being taken to deal with the latter issues, and the Radio Direction Finding (i.e. Radar) staff were in contact with the Air Ministry regarding technical issues. But the most audacious suggestion was to come later in the month. On 16 July Dickins floated a proposal to attack Essen from the east.34 In a minute to CNavO, W/C RDF,35 and G/C Ops he forwarded the paper to his colleagues asking for their thoughts before sending it to the seniors. The paper was a detailed review of the twelve attacks against Essen that had taken place between March and June and it reviewed results and factors in extensive detail. Of the few attacks which had taken place in relatively good weather only the single attack from the east had shown encouraging results. Reviewing aircrew morale it was noted that this raid and the just completed massive thousand bomber raids (described in the next chapter) had been relatively successful because, in the initial stages, there had been concentrated flare laying. Other attacks had been less focused from the very beginning, leading to the conclusion that the crews who do penetrate the defences and seek out the aiming point receive such a battering that their determination has to be of the highest order to persevere to the target. The reception given these crews does not, of course, encourage the weaker crews to go in themselves. The process is in fact cumulative – the better the attack is going and the more concentrated it is, the easier it is for the following crews to increase concentration still further.36

74

The Science of Bombing

These observations led to a number of recommendations: higher concentration of attackers; practice flights, so that crews would be comfortable and accomplished at using TR 1335; the use of flares to illuminate a check point from which a short timed run to the target could be accomplished; and the employment of marker bombs, dropped on the aiming point by selected crews. To build morale and enhance the chances of success the scientists recommended trying something innovative: ‘It is considered essential, if it is to be successful, that the attack should be on a large scale, should be concentrated in time, and that weather considerations should be favourable. Any promising new tactics or aids to navigation should be tried. The morale of the crews is lowered when tactics which have failed repeatedly are employed, and under such conditions they will achieve little.’37 Indeed, they said, ‘It is necessary to ensure that the morale of crews is very high for any attack on Essen. This can be achieved if an extra large attack were planned especially if the crews have been made aware that “something special” is to occur.’ The something special was then described in considerable detail with the boffins laying out an attack which optimized all of the elements they had described. The reaction was markedly positive. CNavO called it ‘a most constructive report, the remarks about morale being particularly apposite.’38 G/C Ops commented, ‘I think this paper is very sound and I have no fault to find with any of the arguments. Like CNavO, I am in particular agreement with the remarks on morale. I think we have proved conclusively the inefficacy of our present methods of attacking ESSEN and if we are to attack it again it is essential that we should radically alter our methods. Oboe and the Marker Bomb seem to be the obvious basis for a new scheme of attack.’39 In fact the scientists’ paper had not mentioned Oboe, a new and more accurate bombing device then under development that consisted of two ground stations transmitting pulses which were rebroadcast back from an aircraft and permitted precise bombing guidance. They had, however, been discussing its use in other documents and were happy to add it to the proposal as they now forwarded the revised document to Saundby. He replied: ‘I have made some amendments to this excellent report. It is of such importance that I would like to issue it as a Bomber Command Air Staff Note.’40 When Harris saw the concept he directed that ‘It should go to the [Groups] with a note telling them to study it & brief the crews from it on the next Essen attack.’41 The credit for using Oboe to drop flares over the target goes to Dick-

Arthur Harris: February–December 1942 75

ins and the staff. At about this time Mr A.P. Rowe at TRE had written to Sir Henry Tizard to report on the proposal. Rowe ‘“wanted him to know of a brainwave Dickins had had regarding an early operation.”’ Rowe thought that, once the first flares were in place, follow-up aircraft could then ‘bomb the flares and drop more flares.’42 Indeed, the Command had had little initial interest in Oboe because of its limited ability to handle significant numbers of aircraft which made it impractical for raids of any size. The ORS had now rethought the problem and concluded that if Oboe could be made to handle one aircraft every three minutes then these aircraft could mark the target for the Main Force. These ideas were put forward in ORS S 53 (discussed below) and accepted by the Air Staff, who then asked for all possible priority in fielding the device.43 Even though Gee and Shaker had contributed to marked improvements in efficiency, the results of the first half of 1942 still suggested that Bomber Command was largely ‘a “fair weather” and an “easy target” forces.’44 This view was in keeping with the opinion held by many in the command that ultimate success in the bombing campaign would require the creation of a specialized target-finding force able to identify the target for the followers regardless of weather or defences. Shaker was in fact predicated on using selected crews in the illuminator and target-marker roles. Indeed, the idea of such a specialized force had been circulating for several months.45 In early November 1941 Saundby had distributed a paper which dealt with the problem of locating targets at night and proposed the formation of ‘special target-locating squadrons.’ There is no indication if the paper might have come from the ops staff, the navigation staff, the ORS, or indeed if it was something that had been commissioned by the SASO or AOCinC or even the Ministry. In any case, the response from the Groups was evenly split: two firmly against and two open to the idea.46 While Harris was fundamentally opposed to forming what he called an elite force, independent from the regular squadrons, he was not in fact against the target-finding concept. He himself was frustrated by the inefficiency of raids, stating: ‘somehow or other we must cure this disease, for it is a disease of wasting bombs wholesale upon decoy fires.’47 But Harris’s greater frustration would seem to be the result of irritating young officers in London. Group Captain S.O. Bufton, the Deputy Director of Bomber Operations and a former Bomber Command squadron and station commander, and his cronies presumed to have Bomber Command’s problems neatly identified and solved. All Harris had to do was follow their counsel – something few commanders would be inclined

76

The Science of Bombing

to do when they had just been appointed to make good their organization’s woes! Harris and Bufton were at odds over the tactical aspects of pathfinding, Harris feeling that the bombers could not achieve precision results, Bufton believing that eventually they could, ‘for it was, of course, the ultimate aim which conditioned the processes of design, scientific investigation and production which resulted in the equipment, and of training which resulted in operational aircrews.’48 The debate over the formation of a formally constituted target-finding force dragged on into June and it was only when the CAS ruled in favour of the Air Staff’s proposal that Harris was obliged to set up the Pathfinder Force (PFF) under Group Captain Donald Bennett on 11 August 1942.49 Ironically, the Germans had begun to jam Gee just days before, and thus the PFF would be forced to operate for some time without any better equipment than had been available prior to the beginning of the year and also without any better target-finding aids than those of the main force followers.50 Yet once the PFF was in operation improvements became evident. The most obvious of these was an enhancement in concentration. From March to August some 35 per cent of bomb release photos had been plotted within three miles of the centre of the bombing concentration (also called the Mean Point of Impact or MPI). That figure rose to 50 per cent in the period from August 1942 to March 1943. On the other hand, accuracy of the attack was equally important and here the PFF was less successful: from March to August 32 per cent of night photos were plotted within three miles of the aiming point, but this value rose to only 37 per cent for the subsequent period. The spread between the concentration and accuracy values became known as the systematic error and led to vigorous action in attempts to find a way to ensure that the PFF could achieve greater accuracy in marking the target.51 The nature of these PFF problems had been distilled by the scientists and Harris was most impressed by their work. In his words: The bombing results after the formation of the Pathfinder Force were investigated by the Operational Research Section – a body of brilliant young civilian scientists and technicians at Bomber Command headquarters, who did work of inestimable value in subjecting all aspects of our operations to an impartial scrutiny – and it was discovered the bombing concentration had undoubtedly increased, but at the same time another factor made the attacks as a whole hardly more accurate than they had been before. The bombing was certainly more closely concentrated round the marker

Arthur Harris: February–December 1942 77 bombs than it had previously been round the aiming point, but there was no advantage in this unless the marker bombs had in fact been dropped on aiming point, which was not always the case. Any error in the placing the marker bombs led the whole attack astray.52

The scientists had been involved in the ‘Assessment of PFF effectiveness’ from the outset, opening a file with this title on 22 September 1942. At that time they submitted a short review of PFF results to the CinC, as he had previously requested. From a detailed examination of thirteen raids they concluded that results within Gee range were more successful than against more distant targets: ‘This suggests that [the Pathfinders’] main advantage has been in concentrating the attack by means of target markers.’53 Outside of Gee range, accuracy was not as good, but concentration was still much improved. The boffins also reminded Harris that the PFF crews were still fairly new to the business and that improved results could be expected with time. Acknowledging his thanks, Harris asked for another review after a further twelve ‘shows.’54 This next assessment of the PFF was conducted in November with a report being tabled near the end of the month. Covering the full period of PFF employment, operations from thirty nights (twenty-one of those over Germany) had been reviewed. Of the twenty-one German raids, thirteen had been fully or partly successful. Five had failed and three ended up over the wrong target. Only 24 per cent of attacks were found to have been within three miles of their targets, which gave ‘much scope for improvement, as it represents only about 5% within 1 mile.’ The scientists felt that impending introduction of H2S, an aircraft-mounted ground-mapping radar, and purpose-built marker bombs (Target Indicators) would lead to better results. Significantly, the boffins noted that, with bad weather being a factor on about 30 per cent of the nights, the use of H2S by the entire force might lead to ‘considerable further improvements.’ They were not so gullible as to think that just any navigator would be able to master the on-board ground-mapping radar and called for the urgent provision of data on this question.55 There is no indication of the impact of this short report, but it would undoubtedly have been bittersweet for the aircrews, for here was another attempt to find a cure for the fog of bombing that seemed to have a number of bugs which had yet to be worked out. At least the scientists had begun the necessary analysis. Although target markers were needed, the answer to accuracy lay in ensuring that the PFF itself could achieve a better standard of naviga-

78

The Science of Bombing

tion and target recognition. And hopes for these improvements were in no small way tied to the arrival of new aids such as Oboe and H2S.56 As mentioned, the limitations of Oboe, in particular its ability to guide only one aircraft at a time, meant that some solution needed to be found to maximize its obvious potential. On 18 June the boffins sent Report S 53 ‘The Operational Use of Oboe Mark I for Target Location’ to Saundby. The extensive analysis described the service trials of Oboe and proposed a means of integrating it into the existing target-finding process; it became the first document in HQ BC File 900 ‘Operational Use of Oboe (Blind Bombing).’ The scientists had reviewed the trials of the device and emphasized the excellent accuracy achieved in bombing: ‘50% of the bombs fell within an ellipse approximately 400 yds. × 200 yds.’ In addition they noted ‘that there is an immediate application for Oboe Mk. I as an aid to target location.’ They described the process for getting the Pathfinders on to the target: ‘It is proposed that an Oboe aircraft should drop coloured marker bombs accurately over the target at Zero hour. These bombs burst at about 2,000 feet, giving an unmistakable display. The burning elements take about 20 seconds to reach the ground, where they burn for about five minutes. The Target Finding aircraft would therefore start with an unmistakable mark over the target which would last for about five minutes.’ The report then described how the first and then a second Oboe aircraft would continue to mark the target over a forty-minute period, but at a rate of only one marking every ten minutes owing to the technical limitations of the system. ‘It would of course be an advantage if the Oboe aircraft could drop their marker bombs every 5 minutes and thus keep a continual mark going at the target, but even if this cannot be done, it is thought that they would have a most valuable effect in assisting the Target Finding Force to keep on the mark.’ The scientists went on to suggest the composition of the Oboe force and the tactics needed to get Oboe markers over the target in sufficient numbers to affect the process. One of the requirements was for the ‘provision of marker bombs.’ Finally, unlike most ORS reports, S 53 ended with recommendations the most important of which stated: ‘It is recommended that the use of Oboe Mk. I as an aid to the location of targets in the Ruhr should be put up as an extremely urgent operational requirement.’ Other recommendations dealt with the system needed to effect the proposal and called for the establishment of two ground stations, the provision of ‘six suitable aircraft fitted on the highest priority,’ training of crews, and the selection of navigators with Gee experience and pilots with beam-flying expertise.57

Arthur Harris: February–December 1942 79

On 21 June, only three days after the report’s production, Saundby, with Harris’s approval, sent it to the Air Ministry with a strong recommendation for immediate action.58 The Ministry responded within just four days, not only accepting the concept, but making several proposals to improve the system. Key among them was the suggestion that ‘we should ensure to the utmost against failure at this vital stage of the operations by duplicating both the ground stations and the “OBOE” aircraft. This would make it possible to drop two sets of markers every ten minutes or alternatively to stagger the two systems and drop one set every five minutes, thereby effecting a more frequent refocusing and reducing the chances of failure.’59 The scientists had clearly hit upon a winning tactic for precision engagement of targets as far away as the Ruhr. The introduction of Oboe and then H2S would not come until the end of 1942 and would mark the beginning of the next phase of the bombing campaign. With them would come the transition from setting right the problems of the past to flexing the muscles of the bombing force with new resolve.

6 Concentration and Other Curatives: February–December 1942

If on the plus side of the efficiency equation the crews and the headquarters were learning how best to use the new technologies that were coming into service, they had also to deal with increasingly deadly defences. Indeed, the bombers were not only confronted with challenges of finding the target, but they also had to contend with the rapidly improving German air defences. A fairly complete picture of the system was available by May 1942 and it had become strikingly apparent that the previously ‘haphazard’ routing of raids was no longer possible. ‘It was also clear that concentration both in time and space was vital, not only over the target, but on the route as well.’ By sending in an attack en masse, the defences could be overwhelmed. As recorded in Harris’s Despatch: ‘“The tactical aim of Bomber Command in this period can be described in one word ‘concentration.’”’ In fact, one of the aims of the thousand bomber raids that would be launched in the summer was to show that concentration could lower losses.1 Indeed, in the preceding months considerable success had been achieved in concentrating the attackers; the thousand raids were intended to put up a big force that would really saturate the defenders.2 Harris himself was convinced the force of 250–300 aircraft was wholly inadequate to saturate the then existing defences of a major industrial town of half a million or more inhabitants. But if we attacked with a larger force, supposing that we could get one, should we be able to organize it in such a way as to get a really high concentration over the target? It was becoming obvious that the degree of concentration we had achieved so far was not going to be good enough, but there was no previous experience to show whether it would be practical to put many hundreds of aircraft over a target at a rate of, say,

Concentration: February–December 1942 81 ten a minute; in the past such a concentration would have been thought as absurd as it was dangerous, for this was six times as high concentration as had been aimed at towards the end of 1941.3

Much of Harris’s thinking on this matter came from extensive ORS investigation that had been taking place throughout the winter. Dickins and the scientists have been given credit for developing and expounding the concentration technique. According to Dickins, ‘We had to reduce it all to mathematics, and to work out the actual chance of collision. And it became quite obvious to us at ORS that while a collision was a half per cent risk, the chances of being shot down by flak or fighters was at three or four per cent risk. So we could allow the collision risk to mount by quite a bit, provided that in doing so we could bring down the losses from other causes.’ The scientists ‘envisaged ten bombers crossing a given point every minute which meant ten aircraft scattered randomly in a box of sky three miles long by five or more miles wide and nearly two miles deep.’4 Concentrating the raiders in time and space had, in fact, been a topic of repeated discussion, and often debate, for the previous year, and the ORS had been caught up in the subject from the outset. On 30 August 1941 G/C Ops had minuted the SASO concerning a 4 Group study which indicated that losses seemed to decrease proportionally with increased concentration. While not convinced of this relationship himself, he felt bound to pass the matter along. Two days later Saundby asked Dickins to review the issue. Dickins, too, was not ready to make a conclusion but said that he would be investigating the question in depth. ‘In the meantime,’ he said, ‘I would endorse the proposal to concentrate aircraft in time and place so that while over enemy territory as many aircraft as possible are in the same “room” at the same time.’5 In November the subject of concentration resurfaced. Now, in reviewing a letter from the Deputy CAS (D/CAS – A/V/M N.H. Bottomley, Saundby’s predecessor as SASO at High Wycombe), G/C Ops noted that actual concentrations were only about half of the desired value. The SASO and CinC agreed with the observation and with the Ops staff’s recommendation to try doubling the planned concentration in an attempt to get the actual rate to the desired level. This exchange may well have been precipitated by an ORS report completed at the beginning of November. In it the scientists referred to two recent attacks: In the case of Cologne it would appear that the average loss per sortie does in fact increase as the concentration over the target decreases. In the case

82

The Science of Bombing of Frankfurt the same thing is noticed, if we accept one night which corresponds with the night of full moon. Cats-eye fighters may have been used on this night. From the very flimsy evidence available it would thus appear that for attacks on these towns there is something to be gained by concentration.6

On 3 December the D/CAS wrote again, this time to criticize the plans for increased concentration, recommending that there needed to be concentration not just over the target, but also over defended areas. W/C S.C. Elworthy, the understudy to the G/C Ops and, a future CAS, indicated to his seniors that the problem in replying to these criticisms was that there were no effective data for measuring the ‘efficacy of concentration in time and space.’ Dickins, however, proposed experiments for varying concentrations in time of between two hundred aircraft per hour and fifty per hour and in space with varying penetrations of defended belts using raid frontages from twenty to fifty miles. After some heel dragging by the Groups (as they would be called upon to make annoying modifications to take off times should the weather change throughout the day of a raid), the proposals were implemented.7 The scientists completed their study of concentration with respect to time on 11 March 1942, concluding that a concentration of less than 50 raiders per hour had a measurably higher loss rate than those of that or greater density.8 The boffins were still not convinced, however, that their data were conclusive and indicated that the subject would bear watching. ‘The results,’ they said, ‘confirm however, the suggestion that high concentration will help reduce the losses and raids are being planned accordingly. The effect of higher concentration will be watched with a view to confirmation of this present trend.’ The scientists recommended a concentration of at least 50 aircraft per hour when there was no moon and at least 80 aircraft per hour on moonlight nights. Additionally, they recommended larger raids, as the loss rate appeared less for large forces than for small ones. In sending this study to the Ministry, Saundby stated that the report had led to increases in concentration on recent raids using the figures proposed.9 To achieve this density, the letter said, the staff were actually planning concentrations of 150–200 aircraft per hour with the aid of Gee. Experiments and observations continued, and by the end of May 1942 the Section was able to report a definite relationship between concentration (both in time and space) and losses. As concentration increased, losses dropped, particularly when the moon was down. At Dickins’s rec-

Concentration: February–December 1942 83

ommendation this updated information was also sent forward to the Air Ministry on 4 June.10 On the same day, Air Commodore Harrison, the D/SASO, asked Dickins for an abridged version of this report as he wanted to circulate it to the stations for presentation to the crews.11 If concentration could theoretically swamp the defenders at a given time and place, the logically related concern was to ensure that the raid would avoid areas of heavy enemy defences. If it could be predicted where the defenders were, then the force needed accurate navigation to get to the target relatively unscathed by avoiding those locales. Here too the ORS had been studying the available data. In early February 1942 the responsibility for statistical analyses on losses shifted formally from the Intelligence staff to the scientists. Saundby had become concerned that there was a duplication of work and asked the Chief Intelligence Officer (CIO) to confirm the coordination between the two offices. The Intelligence 3 section was able to report back that the ORS was indeed using data provided by the intelligence officers to prepare their analyses. Despite this reassurance, G/C L.F. Pendred, the CIO, felt that there had been some doubt of his staff’s data even as high up as the Air Ministry, where DB Ops felt that the ORS should be the only source of long-term analyses and conclusions on losses. Pendred suggested that it was time that his staff got out of the business of long-term data collection and that this would be best left to the ORS. He said that it would be unacceptable for the CAS and CinC to receive conflicting reports. Dickins for his part confirmed that the ORS was now preparing long-periodicity reports and that they would continue to work closely with Intelligence 3 in assembling data. One can imagine that the Intelligence staff, which had until this time been amassing, analysing, and reporting various data, would have reasonably assumed that this was their mandate and that they should continue to do this work. While the file does not indicate any emotion at the working level, one may well presume that the reassignment of this traditional Intelligence task might have been somewhat awkward or embarrassing. Nevertheless, the SASO immediately approved this realignment of functions.12 Thus, in April, on receipt of a query from the Air Ministry, Saundby turned to Dickins to look into the continuing high loss and damage rate of the Manchester bomber. A/Cdre J.W. Baker, the Director of Bomber Operations (D/B Ops), had listed a number of possible causes and had indicated that he would be grateful if the ORS could add this investigation to their already full schedule. At Saundby’s request13 the scientists had reviewed the matter in some detail by the end of May, comparing two

84

The Science of Bombing

heavy bombers types. They now felt the higher losses of the Manchester were due both to its reliance on just two large fuel tanks compared to the Stirling’s fourteen fuel cells and to the Manchester’s two liquid-cooled engines compared with the Stirling’s four less vulnerable radials. Somewhat naïvely perhaps, Dickins thought the information would be of interest and asked permission to circulate it. Saundby replied immediately telling Dickins not to distribute the obviously bad news to the Groups. The final version of the report was sent to the Air Ministry on 4 June. It indicated that the fuel and engine issues were the most likely, but as yet unproven, causes for the higher losses; all other possibilities had been effectively discounted. In his covering letter Saundby went further, saying that the fuel tank and engine problems seemed the likely causes. Meanwhile he also minuted the file to Harris so that the AOCinC would be sure to see the data for himself.14 The scientists had been tracking the causes of losses and would continue to do so. Concentration was proving to be a major curative, but there were many aspects of casualty analysis, such as the different loss rates between aircraft, that needed to be studied further to achieve a complete understanding of the complex warp and weft of the bombing campaign. In February 1942, for example, the boffins had had the opportunity to examine the circumstances of non-combat casualties. In ORS Report 28 ‘Preliminary Note Comparing the Number of Casualties Not Due to Enemy Action of Aircraft on Operational Sorties During the months October–December, 1940 and 1941,’ completed on 24 January 1942,15 they reported that losses as the result of crashes and forced landings (from all causes) had actually dropped from 4.5 to 2.8 per cent of the sorties dispatched. Losses over enemy territory were up by comparison, so they cautioned that the overall difference might be less. The report highlighted engine problems on the Wellington and caused the engineering staff to examine what might or might not be behind the phenomenon, leading them to acknowledge that the aircraft had been operating at its design limits for weight and endurance, thus putting a significant strain on the engines.16 The scientists put out another report in early April 1942. This one concluded that the accident rate had remained steady at about 3 per cent of sorties over the period, with 70 per cent of those resulting in major damage or loss of the aircraft being due to fuel shortage, landing accidents, and engine failures, ‘and it is considered that a more detailed investigation into these crashes would be well repaid.’ By comparison the scientists reported that losses due to enemy action had risen from 1.9 to 3.1 per

Concentration: February–December 1942 85

cent. The study also noted that Whitleys, ungainly and slow compared to the newer bombers, had the highest losses of any of the types looked at. In submitting the report to Saundby, Dickins repeated the key figures, also noting that weather had been a contributing factor in just over a third of the accidents.17 He also thought that the Air Officer Administration, responsible for aircraft maintenance, and the AOT, responsible for training and crash investigations, should see the draft before it was published. The AOA’s reaction seemed fairly low key. His only comments related to the non-standard format that the scientists had used in typing the report. The AOT thought the document ‘a valuable statistical report’ but wondered why the scientists were repeating the data collection that was already done by the statistics staff at the Air Ministry. He did recommend that the report be circulated to Groups but not to the units, particularly given the Whitley losses.18 It was towards the end of May that the report was ready to present to Harris. Now, however, Dickins was able to report that the AOA and the AOT would in fact be investigating the causes of fuel, landing, and engine related accidents more fully. Harris’s reaction was positive: ‘Very instructive. Send to groups. Nav and AOT should investigate the cause and cure of the high % of fuel shortage accidents not due to weather.’19 A subsequent undated minute (but probably in the late June period) indicates that work had been done and there was now some urgency both in providing pilots with better data on fuel management, and in installing fuel flow meters in aircraft.20 Work was also being done to minimize errors in navigation and the advent of new navigation aids was seen positively. More attention would in future be paid to losses due to accidents. Starting in August 1943 the Command put out a monthly ‘Bomber Command Flying Accident Review’; Issue 9 in April 1944 included a graphical history of accident rates, which had dropped appreciably in the spring of 1942 and again in 1943. By early 1944 the accident rate, had fallen to approximately .002 per cent of all sorties.21 It was also in the spring of 1942 that the scientists began looking at the effect of icing on missing rates. Dickins asked the Chief Engineering Officer (CEngO) to review the scientists’ findings before sending the data to Saundby. Comments from one engineering staff member included the criticism that there would be a greater number of casualties due to enemy action on nights with no icing as these nights were normally the clear nights when fighters were present. Another engineering officer ‘dared’ not draw any conclusions from the data. Dickins accepted the first comment, but in his note to the SASO said that there was indeed sufficient data to draw conclusions. Subsequently the report, which Dick-

86

The Science of Bombing

ins noted did not attribute any undue losses to icing, was seen by Harris, but there is no indication of his reaction.22 Towards the end of 1942 the Air Ministry did, however, institute a monthly Icing Return, which continued throughout the war. The October 1942 report showed that 3 per cent of all sorties had reported icing and that about half of these had aborted for this reason.23 In May, after the preparation of a draft report the previous month, a first attempt at assessing losses over the year leading up to April 1942 was made. The boffins concluded that losses (which had initially been in the 3–4 per cent range) had been creeping up until the fall of 1941, but that there had been a respite over the winter. This, they said, was due to the nature of the targets attacked during the winter months, largely coastal towns in occupied Europe, and so these figures should not be compared with the higher losses over heavily defended Germany. Their overall conclusion was that ‘our own tactical measures and the use of more powerful bombers and greater concentrations are approximately keeping up with any improvements in the enemy defences.’24 Also in May, following a report on Whitley losses, Dickins suggested to Saundby that there seemed a possible link between liquid-cooled engines and higher losses. You will remember the theory put forward some time ago that the higher losses of Whitleys was due to their liquid cooled engines. While we now know that it was largely due to their lack of manoeuvrability we have thought it worthwhile writing up the results of an investigation done some months ago. I accordingly submit the paper ... for your approval prior to circulation. We still believe that there may be something in the liquid cooled story – the Halifaxes, Manchesters, Wellington IIs and Whitleys all have higher losses than the others and we are going into it further.25

Included was data for the period from June through December 1941 which showed that among the heavies the air-cooled Stirlings had a 3.7 per cent loss rate, comparing favourably to the 4.5 per cent for liquidcooled Halifaxes and 4.8 per cent for the previously described Manchesters. This work would prove valuable in the critical review of the Halifax which was soon to begin. Dickins sent Harris a preliminary investigation of Halifax losses on 10 August 1942.26 The scientists had identified three potential causes for high casualty rates. First, Halifax squadrons might be staffed with largely inexperienced personnel. Second, ‘the greater visibility of Hali-

Concentration: February–December 1942 87

fax exhausts namely, about 1600 ft. directly astern, enables an attacking fighter to attain and maintain contact more easily.’ Last, the authors indicated that there was ‘some doubt as to the stability of the Halifax making evasive turns when fully loaded and equipped, due possibly to lack of reserve speed.’ More work, the boffins stated, would need to be done to confirm the first and last suspicions. Within days Harris thanked the scientists for this work, indicating that the staff was already taking action with respect to flame dampers and aircraft stability. The CinC concluded his remarks saying: ‘a further point for consideration is that the Halifax – being the slowest – normally brings up the tail of the procession home after an attack and is therefore more likely to get nibbled out by the fighters? [sic]’27 Dickins was quick to get back to Harris, stating that the proportion of Halifax aircraft leaving the target late over the period of 14 weeks considered is not greater than any other type. The Halifaxes have only infrequently been last over the target and the figures for losses on these occasions are too few to draw positive conclusions. It is possible that returning aircraft not do not follow a planned route so well as those going in and the geographical positions of the Halifax Stations may have produced a special inducement for this type to go astray.28

In early October Dickins was able to provide an update on the handling issue, and coincidentally on the matter of engine exhausts. He wrote, further evidence of the importance of improving manoeuvrability, which includes performance, is given by the facts that: (i) operational experience of flying the aircraft, regardless of the experience of other types, has been found to affect the pilot’s chances of survival very profoundly. (ii) a further case of a Halifax returning after being out of control and turning on its back occurred during the raid on Kreveld in October. It is therefore recommended that all possible means of improving performance and stability of the Halifax be given urgent priority.29

In the same note Dickins indicated that the CEngO would be corresponding directly with the CinC on the matter of improving Halifax performance and stability. Dickins went on to say that engine exhaust covers were having a negative impact on speed. Work was being done by Aeroplane & Armament Experimental Establishment (A & AEE) and talks had been held with Handley Page, the manufacturer of the Halifax; it seemed that

88

The Science of Bombing

the exhaust covers now in service were causing excessive drag. Dickins felt that the most effective short-term solution was to remove those covers and to attempt to use better anti-glow paint on the exhausts. Regarding Handley Page’s efforts to produce better covers Dickins said: ‘These, if ever available, are bound to be long-delayed and it seems therefore most desirable to prepare for flying without shrouds.’30 It appears clear from these words that Dickins shared the Service officers’ doubts about Handley Page’s ability to measure up. The boffins had been interested in the effect of experience on losses for some time when they published their first study in the fall of 1942.31 In Report B 116 they examined the Halifax losses in 4 Group. Sending the draft to the SASO, Dickins commented on the value of having pilots gain experience on easy trips before heading for Germany. It will be seen from the tables that there is a marked increase in the number of sorties performed before becoming missing as the number of trips as second pilot or against lightly defended targets is increased. The figures suggest that experience in flying Halifax aircraft against lightly defended targets is very desirable before going on to main targets, and that if each pilot could do three to five trips as second pilot or as first pilot against lightly defended targets, a substantial reduction in the overall missing rate against main targets would be achieved.32

Dickins offered evidence of these conclusions just days later, drawing from the results of recent attacks against Flensburg. Of fifty-five total sorties there were twelve missing among inexperienced pilots and just five for those with experience. Saundby minuted the data to the AOT asking for comment. The latter admitted both that there were generally higher losses among new pilots, and also that the Halifax was more difficult to handle than the Lancaster or Stirling and might require longer conversion to type training. By the end of October the report was ready for release and Dickins sent it to Harris. He repeated his comments, but added some thoughts about experience on other types. It had been noted that ‘it is experience in flying Halifaxes which is important and not merely operational experience since pilots with operational experience on other types have a shorter operational life on Halifaxes than those who have had several second pilot trips [on Halifaxes] before becoming captains.’ Finally, Dickins said, the scientists were now examining the experience of 3 Group to see if the results were unique to Halifaxes. In response Harris asked for the work to be completed as soon as possible

Concentration: February–December 1942 89

and then to have a review done of 5 Group.33 It was clear that the CinC saw the value of this analysis. On 14 November AOC 4 Group wrote Saundby to provide details on the Group’s plans to address the problems. To build up experience they hoped to find ‘fresher’ targets where possible and to use recently started mine-laying operations as well. Cross-country flights were being instituted and addition training trips would be implemented including at least two trips as second pilot. The investment in training would mean a delay in getting new crews on operations, but it was considered a worthwhile undertaking. Saundby passed the document to Harris, who responded the same day, if somewhat grumpily: ‘He can try it. It is the aircraft not the crews that require improvement.’34 When word of the ORS study reached the Air Ministry a request came for the information. The boffins prepared a précis of their earlier work and passed it up the chain of command for approval before sending it out. Saundby recommended it for distribution: ‘[It] is purely factual and expresses no opinions and I therefore suggest that, if you agree, copies might be sent to the Air Ministry.’35 Harris did not agree. While these researches are invaluable to enable us to see where we [are] going and what to do about it I am not prepared for the existing conclusions to go outside the limited circle of Heads of Branches in this HQ. The figures could only (and perhaps wrongly) satisfy the curiosities and/or arouse alarm and despondency. When we have investigated Lancs, Stirlings and Wellingtons we may or may not find something really startling beyond the high casualty rate on Halifaxes which we already realise.36

The boffins completed their analysis of Lancaster loss rates by Christmas. In looking at the results provided by 5 Group they concluded that there was no apparent correlation between experience and losses. They did observe that the restriction of inexperienced crews from the more difficult missions at the start of their tours would likely have a mitigating effect on losses. Overall the loss rate in the Group sat at 2.9 per cent, which meant that a crew had a 42 per cent chance of completing a full tour. Both Saundby and Harris acknowledged the study, but made no observations.37 In determining what in fact was happening, part of the challenge for the scientists as well as the flyers was the nature of aircraft casualties themselves: there just did not seem to be that many aircraft coming back to home stations with significant damage. In search of clues, the scien-

90

The Science of Bombing

tists studied the effect of anti-aircraft fire in conjunction with Brigadier Schonland’s Army Operational Research Group. The results suggested that bombers could absorb a fair bit of AA fire, but it was also noted that returning damaged aircraft usually had few hits. The scientists hypothesized that losses must then be due to some catastrophic in-flight circumstances. More work led to a focus on in-flight fires and explosions of which there was significant eyewitness evidence. The conclusion reached by the ORS was that losses could be reduced through the use of inert gas within the fuel tanks to minimize the chance of fire. As a result, by April 1943 Bomber Command was ‘demanding’ that all aircraft be so equipped.38 The actual work was much more involved than these words suggest. The Bomber Command file on ‘Fire Risk in Bomber Aircraft’ shows that throughout 1942 there was considerable technical and scientific discussion about the causes of in-flight fires. But it was only at the end of the year that Dickins approached Saundby and the engineering staff with a proposal to firm up the evidence. Dickins pointed out that considerable work had in fact already been done on the development of an inert gas system that would replace the combustible fumes left in tanks as the fuel was expended, but there was still a need to ‘settle the technical facts so that the desirability of introducing a scheme of this sort may be decided as soon as possible.’ His bottom line was unequivocal: ‘Our analysis of bomber losses has suggested that the proportion of aircraft set on fire by enemy action is considerable and it would appear that if the fire risk in the petrol tanks can be reduced, a substantial reduction in losses might be achieved.’39 The associated report, S 74, suggested that a minimum of 40 per cent of losses were attributable to fire and there seemed strong indicators that flak – particularly at altitude – was the prime cause for the fires. What was needed now was a series of detailed experiments to confirm these conclusions and to point to definitive corrective action. The reaction by CEngO was likely somewhat surprising. In a long communication he said essentially that fires were going to happen and needed to be accepted. Is it not unreasonable to expect that an aircraft shot down, carrying 1,000 gals. of petrol would go on fire, as I imagine, except for a lucky shot, the aircraft are pretty well shot about before they go down? I am all for doing everything reasonable to prevent fires in aircraft, but when one stops to think a four engined bomber can be broadly described as four incandescent masses entirely surrounded by oil and petrol pipes, the

Concentration: February–December 1942 91 whole backed up by thousands of gallons of petrol in light suitcases. Add to this the fact that incendiary bombs are carried and also subject to enemy action. I suggest that the number of fires due to enemy action is extremely low, especially if they are compared with the number of sorties. I feel that the suggested 230 lbs. additional weight for the nitrogen apparatus is not justified.40

He based this recommendation on his belief that most fires were due, not to enemy action, but to fuel leaks which allowed fuel and vapours to come in contact with hot engines. There was apparently no consensus on this opinion, even among the engineers, as some days later G/C Eng noted his belief that ‘the development of fireproof tanks surrounded by inert gas should be actively developed although it does mean an increase in [aircraft weight].’41 The matter was taken out of the staff’s hands to some extent by the receipt of a letter from 5 Group explaining their experience with in-flight fires. Based on this new evidence, and referring to the scientists’ previous work, the SASO asked the boffins to prepare a paper for submission to the Ministry ‘with definite recommendations for the reduction of fire risk.’42 Trials soon followed and by April 1943 it was possible to conclude, following a demonstration for Dickins, Air Cmdre Roach (CEngO), and their staffs at the RAE, that the explosion risk to fuel cells at operating pressures and temperatures could be ‘completely eliminated’ by the use of an inert gas. In these circumstances the ‘immediate’ introduction of an inert gas system of some ‘230/250 lbs’ was strongly urged.43 It was Roach who passed the report to Saundby, and his thinking seemed clearly to have changed in the intervening weeks. There is no doubt that the explosive mixture in our petrol tanks, which exists during the greater proportion of most of our operational flights, constitutes an extremely serious hazard and the elimination of it should lead to a substantial reduction in losses. It is recommended, therefore, that immediate action should be taken to incorporate a system for the introduction of nitrogen into the tanks, which will eliminate this hazard, on the highest priority. A demonstration was also given to the proposed method of dealing with fires in tank bays which may result from the puncturing of self sealing tanks below the petrol level. There is little doubt that the system developed for dealing with this would be successful in extinguishing the majority of tank fires and it is recommended that its development should proceed on the highest priority. If you agree with the above proposals, it is

92

The Science of Bombing suggested that a strong letter, preferably over the C.-in-C.’s signature, be sent to the Air Ministry requesting that these developments be put in hand on the highest priority.44

What had started with the boffins attempting to connect a number of rather disparate studies of in-flight fires had now taken a definite and crucial shape. There was more to losses than faulty aircraft design and systems; the crews themselves could on occasion prove to be their own worst enemies, and the scientists had to be ready to report these alarming and at the same time discouraging incidents. In the closing days of 1942, for example, Dickins had sent Harris two commentaries on a costly Baltic raid mounted on 17/18 December.45 The accompanying report indicated that of the 104 aircraft dispatched 18 had failed to return. The analysis described, and Dickins further emphasized, that fully one-third of the returning aircraft had been flown at altitudes above 2000 feet, making them perfect targets for GCI fighters, particularly as the raiders had failed to maintain the planned concentration. Dickins pointed out, too, that about 25 per cent of the losses were due to light flak, which was very effective at altitudes up to 6000 feet in good moonlight (as had been the case on the night in question), and most lethal in the 250–2500 foot band. Saundby acknowledged the finding, but Harris made a stern comment. It had been understood, he said, that there was light flak in the region, but the idea was to plan the route through the most lightly defended areas so as to avoid it. Given the excellent visibility, it should have been possible to fly the route at low altitudes thereby ‘entirely’ negating both flak and GCI fighters. ‘If orders are not followed by reason of disobedience or viability then the aircraft so flown make the worst of all three dangers.’46 The reverse of the question of knowing what the flak and fighters were able to do was the ability to navigate around them; taken as a whole these symbiotic needs presented a classic operational research relationship. Gee would undoubtedly be of benefit, but navigation generally was another of the flyers’ weak points. Harris had written to the Air Ministry just weeks after taking over the Command to indicate his concern. He focused on the problems which he believed had confounded effective navigation: a. The non-availability of certain navigating instruments essential to the efficient prosecution of bomber operations.

Concentration: February–December 1942 93

b. The poor service prospects of the air observer. c. The lack of Squadron Navigation Officers. d. The lack of attention given to the navigational aspects of the tactical planning of air operations. e. The poor navigation training given to pilots. The CinC expressed concern that there was too much reliance being placed on radio aids to navigation (such as Gee) and too much authority being given to the signals branch, which operated these systems. He pointed to the absence of a navigation branch in the Air Staff saying that navigation was more than just a training matter, from which it may be reasonably inferred that the previous concept of navigator production was that once trained, navigators were expected to be competent and self-sufficient in their duties.47 The boffins suspected otherwise. The scientists had been studying the questions surrounding navigation for some months. As mentioned briefly in chapter 4, December 1941 had seen the conclusion of a large trial which attempted to gain a better understanding of what crews actually thought they were seeing at night. The Section gauged from this that there was a ‘big difference in the reliability of various ground features.’ Lakes and rivers, although very popular, had proved to be particularly unreliable, while coastal features including docks were relatively ‘trustworthy.’48 The boffins’ report was circulated to all command organizations and emphasized the requirement to focus on map-reading skills, both in theory and in practice.49 They were evidently not reluctant to offer criticism but, as importantly, the Service officers did not shy away from getting the scientific views out to the flying stations. Harris was also using the information that the scientists were collecting to make changes to tactics. He opened a Commander’s conference in mid-April 1942, for example, by stating that he wanted to focus discussion on the use of TR 1335 and a Target Finding Force, but had a few general concerns that needed airing. The first of these was his desire to have raids against the Ruhr approach from the south both to avoid the areas of greatest defence and also to spread out the night fighter force. The group commanders complained that this was a more difficult approach from a navigation perspective, but the CinC ended the discussion by asking the AOCs to impress upon their crews the need for this scheme as it would avoid the heavily defended zones and help reduce losses.50 In a similar vein, the next month Saundby asked the ORS to comment on letters from 3 and 4 Groups which recommended various means to shake up routing with a view to minimizing losses either by avoiding

94

The Science of Bombing

high-threat areas or swamping them through concentration. The Ops staff had already reviewed the letters and provided comments, but Saundby asked for the scientists’ views nonetheless. The boffins now looked at the shift from multiple routes to the single route policy adopted in March. Statistics were reviewed and no discernible advantage could be found for using indirect routes. The scientists reminded the Air Staff that there were other variables which might have an effect on losses during similar periods. Dickins concluded his reply by suggesting that on nights of bright moonlight one alteration in course might help to throw off catseye fighters, which relied on natural and artificial illumination to identify and close in on their targets.51 Taking all this information into consideration, ultimately it was still up to Harris to weigh the pluses and minuses of launching a raid that would require mustering the entire strength of the Command and this was what he intended in the late spring of 1942. On the plus side, such a raid would show, if successful, what Bomber Command was capable of and also do much to instil confidence in both the crews and external doubters in the other Services. On the minus side, if it failed there could be dire consequences. In particular among the thousand aircraft intended for the raid would be those of the operational training units, and taking those crews out of the training pipeline, should they sustain significant losses, would have a long-term and detrimental effect on Bomber Command’s capabilities.52 Prior to target selection Dickins was called in to brief Harris. The scientist was firm that the attack should take place within Gee range and repeatedly recommended Cologne, despite Harris’s desire to attack the more distant Hamburg, which, as a port, Dickins agreed should be easy to find. ‘At the end Harris said “I still want to take on Hamburg.” “Stay within Gee coverage,” advised Dickins. “Go to Cologne.”’53 The Operation Order for the raid continued to show Hamburg as the primary target,54 with Cologne as the alternate, but Harris had actually heard and accepted Dickins’s counsel. ‘He had come to rely on Dickins’ flair for finding out before a raid what to expect from it.’ 55 Indeed, Dickins’s recommendation to attack Cologne was ‘strongly supported by Saundby, Air Commodore Harrison and Group Captain Elworthy, the three principal planners.’56 Thus the first of the ‘thousand’ raids was sent out against Cologne on 30 May. The force had been assembled some days earlier, but this was the first night on which the weather was adequate for visual identification without flares as it had been decided to use a modified Shaker at-

Concentration: February–December 1942 95

tack. The raid was a success: the ORS estimated from the available night camera evidence that six hundred of the force had attacked the target area. Losses were reasonable at 3.8 per cent and six hundred acres of the target had been destroyed.57 Yet while Harris was pleased with the results, he recognized that this was no indicator of guaranteed similar successes in the future. In fact, he believed that it might be more difficult to achieve concentration in the Ruhr, given that there were no good landmarks to use as visual crosschecks for Gee indications. On the positive side, he said, the Cologne attack had shown what could be achieved. With the force still intact he therefore launched an attack against the Ruhr two nights later.58 The second thousand raid took place on the night of 1 June; the target this time was Essen. Here the results were more typical. The visibility was between one and five miles in smog and haze, and, despite the use of the complete Shaker sequence, crews did not have much success in finding the target area. While 767 crews claimed to have been over the target, the daytime reconnaissance flights of the following week showed little evidence of this. Losses were slightly lower than Cologne, at 3.2 per cent.59 Despite this average performance, Harris was able to conclude that concentration of the bomber stream was an effective protection.60 Here again, however, one wonders if it was Harris who drew the conclusion or if it was the ORS that brought the matter to Harris’s attention. The last of the thousand raids was dispatched on 25 June against Bremen. This night the target was completely hidden by clouds and the force had to bomb through the solid undercast. A large incendiary force bombed blindly on Gee and many of the followers were able to bomb on the glow of the fires which penetrated the cloud deck. Follow-up photo reconnaissance revealed that, while not a full success, the raid had achieved good results, particularly given the conditions. Less favourable were the losses which amounted to 4.9 per cent of the force. When the losses from 91 Group, one of the training groups, were removed, the losses fell to 3.2 per cent. Losses among the heavy bombers dropped to just 2.6 per cent when Halifax losses were discounted. These numbers were significant and bear comment.61 As mentioned, the possibility of heavy losses to the training organizations, the Command’s Operational Training Units and Heavy Conversion Units (which prepared crews to fly the front-line bombers in combat conditions), had been a particular concern going into the raids. Indeed it became the subject of a special session called by Portal at the end of June. At the meeting Harris expressed the opinion that while there was

96

The Science of Bombing

room to reduce the impact on the OTUs, the value of a real mission to a crew nearing the end of their training was much greater than the usual leaflet or mine-laying sorties. Aircraft losses, he said, were ‘compensated for by the increased keenness and efficiency of the maintenance personnel, which resulted in a higher rate of serviceability [among remaining aircraft]. He was convinced that the intangible advantages of using O.T.U’s in operations far outweighed the tangible losses.’62 It was only towards the end of October that the ORS was able to conclude their analysis of the OTU participation at Cologne and in other major raids that had taken place in previous months. They had examined a range of issues which generally focused on what the OTUs had contributed to the attacks and what the losses of time and resources had done to crew production. The scientists were able to conclude that OTU crews had contributed significantly when raids were conducted in good weather, but not so much in bad weather; they therefore recommended that where the weather could not be guaranteed it was better to cancel the OTUs at the last moment rather than waste the sorties. In terms of training output it was concluded that so long as training was cancelled only on those nights that the OTUs were used and that aircraft losses were made good ‘promptly,’ there was no impact on graduate production. Harris’s thoughts were repeated: ‘In these circumstances the increased keenness will make good for the loss of flying incurred.’ They cautioned finally, however, ‘When, as at present, there is a shortage of crews in the heavy squadrons, [Heavy] Conversion Units and Flights [as opposed to OTUs] should not be used.’63 Thus, while their quantitative examination generally confirmed Harris’s intuition, the boffins had not in any sense rubber-stamped his views. The official history ends its discussion of the thousand raids by suggesting that while the attacks had shown the potential of large raids and taught valuable lessons with respect to organizing and operating such large numbers, at the same time they had shown that much more needed to be done to make good the concept of concentration, particularly in the face of ‘an alert enemy and variable weather.’64 The weather had been a known but at the same time unpredictable adversary. The unexplainably high losses on the night of 7/8 November 1941, losses which had contributed directly to Peirse’s replacement, give some indication of the havoc that the ‘met’ alone could play on the bombers. The ORS was asked to attempt to determine the cause behind the high loss rate on this night, which saw four hundred sorties dispatched in attacks against some eight targets. It was apparent that

Concentration: February–December 1942 97

those Whitley sorties sent to Berlin had experienced approximately double the expected loss rate (20 per cent versus 10); similarly 13 per cent of the Wellingtons sent to Mannheim had been lost compared to a previous average of 2.6 per cent. The scientists determined that there had been a strong west wind and icing conditions and that, on return, many aircraft from these two attacks had landed away from base due to low fuel issues. The ORS concluded that fuel shortage was to blame, but made no recommendations for corrective action.65 On the question of enemy action Harris later indicated that the enemy did not make many changes to defensive arrangements in 1942, but that they did introduce what he called ‘a new and deadly tactic.’ First observed in the month after he took over command, enemy fighters would climb steeply from beneath the bomber and open fire at close range. This was a difficult tactic to defend against since none of the bombers had adequate vision beneath the aircraft. In time crews began to use ‘corkscrewing’ and repeated banking so that they could look down sideways to detect threats.66 It appears that serious thought was not given to countermeasures until late in the year. By then it was noted that better visibility and more firepower were needed for this quadrant so that the flyers could neutralize the Luftwaffe’s adaptation to concentration. Both the Ministry and High Wycombe staffs were in search of alternative means to get past the defenders. These included low altitude penetrations which were subject to flak, as well as multiple course changes which made accurate navigation almost impossible for overstretched crews.67 The scientists kept track of losses, putting out monthly summaries. In addition, the ORS made investigations into particular raids where circumstances seemed to warrant. In an attempt to define the causes, one particular report examined the events of two costly raids against Kassel (9.8 per cent losses) and Saarbrucken/Nuremberg (11.5 per cent losses) which took place at the end of August 1942. Data was displayed clearly in tables and map plots, and a variety of sources of information, including signals intercepts, were analysed and compared. It was concluded that during the attack on Kassel enroute flak losses had been higher than usual, while it was also noted that visibility and illumination over the target had been ideal for night fighters. During the second raid it was determined that there had been a ‘record’ number of controlled fighter sorties, again with excellent weather for interceptions; as well, there were indications of fuel shortages. The overall conclusion was that ‘on both of these nights weather was ideal for fighter interceptions,’ leading on

98

The Science of Bombing

each occasion to approximately double the average loss rate.68 While the report made no recommendations it may be seen that such would not have been required. What was more important was that, based on the evidence, the operators could fairly confidently conclude that these raids had regrettably run into bad luck. Referring to these investigations in his memoirs, Harris was almost effusive about the contribution of OR section. Under brilliant young Dr. Dickens [sic], Bomber Command’s Operational Research Section’s investigations always enabled us to know exactly where we stood. In August [1942], they reported that in the previous two months between a third and two-thirds of all our losses – our total losses amounted to 5.6 per cent of all sorties – were caused by radar-assisted enemy defences, which included radar-assisted guns as well as ground-controlled fighters. They also estimated that effective countermeasures against radar transmissions would probably cut our losses by about a third, and, because bombing accuracy was seriously diminished by the strength of enemy defences in the target area, would increase the efficiency of our attacks. In fact, there was so strong a case for the immediate use of radio countermeasures that we made another application to the Air Ministry, asking them to be provided at once; we said this was a matter of the greatest importance and urgency.69

In this Harris was referring to the use of Window, a radar-jamming technology which used metallized strips of paper dropped in great quantity by the attackers to fog the German radar screens, and which, while not authorized for use at this time, was to come into service in 1943. At the end of the year Dickins’s staff put together report B 120 ‘On the Routeing of Night Bombers through areas Defended by Fighters,’70 which attempted to find a compromise that would minimize losses from ground-controlled interceptors and the free-lance or catseye defenders. The problem was that minimum concentration in the stream favoured GCI fighters since the radar controller could ‘see’ individual bombers fairly readily, while, conversely, high concentrations make interception easier for the catseye fighters. It was the scientists’ job to find the optimal concentration that would minimize both threats without compromising raid effectiveness. They used the evidence taken from two raids described above which had been analysed in report S 64 ‘Investigation into Bomber Losses Sustained on the Nights of 27/28 and 28/29 August 1942.’ After looking at all the variables, including the risk of loss to flak, the scientists now concluded that

Concentration: February–December 1942 99 under conditions of good visibility, even when effective countermeasures to enemy G.C.I. are not available, it pays to order the routing of bombers so that they will fly in a band of width of 40 to 50 miles. If fairly effective countermeasures are available it pays to do this even when visibility conditions are not so good. In practice the routeing over a wide front would be achieved by routeing the aircraft over two or three lines between 5 and 10 miles apart. It might be an advantage for the lines to cross in the neighbourhood of gun defended areas.71

It was one thing to report problems and another to make recommendations on resolving losses, but by now the scientists were comfortable enough within the Headquarters to do just that. Earlier, in August, Dickins had already reported the steady rise in the loss rate. While devices like Window and Monica, which warned the crew if another aircraft was approaching, were soon to be available, he still believed that training was an important tool in preparing crews for what they would encounter over hostile territory. He advocated the establishment of small fighter training flights like the one already in operation with the AFDU; in his proposal each Group would have one such ‘circus’ of two or three fighters which could be used to give all new crews an orientation to fighter attacks. This training, he was convinced, would lower the loss rate. AOT staff were able to confirm that they were already in the process of establishing more fighter detachments. There was also work in progress to open night vision schools.72 The boffins and the operators were, it is clear, in full agreement. Harris later argued that his first year in command was spent fixing some of the problems which had been identified by the end of 1941. ‘I was unable to begin any real bomber offensive for a whole year after I took over the Command for lack of aircraft, proper equipment and trained crews, and also because I was compelled to use what force I had for many other purposes besides the strategic bombing of Germany.’73 As has been seen from the preceding discussion, he might well have added to his list the lack of effective tactics. His problems had included in a very real way the continuing chronic inability of the bombers to find and hit targets in the dark and the difficulty of protecting the force from enemy action. At the beginning of the year the scientists had been instrumental in developing the tactic which all hoped would solve, at least to some extent, these problems. As seen in the previous chapter, the ORS had proposed and then developed the Shaker attack protocol, which would give the majority of any raid a relatively clear look at the

100 The Science of Bombing

target area and then an accurate aiming point to bomb. Now they had comprehensively investigated the question of raid concentration showing categorically that high densities reduced losses to enemy action while also contributing to more effective and efficient bombing. Both of these proposals were based on their early work in bringing Gee effectively into use across the command. Finally they had been early supporters of the pathfinder process and had been able to study and report on the PFF’s successes and limitations from the first use of the force. In all of these undertakings the scientists had collaborated with the Bomber Command staffs and with the senior leaders as well. In many cases the scientists’ advice had been sought by Harris and Saundby and their research had been lauded by seniors both in the command and at the Air Ministry. There is little question that the boffins, too, had learned along the way, but now they and the Command as a whole would be called upon to put that learning into practice as the campaign entered a new phase.

7 Putting Concepts to the Test: January–February 1943

While much had been attempted in 1942, it had not been a year of particular success. True, the Command had begun the process of converting to the modern and vastly more capable heavy bombers. As well, Gee had been introduced with significant impact and the Pathfinders, while limited by equipment inadequacies, had made a modest contribution. But these improvements were still not enough to make bombing operations sufficiently efficient or effective when compared to the efforts, both human and materiel, that were being invested. According to the official historians, the outlook for the coming year (actually the fifteen months between January 1943 and March 1944) was ‘less bleak.’ More squadrons were being established and an entire new group – No 6 (RCAF) Group – was to be activated. The shift to new fourengine heavies was well along, and the Lancaster in particular was proving itself beyond all expectations. Other systems were now coming into service: Oboe, the precision navigation aid, had seen first operational use on 20 December 1942 and H2S, the on-board ground-mapping radar, entered service within the first weeks of the new year. The first effective target indicators (TIs) were used on 16 January and a fourth device, ‘Window,’ was available to jam enemy radar, although it would be held in reserve until the summer.1 Taken together, the growth in quantity and quality of the bombers, coupled with the new equipment and enhanced tactical options, convinced Harris and his staff that the campaign should start to see larger returns, returns which would result from three major series of engagements: the Battles of the Ruhr (March–July), Hamburg (July–November), and Berlin (November 1943–March 1944). The successes and failures of these actions would play out in the skies over the continent, but at the same time the scientists, working in the back halls

102 The Science of Bombing

of High Wycombe, would also be hard at work ensuring that the full potential of these new aircraft and technical aids was realized. This chapter and the three that follow present an examination of the Section’s work through a review of their reports, and an assessment of the impact of the advice provided by the scientists. While most histories treat the Ruhr, Hamburg, and Berlin actions as part of a single phase in Bomber Command’s war, there were important differences between the circumstances of the attacks on the Ruhr and those of the attacks on Hamburg and Berlin. Up to the end of the Battle of the Ruhr in July, the Command worked largely within range of both Gee and Oboe, and distances over enemy territory were relatively short. The assault on Hamburg, beginning at the end of July, took the bombers deep into German territory, where they would continue to operate in the winter months during the Battle of Berlin. Although in late July the Command deployed Window, which was intended to mask the bomber stream from German radars, it could not provide a perfect or constant defence. The hours of passage over enemy territory, therefore, meant that the crews faced additional risk from fighters and flak. More, these deep attacks, which reached well beyond the range of Gee and Oboe, were solely dependent on the new H2S to provide navigation and target-finding assistance. Before these three battles would begin there was another shorter phase of re-equipping and learning that had to take place. First, during December 1942 and January 1943 experimental raids were conducted to develop the best tactics for Oboe. A new and exceptional light bomber, the Mosquito, with its high operating altitude, made a number of small raids and these were followed by somewhat larger attacks using Mosquitoes and Lancasters that also introduced the technique of ‘skymarking,’ the use of TI flares dropped by the Pathfinders above undercast cloud cover. Slowly, evidence was accumulated that these techniques were working; for example, previously unscathed facilities in Essen were now showing the effects of bomb strikes. However, the ORS continued to express reservations, largely because there was insufficient evidence of the accuracy of bombing. Worse, for the moment the scientists had no way of judging the accuracy of the skymarkers or the bombs dropped on them as the former literally drifted with the wind while the bombs fell unobserved.2 Other experiments were easier to assess. The first use of groundmarking with target indicators took place in two raids against Berlin on 16/17 and 17/18 January 1943. Subsequent photo evidence confirmed new and relatively concentrated damage, but the majority of the raiders had scattered their bombs through the suburbs. From these raids it was

Concepts: January–February 1943 103

fairly evident that using TIs without some means of getting the marking aircraft over the target was of little advantage.3 H2S was used operationally for the first time on 30 January. Pathfinders leading the attack on the port of Hamburg that night reported that their on-board radar screens, called Plan Position Indicators (PPIs), had presented pretty much what they had expected to see in terms of ground and water details. This information had allowed them to mark the target effectively, as seemed to be confirmed by the testimony of the main force crews that followed. Nevertheless, little damage could be found during a subsequent reconnaissance.4 Five additional attacks using a combination of Oboe, H2S and TIs were attempted between 2 and 9 February with similar lack of demonstrable results, but the attack on Wilhelmshaven on 11 February did bear fruit. In this case skymarking based on H2S led to the ‘spectacular’ destruction of an ammunition depot.5 Taken together these raids permitted the development of tactics for various weather and cloud conditions. (These techniques are depicted in appendix 8.) Skymarking was to be used if the cloud cover was 8/10ths or more. In conditions of between 5 and 8/10ths the pathfinders were instructed to drop their TIs based on H2S. If less than 5/10ths cloud was encountered the pathfinders were to drop flares using H2S and then visually aim their TIs. The latter tactic, the visual ‘Newhaven’ technique would later result ‘in the most remarkable successes.’ Conversely, it was found that blind marking and bombing was often more reliable and accurate than the visual techniques as crews could not easily be duped either by decoy fires or by misjudgment if they questioned the positioning of the TIs.6 The scientists had been hard at work analysing all of the raids and the relative success of each. Interim raid reports were put out as soon as possible after attacks, particularly when things had not gone well. Such was the case following attacks against Wilhelmshaven on 18/19 and 19/20 January. In the first case the scientists were able to reconstruct the raid minute by minute, comparing the comments of the PFF crews. From the anomalies in the flyers’ reports the boffins were able to conclude that the Germans had used decoys and a smoke screen to throw off the attack, which, but for having missed the aiming point, was generally well executed. This said, the ORS did determine that the first markers fell about 1½–2 miles short of the target and that undershooting, the main force crews dropping their bombs progressively more short of the target, increased during the attack. While Wilhelmshaven was a coastal target, the raid had approached over land, leading the ORS to suggest that in fu-

104 The Science of Bombing

ture coastal raids should approach from the sea so that markers dropped short of the aiming point would be extinguished in the sea and thus not contribute to any undershooting of the main force.7 Particulars of the following night’s attack were equally detailed and equally disappointing. Low cloud, a smoke screen, low tide, and newly constructed buildings all contributed to false H2S pictures, but it was also suspected that the bulk of the PFF marker crews might have used the H2S for navigation only, bombing visually despite direction to the contrary. The primary markers had achieved good concentration and the bulk of the raid bombed on these indicators, which unfortunately were focused about four miles north of the aiming point. Ironically, concentrating the H2S markers at the beginning of the raid had kept the bombing tidy, but had done nothing for finding the target. To combat the apparent confusion due to changes in ground features the scientists recommended that H2S (also called Y) aircraft be provided the most up-to-date information possible on changes in the target area which might affect radar returns.8 It was not until the first week of March that this analysis became available. Saundby, appointed to the newly created position of Deputy Commander, sent Harris the OR reports on Wilhelmshaven on 6 March, highlighting the key issues for the CinC. It seemed plausible to Harris’s deputy that the mistake on the first night could be attributed to visual errors caused by the smoke screen. The second night’s problems might possibly have been the result of extensive use of metal materials in new construction in the area. Given this particular scenario, Saundby now ordered ‘that the areas around possible H.2.S. targets are to be examined and any new built-up areas marked on target maps and sent to the pathfinder force.’ He also directed the SASO to ensure that the PFF received copies of the reports along with his direction.9 In hindsight it might have helped had Saundby offered sterner criticism of the PFF’s failure to follow specific procedure, but the scientists had unquestionably drawn the problems to the attention of their seniors. Harris remained concerned that, even with all the enhancements to capability, it was still frequently impossible to get bombs on target on many nights where the weather was uncooperative. The problem was in part resolved by reassigning crew functions to ensure better integration of Gee and H2S data.10 Simple as this sounded, the research behind this decision was much more extensive. While trials of the new equipment and tactics continued, the boffins were heavily engaged in the analysis of past practice. They had long been concerned about the visual identification of navigation and target landmarks and in January 1943 they

Concepts: January–February 1943 105

were ready to table an extensive analysis of the issues. Dickins passed the draft to members of the training staff and the CNavO for comments, suggesting that the document receive wide distribution, but that first it be reviewed by the Headquarters and Groups so that a fairly definitive statement could go to the Ministry and other agencies.11 The report looked at a number of interrelated factors which affected the visual identification of targets, including the appearance of the ground, enemy opposition, aircraft design, the bomb aimer, and training. While the scientists recommended additional extensive investigation into all areas, they had already assembled several pages of data, observations, and conclusions. They began by saying that, while there had been improvements in the previous year, ‘nevertheless, a proportion of the force still fails to find the target, and on some occasions even expert crews are unsuccessful. The elimination of such errors would, of course, lead to an increase in the effectiveness of the bombing offensive.’12 Weather, they said, was the single most important factor in raid failure, but, as little could be done about poor conditions, it was their intent to focus on those nights and conditions of marginal weather – nights with low illumination and moderate haze – where improved performance by the force could have a real impact. They called for studies of object recognition at the limit of visibility and focused their suggestions particularly on the bomb aimer, looking at both selection and training of individuals as well as the working conditions of the bomb aimer and his ‘office’ in the nose of the aircraft. In the latter case they called, for example, for an ability to stow all equipment and for the use of black paint rather than the green then in use, which was hard on the eyes. As far as PFF bomb aimers were concerned, they too needed extensive training to maximize the experience that they brought to the role. As a general conclusion the scientists stated that no one problem could be considered the fundamental weakness; rather, they said: ‘It is the cumulative effect of a number of factors, each by itself appearing rather insignificant, which has contributed to the failures experienced.’ They called for ‘a steady and parallel attack on all’ and recommended the formation of a standing committee to conduct a regular review of all issues.13 Signalling his agreement, the Air Officer Training, A/Cdre Capel, sent the draft report on to Saundby, adding extensive comments on the early establishment of a series of twenty night-vision schools that was already being planned. He ended with an endorsement: ‘I expect the C.in.C. will like to see [the report] to know what action is being taken to get ahead with this training.’ Saundby too was in agreement and passed

106 The Science of Bombing

the document to the CinC. Harris asked him to immediately set right the technical deficiencies that the boffins had identified in the bomb aimer’s compartment: paint colour, lighting, and steadiness of the bombsight mounting. He also agreed with the formation of the committee that Dickins had recommended, but he expressed reservations: ‘My own view is that in the general run of crews “look see” target identification at night is impracticable on most occasions. We should concentrate on the Pathfinders & on aids for them first with a view to lighting up the target by flares, markers and fires.’14 Little time was lost in assembling the participants for the committee, but the list of attendees was somewhat more limited than what might have been envisaged by the scientists. In the end only Saundby, A/V/M R.D. Oxland, the new SASO and former AOC of 1 Group, G/C Elworthy (representing the Air Staff), and Bennett were joined by three scientists: Dickins, Mr G.A. Roberts, and Mr G.W.H. Stevens, the ORS’s senior researcher for problems of visual identification.15 Attendance had been limited, for, as Saundby stated, there really was little to be gained from trying to implement major changes to visual identification methods among the main force crews. In the first place, the majority of raids were conducted in poor or fair conditions, making visual identification of limited value, and second, once a target was lit up with flares, markers, and fires there was little ground detail that the average main force crew could even hope to see. The D/CinC, as Harris had suggested, concentrated on improving the results of PFF crews, and hence Bennett’s participation. This strategy adopted, the balance of the meeting focused on setting right three problems which captured the essence of the report’s recommendations: redesigning the nose section of bomber aircraft; defining and acquiring aids such as night binoculars and flares; and setting up additional training. Much could be done to optimize the nose of the Lancaster, as this was the PFF’s principal aircraft. The pathfinders were tasked with deciding what modifications were needed to meet the recommendations put forward by the scientists. Where the problem was a major one requiring national-level modifications, the Headquarters would step in. With regard to crew effectiveness and cooperation Bennett felt that these issues could only be properly worked out on actual operations; there was no easy solution, he felt, because of the unique personality of each crew. Saundby, while not disagreeing, believed that much could be done by instituting practice bombing from operational heights; he had already asked the AOT to find suitable ranges. Finally it was agreed that practice

Concepts: January–February 1943 107

in estimating distances was important and the ORS was asked to assist the pathfinder bomb aimers in finding the best way to use a new device that was being investigated for this purpose. Just two weeks later Dickins received a progress report; he examined each item carefully, asking for further updates and directing the boffins to ensure maximum coordination of effort with outside agencies. Many of the changes proposed by the ORS were either under consideration or already in progress. Work on a new ‘ideal’ nose, initially for the Lancaster, was on the drawing board; it would have better visibility and stowage, and proper painting and lighting, not to mention heating. Training and simulation devices designed to let bomb aimers have a better idea of what they should or could be seeing from altitude were also under development. Not only was it clear from this update that the scientists had zeroed in on the key problems facing the bomb aimers, but, given that the update went directly to Dickins, it was also evident that the scientists had been given the lead in effecting these improvements. 16 The associated problem of navigation was similarly near the forefront of the Command’s concerns. Indeed, there was, as always, little point in worrying about bomb aiming if the target area could not be found. At a navigation conference held some weeks earlier in January, the chair, G/C H.R. Graham, the CNavO, started off with a scathing rebuke: ‘It should be plain to everyone that the failures of Bomber Command were primarily navigational, and there was much scope for improvement in [navigation equipment, layout of equipment in the aircraft,] and navigation technique. Our aim is to be able to navigate an aircraft so that the timing will be precise, and errors in position will not exceed two miles.’ Graham deplored ‘the scant attention paid by the majority to navigation.’ He then went on to lay this aspect of the problem at the feet of the Group and Station navigation officers, holding up the PFF as an example of what should be possible.17 It is intriguing to note that of the several navigation conferences spanning three and a half years this is the only one at which the ORS staff were in attendance. It seems odd that, when the central topic was always the effectiveness of navigation as a contributing factor to effective bombing or force concentration the ORS would not have been present in every case. There is little doubt, however, that those who needed to talk to the scientists about navigation problems had ample occasion to do so on a day-to-day basis. As mentioned, weather was the biggest challenge to accurate navigation and one of the most vexing weather-related problems facing planners and flyers was the question of wind direction and speed. Winds

108 The Science of Bombing

above the surface, referred to as ‘winds aloft,’ vary in direction and velocity with altitude and location. Not knowing the winds made it difficult for planners, dealing with aircraft coming from different locations and with different cruising speeds and altitudes, to achieve concentration both enroute and over the target. In January 1943 the Navigation staff put out proposals for modifying existing practices. Among them, crews were to use a common forecast wind and stick to their predesignated speeds regardless of the actual winds encountered. Thus, bombing could occur as planned regardless of whether the winds were stronger or lighter than anticipated. It was recognized that, as part of this approach, Zero hour (the moment the bombing began) would not be a particular time, but would become whatever time the raiders actually arrived over the target. The proposal brought a wave of responses varying from agreement to detailed rebuttal, some comments even suggesting that the wind was not always the reason behind poor concentration.18 While the OR section had not been involved in these discussions, in December 1942 they were asked by the OR Centre to provide data on the effect of weather on operations to the rapidly expanding US Army Air Force (USAAF). The staff assembled a short analysis, which Dickins put before Saundby at the beginning of the new year. Revisions went back and forth between them and Harris for several days before the final document was ready for Dickins’s signature.19 In it the scientist provided a number of fundamental observations in addition to specific statistics. This Command suits its operations to the prevailing weather conditions, and an analysis of past operations, therefore, does not necessarily indicate the relative importance of blind landing and blind bombing equipments. For example, we have to lay a considerable quantity of mines each month, and these operations can successfully be carried out in much more cloudy weather conditions than bombing. Further, when we have been operating intensively, it is necessary to rest crews and overhaul aircraft, and advantage is taken of weather conditions to do this. Also, we have, at the moment, a large variety of targets spread over a wide geographical area available to us. If, however, we had to give priority to one particular target, we should find that the weather conditions over the particular target in question were usually a much more serious restriction than the weather conditions at home bases. Bearing in mind the above points, we have endeavoured to produce the answers to the questions which you ask by a study of the night operations during the period June–October 1942. Our conclusions are as follows: (I) No operations have been carried out owing to weather conditions at

Concepts: January–February 1943 109 base on 27% of occasions, which includes 6% in which the weather at the target was also unsuitable. (II) On a further 5% of nights, not sufficiently good weather conditions for any operation to be undertaken. These figures relate to any weather conditions preventing takeoff, landing, or target identification. As regards the percentage aircraft lost due to fog conditions over British Aerodromes on return from operations, the proportion of sorties which resulted in crashes due to the weather during the period June–October 1942, was about 0.2%, but not all these aircraft were written off. The above data refer to night operations, similar data for day operations are almost impossible to obtain since the possible types of operations are so varied.20

What is particularly significant about the document is that, while including a range of statistical data, it also describes a variety of policies employed in planning and conducting operations. That these should go out of the Command – and to a foreign military no less – over the scientist’s signature is clear evidence of Harris’s confidence in and regard for Dickins. While much attention had been maintained on improving bombing, the scientists and the senior leadership had been equally focused on the question of casualties. As always the discussion could be split into two interrelated topics: defence against attack and the particularities of losses. Prepared in the last weeks of 1942, ORS Report G 66 ‘Review of Sorties, Losses and Interceptions of Bomber Aircraft in Night Operations during the Period 1st Aug., 1941–31st October, 1942’ offered a detailed review of losses due to fighters.21 The scientists noted that the increased use of heavy bombers over the preceding six months now provided sufficient data to conduct analyses and make comparisons to earlier aircraft types. They also said that special sorties like minelaying were not indicative of a typical mission with respect to the fighter threat and so had not been included. A long list of significant conclusions emerged. First, the scientists indicated that, while total losses had previously been dropping, there was now a definite increase and that this increase had been paralleled by an increase in night fighter activity. In analysing the losses it was seen that while North Sea coastal attacks had experienced the same loss rate as in the past, those against the Baltic and southwestern Germany had experienced higher losses where, it was noted, fighters were particularly active. All told, it was felt that losses to fighters had doubled. Analysis had also been done by type of aircraft and it was found that there was a relation-

110 The Science of Bombing

ship between losses and the ‘liability to interception.’ ‘The comparative immunity of the Lancaster is very interesting. It is probably due to good performance and manoeuvrability.’ Lancaster interceptions were 5.1 per cent with 3.3 per cent losses; Halifax figures were 10.7 and 5.6 per cent respectively. For both types 33 per cent of these interceptions had developed into attacks, although generally these figures as well as actual losses were lower in winter months. It was also noted that most attacks occurred during good moonlight. These trends did not tend to be reflected over the Ruhr and it was suspected that the concentration of defences in that area accounted for the anomaly. The scientists further observed that 75 per cent of fighter attacks developed from directly astern or the stern quarters and that the majority occurred during the journey home. Finally it was noted that there was little use of aircraft lights or searchlights for attacks. All of this was excellent information for those directing and planning the campaign and for those flying the bombers. Harris and his operations staff were able to see a need for enhanced countermeasures against fighters (and the fighter control system) and could not avoid recognizing the relative limitations of the Halifax.22 Tactical Committee Paper TC 31 was published in February and it continued to recommend concentration as the best defence against AA searchlights and guns.23 The new report also recommended the use of immediate ‘violent evasive action’ should an aircraft become ‘coned’ (caught in searchlights). Significantly, however, one of the Operations staff took some exception to suggestions of evasive action, saying that experience showed the best way to escape searchlights was to proceed quickly from the area, rather than causing delay by evasive manoeuvring.24 This notion would not escape the scientists. The problem of dealing with fighters was a vexing one and 5 Group did not always like the information and decisions coming from the Command. In early 1943 the Group’s SASO voiced his unhappiness to his AOC, A/V/M W.A. Coryton. ‘I feel it is incorrect to say that casualties due to “Catseye” fighters are independent of concentration. It seems to me that a “Catseye” fighter has a better chance of picking up a bomber when the density of bombers per unit of space is greatest, i.e. if the bombers are concentrated in height. Conversely if the bombers are dispersed in height, then there is less risk of an interception by a “Catseye” fighter.’25 Coryton asked his ORS officer, Mr J. Curry, to consider the matter and the Group’s scientist provided a detailed response.26 First, said the boffin, concentration was essential. But at the same time there was, in practice, dispersal in height when different Groups flying differ-

Concepts: January–February 1943 111

ent aircraft types participated in the same raid. It was desirable to have this dispersion, as it did actually cause problems for the GCI controllers attempting to coordinate intercepts, and dispersion usually gave the bombers some height to trade for speed when crossing fighter belts. But Curry cautioned that too much dispersion and varying airspeeds were bad for the essential concentration in time. From these comments the AOC and SASO were able to work out the best options for height dispersal given either GCI catseye and pure catseye threats at 3000 and 6000 feet respectively. These figures could then be applied to 5 Group independent missions that were to grow in number in the coming months.27 Moreover it seemed that the science, once explained, had convincingly won over the otherwise sceptical SASO. It was ironic that Curry, having established the credibility of the ORS’s recommendations, would find himself in a similar debate some months later. One of the planned defences against fighters was Monica, an on-board radar device designed to warn of approaching aircraft, but its development was slow and frustrating. After a protracted gestation, hopes of getting the system into use by the end of 1942 had faded. In January 1943 Dickins presented the boffins’ review of the situation. While it was acknowledged that Monica would save aircraft, it was also accepted that the ‘life’ of Monica, before the Germans introduced countermeasures, was only likely to be some three to six months. Since no replacement was to be expected until the late summer, Dickins wrote: I recommend that we should delay the introduction of ‘Monica’ until the end of February in order that there may be a reasonable chance that it is available for our protection over the summer when danger from fighters is most serious. This delay of 6 weeks will cost us 9 aircraft now but prolonging the ultimate life of the device, will result in a possible saving of 100 aircraft later. [Dickins added a further comment to the file later the same day.] Since dictating the above, the CinC sent for me and we discussed the recommendation in [ORS B 119]. He agreed that the question of the date of introduction of Monica be reviewed on February 1st.28

Subsequent reviews of the delivery of the equipment at the end of January and then February were depressing. By the end of February less than one hundred aircraft had been wired for the modification and only two hundred sets were available for installation.29 Finally at the end of March, with the supply of sets still not much improved, Dickins again

112 The Science of Bombing

provided his advice.30 Given the circumstances, the current low loss rate, the fact that it appeared that Monica and Oboe would interfere with each other, and that the gap between Monica IA and II (an already planned replacement) should be minimized, the scientists proposed an additional delay of one month or ‘until such a time that our losses begin to show an increase and the number of aircraft fitted is such that an appreciable saving would be effected.’ All of this, it was concluded, would need to be tempered against the introduction of Window, the much-anticipated radar-jamming device. Should the latter be approved for release in the coming weeks, then it would be appropriate to begin using Monica immediately as the two had similar purposes. Saundby saw the logic of the proposal and accepted it the same day.31 For no apparent reason almost two months were to pass before Saundby informed Harris of the newest concept for bringing Monica, as well as an updated Oboe, into service. After deliberations with Dickins and the CSO it was recommended that Monica’s use would begin on 16 June. Harris agreed.32 Monica, Window, and concentration were all intended to reduce losses, the numbers and causes of which the boffins continued to track with care and great concern. In January 1943 Dickins had passed Saundby a report dealing with casualties due to flak which had been prepared by the Operational Research Group. The file showed that of 15,167 sorties studied, 1557 had come home with damage and from these 55 aircrew casualties (13 serious or fatal) had been reported. Using the average number of flak holes in a heavy bomber it was concluded that the chance of flak first making a hole in a bomber, and then wounding a crew member, was 0.0011 per cent. Despite this apparently small risk to life and limb, when Saundby saw the data he was less than convinced that it should be widely circulated, as the ORG had requested. ‘I am all against circulating to air crews, for instance, statistics about the vulnerability of their bodies.’33 Perhaps the D/CinC would be less discouraged by the data contained in ORS report S 77, which the scientists completed later that month. Of 20,223 sorties recorded, the equivalent of 124,390 person sorties, there had been only 203 serious casualties or fatalities from flak or fighters. Further, of these, there had been only four cases of pilots being injured. Apparently, however, Saundby’s concerns remained and he replied: ‘We must restrict the circulation of this.’34 Dickins proposed a reduced distribution and the addition of a comment on the report’s header: ‘This report has been prepared for the information of those concerned with protection of personnel in bomber aircraft, and it is of the greatest importance that the information contained in it should only be allowed to

Concepts: January–February 1943 113

come to the notice of those officers directly concerned in this work.’35 Why this reluctance? It would seem plausible that Saundby was uncomfortable with giving the crews any reason to be more doubtful than they must already have been about their futures. One can conjecture that his reasoning went something like this: It is bad enough to think about the loss rate without having to consider the specifics of death or injury from flak or fighter. Yet surely he realized that the crews were only too aware of comrades who had been killed or wounded by enemy fire. A final thought on the systemic dishonesty of statistics seems appropriate. One wonders why the scientists seemingly ignored the fact that missing crews must have suffered casualties in the early moments of their demise. Would not a crew member only injured by flak, but still doomed aboard a failing aircraft, be a casualty to flak? And what if that crew member was a pilot? It was as if the scientists had wiped these aircraft and crews off the slate before performing their calculations. The boffins were also involved in wide-ranging discussions with other scientists on a variety of issues related to losses. The need for effective exhaust dampening, to hide both the flames and glow from hot engine exhausts, had already been under review for much of 1942 when Dickins approached Saundby in early February 1943 with a proposal. There were so many estimates on which exhaust gases could and could not be seen by enemy fighters that Dickins felt it necessary for the Command to conduct its own trials to come up with some precise and valid numbers. Only then, he thought, would the Command be able to put some effective pressure on the Ministry of Aircraft Production for action. Saundby agreed and directed that the trial be undertaken.36 The results were back before the end of March. For the Lancaster, it had been determined that, when at cruising speed, the aircraft silhouette was visible at a greater distance than the exhausts. The same results were achieved even at high speed (i.e. with the engines running at higher temperatures), except when operating over 12,000 feet when the exhausts were seen first. Given these findings, it had been concluded that the existing Lancaster exhaust flame shrouds were effective. The results for the Halifax were less encouraging. Without shrouds to mask the exhausts and flames, when observed from below on a starlit night, the hot gases were visible at two to three times the distance that the silhouette could be made out. A repeat of this trial had been ordered once shrouds had been installed, but the results were not yet available. More alarming, it was noted that the exhaust could be seen at ranges greater than the warning that would be given by Monica once that device came into service. In these circum-

114 The Science of Bombing

stances Dickins felt that finding effective exhaust shrouds for the Halifax was essential. On this last issue the Air Staff did not agree. W/C Tait, Ops 1(e), believed that Monica would offer adequate warning whether shrouds were used or not. Further, he pointed out, Halifaxes were currently flying without shrouds and with relatively low losses, a point he made again to the SASO in a subsequent minute. Inferring that the Halifax needed to get the best speed out of its engines – something the Lancaster had proven – he summed up: ‘Performance is the night bombers [sic] best defence.’37 While there might be debate over the most effective use of shrouds, and of the new visibility data, it was now much clearer to the staff at High Wycombe, thanks to the initiative of the scientists, that once again the Halifax was the poorer of the two front-line heavies. February saw another collaborative investigation, this time with the ORC. Scientists there were working on alternatives to hydraulic-powered turrets and needed data on the relative robustness of electrical or electro-hydraulic alternatives. Within a week Dickins’s scientists were able to provide a preliminary response. They had reports of several occurrences of hydraulic turrets being put out of action, but ‘while damage forms have reported several cases of damage to electrical wiring in Halifax II’s, in no case was any turret stated to be unserviceable.’ This did not mean that electrical systems could not on occasion be rendered unworkable. ‘All that is it possible to say is that such an occurrence is rare compared with hydraulically operated turrets.’38 Similar requests followed. The first was for data related to engine armour. Could the ORS provide data on the usefulness of the armour plating used to protect engines as there was some interest in deleting any weight that could be saved from heavy bombers?39 The ORS response once again pointed to the fact that there were only partial records, but also that engine armour was located in hard-to-see places and that ‘many instances of slight impressions from missiles may have been missed.’ Some damage had been reported, however, and the ORS recommended leaving the armour in place. Alternatively the boffins did suggest removing the ‘main armour bulkhead’ inside the fuselage for which no reports of damage had been registered. This was a plate of considerable weight which, according to crew reports, was ‘almost invariably left open during flight and when an attack develops, everyone is usually too busy to close it.’40 Based on the scientists’ work during the preceding weeks it might have been argued that the crews were too stretched to be effective much of the time. Theirs was not an easy task in any event; the new devices and as-

Concepts: January–February 1943 115

sociated techniques to which the boffins had contributed had started to offer somewhat better results, but clearly not every attack went right and not every crew could be deemed effective, much less expert. Be these things as they might, the Command was still on the eve of three major battles which would put men, aircraft, scientists and commanders to the test.

8 New Concepts, Old Targets – The Ruhr: March–June 1943

As the scientists and the staff continued with their investigations, the Command now prepared to refocus on its primary targets in Germany. It had had the preceding weeks to become acquainted with the equipment and tactics which were, hopefully, to allow for significant improvements in attacks on Germany’s production centres and elsewhere. The Battle of the Ruhr, waged between March and July 1943, saw the majority of effort sent against Germany’s industrial heart, although a number of raids against other German centres, including Berlin, and some Italian and French targets also took place. The majority of the attacks were of two main categories: main force raids of usually more than three hundred aircraft, and small Mosquito feints and demonstration attacks (ruses) of anywhere from one to perhaps a dozen bombers. Strikingly, the Mosquitoes conducted their operations with a loss rate of only about 1 per cent, while the main force losses during the forty-three raids amounted to 4.7 per cent.1 The first attack of the battle, on 5 March, set the standard for what was reasonably to be expected. It would be the first to use the new technologies and tactics against Essen. The target, as usual, was cloaked in smog and, but for the use of Oboe, the attack, by 442 aircraft, would have been doomed to the failures of the past. On this night accurate Oboe marking and subsequent backing up concentrated the raid so that approximately 150 aircraft could be reliably plotted within a threemile ring around the aiming point. Damage was later assessed as 160 acres ‘laid waste’ with another 450 acres suffering significant damage. The attack was not without cost: 14 aircraft failed to return and another 38 were damaged.2 Harris was very pleased with the performance of the Command.

New Concepts: March–June 1943 117 It was not only that a hitherto invulnerable target had for the first time been seriously damaged, but also that there was no reason why the success should not be indefinitely repeated in attacks on any other target within Oboe range. This had never been the case before, every previous success had been dependent on the caprice of the weather and usually had been won by seizing some opportunity that might never reoccur. But now the whole Ruhr was evidently at our mercy.3

Just what damage could be done by such a raid had been the subject of a broadly conducted review of bomb effects during the winter months. In February and March the boffins had been at work with other research organizations in establishing the relative effectiveness of different incendiary bombs. A report from the Central Interpretation Unit concluded that the 30 lb variant was more effective than the 4 lb incendiary, but the boffins were not convinced by the evidence and had therefore done their own analysis, which led them to a different conclusion. They felt there was no valid reason for believing either bomb was superior to the other and passed their results to Saundby. ‘Am I justified,’ he asked, ‘in deducing from [the report] that there is in fact little to choose between the 4 lb and the 30 lb bomb for effectiveness per lb? If so, this is worth knowing.’ Dickins was reluctant to make a definitive statement in support of either munition but did say: ‘if anything positive emerges at all, it favours the 4 lb bomb and in view of the fact that a small bomb container [SBC] carries a 50% greater weight of 4 lb bombs than the 30 lb bombs, there is little doubt that the former is preferred.’4 In the months that followed there was no change in the ORS opinion. In a letter to the Operational Research Centre in October, Dickins indicated that the 4 lb bomb generally outperformed its larger counterpart in terms of material damage caused, while admittedly the larger bomb had a ‘higher morale value.’ This said, ‘the main reason for [the 30 lb bomb’s] continued use would appear to be the restriction of supply of the smaller bomb.’5 This view was based on extensive work that had been done both during and after the Battle of the Ruhr. At the end of June the scientists produced a paper on the relative efficiency of bombs and bomb loads.6 They hoped to show what loads were most efficient and effective regardless of total weight carried by any one type of aircraft. Taking data from a variety of sources they were able to prove that, acre for acre, incendiary bombs were noticeably more efficient than high explosives. The 8000 lb ‘High Capacity’ (HC) bomb could destroy 1.68 acres per ton; the 4000 lb HC and the 4 and 30 lb incendiaries 3.2; least effective was the 1000 lb Gen-

118 The Science of Bombing

eral Purpose (GP) bomb at 0.56 acres per ton. 7 The effectiveness of the incendiaries could be best exploited based on three conditions: the target must be relatively intact (i.e. still possessing considerable combustible material); the bombing must be concentrated; and the overall bomb load must contain sufficient high explosives (HE) to be a threat to fire services and anti-aircraft gunners. When these factors were compared with the carrying capacities of the heavies it was found that for the Lancaster the best load was 1x 4000 lb HC + 17 1/3 SBCs of incendiaries. The weight of this load was 10,000 lbs and the corresponding damage factor was 9.98 acres. This was less bomb weight than a load of 6 3 2000 lb bombs which the aircraft could carry, but that load had a damage factor of just 6.25 acres. The difference in the Halifax was not so clear cut. The best load was 15 SBCs (5490 lbs and 6.30 acres), but a load of 3 3 1000 lb Medium Capacity bombs + 12 SBCs of incendiaries was not significantly less efficient (7449 lbs and 6.12 acres). Stirlings were seen as incendiary carriers par excellence. Overall, while each raid would have certain variables, such as target characteristics and force composition, it seemed to the scientists that the previously adopted policy of using 70 per cent incendiaries was appropriate and effective. This could all change, they noted, as more and more targets had been peviously damaged by fire, but for the moment they recommended staying the course.8 The report led to some debate. Based on personal experience, the Chief Armament Officer (CArmO), G/C C. Crawford, wanted to see more small delayed-fuse HE bombs, as in his view these had an effect on morale and on the return to normal functioning of a city long after a raid had ended. Looking at those comments G/C Ops said: ‘I am not convinced by the conclusion [of CArmO] and consider it dangerous to draw conclusions from personal experience.’ ‘I consider [the ORS] paper … sound and can think of no other points which should be included.’ Having thus checked with the operators, Dickins sent the file forward to Saundby with a short statement: ‘The investigation leads to the conclusion that the bomb loads used in present attacks are just about right.’9 The second raid of the battle was dispatched against Nuremberg on 8 March. Beyond Oboe range, the attack depended on five H2S-equipped pathfinders to light up the target area with flares and then mark the target visually; these were then to be backed up throughout the raid. The op had been planned with due consideration of the limits of H2S, but in the event much of the marking was wide of the aiming point despite clear skies and only moderate haze. The main force concentrated on the TIs and this led many of them away from the target.10

New Concepts: March–June 1943 119

By comparing these two raids it was possible to see the difference in effectiveness between Oboe and H2S: the former offered accuracy and relative precision, the latter did not. And as it happened, these attacks ‘established a pattern, both in technique and result, from which there were to be few important departures in the major actions which were to follow.’ This said, while a number of attacks were made across Germany and the occupied countries, the majority of raids would take place in the Ruhr where distances from the UK meant that Oboe could be used.11 There was, however, one indication that was independent of navigation device. As results of the raids came in, it became apparent that regardless of the aid used to identify the target area, groundmarking was normally more successful than skymarking for pinpointing the actual target, despite the generally good results of the latter.12 Throughout, the ORS conducted intensive analyses of the various raids. Their reports provided the grist for profound review of events and previous decisions. And sometimes the scientists stepped, however metaphorically, on toes. One such case occurred in April. W/C Saward, the W/C Radar, provided the CSO with his detailed thoughts (which he had prepared in conjunction with the CNavO) about ORS reports for an attack against Hamburg on 3/4 March. He noted the apparent error of one PFF marker which had been led astray by the fact that the tide was out at Hamburg. The error should have been preventable, but in fact it had led the backers up, operating without H2S, to drop their markers on erroneous TIs. Saward appears to have acknowledged that error, but he was not happy with another observation by the boffins. The criticism by O.R.S. of the many main force crews which bombed this Wedel area and yet reported seeing red T.I’s 10/15 miles to the east is, I think, a little unjust. When three distinct areas are marked in an overall distance of approximately 15 miles, and when crews are expected to attack the markers it must be confusing to any crew and very difficult to decide which is the correct marking, bearing in mind that the others may possibly be the result a little cunning on the part of the enemy.13

Overall, thought Saward, ‘it is felt that H.2.S. is being judged on its ability to do something for which it was not designed.’ It had been envisaged that the airborne radar would put a bomber over a specific town but never a specific point within that town, and Saward believed that the device would never permit the degree of precision hoped for. In his view, if H2S was to be used for marking, then large numbers of markers

120 The Science of Bombing

should be sent out and the main force directed to bomb the centre of the TIs so placed. Finally, said Saward, it might be worthwhile to conduct tests against such cities London or Glasgow. Dickins was in agreement with the need for such trials, but his account of the discussion of trials against London in particular was somewhat different from that of the radar officer. Dickins recounted that the proposal was first suggested in ORS B 128 ‘Attack on Berlin 1/2 March using H2S,’ which was published shortly after the raid.14 Subsequently the CSO reviewed the reports and various minutes before sending them to Saundby. His own comments were candid and not necessarily in line with his subordinate. ‘The reports are very depressing and I find it difficult to add usefully to the conclusions and recommendations already made by O.R.S. and Dr Dick[i]ns.’15 Acknowledging the limitations of the equipment, he went on to suggest that the long-term remedy lay with the widespread introduction of H2S across the command. In the shorter term it would be up to the PFF to come up with a solution. Saundby noted all these comments and asked the staff to have the trials against London and Glasgow conducted by the PFF.16 Just as these deliberations were winding up, Dickins sent forward reports on the next two H2S raids: Munich (9/10 March) and Stuttgart (11/12 March), both of which had experienced problems. In the first instance, the PFF had marked the target, but also some other areas, and the Main Force went after these initially. Only determined visual backing up of the actual target had saved the raid. Based on this, Dickins suggested that when weather was suitable the H2S aircraft should act as illuminators, leaving the actual marking to visual techniques. The Stuttgart raid had been well marked, but the main force had arrived late and the attack literally fell short. Dickins made mention of this undershooting and suggested that in view of the approach to the target from the southwest ‘it might be desirable to have made the aiming point 1 or 2 miles to the northeast of the one selected.’17 On 23 April 1943 the ORS issued their third review of PFF operations. A draft of Report B 133 had been sent to Saundby by Dickins three days earlier. In the attached memo the scientist summarized the report, indicating that success when using Oboe for groundmarking was three times better than predicted, but, by comparison, H2S groundmarking had not met expectations. Skymarking results could not be confirmed but it was thought that the same trends were developing. Saundby passed the file to Harris suggesting that the CinC review the detailed conclusions in the actual report.18 Here the various percentages were spelled

New Concepts: March–June 1943 121

out.19 Oboe raids were achieving 35 per cent attack within three miles of the aiming point as compared to 13 per cent in pre-PFF raids on the same targets. Less pleasing were the H2S attacks, which measured only 16 per cent within three miles compared to 19 per cent on earlier raids. Thirteen groundmarking raids had taken place against German targets: only the two Oboe attacks had been ‘unambiguous’ in their marking of the actual aiming points, while of eleven H2S attacks only four showed improvement. Successful H2S attacks had come when the main force had ‘selected the correct markers of all those placed by the P.F.F. Hence, since the A.P. has been marked on most occasions, the most urgent need is to devise a method of preventing ambiguous marking or of enabling the main force to discriminate between accurate and inaccurate markers.’ The report also said that Oboe skymarking had been threefold better than expected, although the sample was small, and that there were no conclusions to be drawn about H2S skymarking. Finally, noted the scientists, even when the aiming point had been precisely marked the bombing suffered from main force scattering. Harris simply initialled Saundby’s minute, but there can be little doubt that this information would be critical in getting the most out of each raid in the coming campaign season, and it is likely that recommendations about focusing the main force on the right TIs contributed in no small way to the concept of the use of a ‘controller’ or Master Bomber. Two weeks later Dickins was able to add more grist to the discussion in the form of the analysis of the raid on Berlin on 29/30 March. Marking had gone well at first, but the main force again arrived late and by then there were markers on the near side of the target area which had apparently been considered good enough for the crews from the main force squadrons. This was also the first raid to use the timed run (a short distance of fixed time and bearing) from a landmark, a procedure which had been approved in a CinC’s conference just days before; unfortunately the scheme had not worked as hoped. The point had not been illuminated or marked as briefed and, worse, ‘the main force saw a concentration of T.I’s and bombed it even though the position did not coincide with the E.T.A. [for the target.]’ Dickins was not ready to abandon the timed run; rather, he thought that many crews had not adhered to the plan and that the plan itself had needed to be more precise about what was to happen, as well as how and when. In his minute to Saundby he said: ‘I would suggest for your consideration that the whole question of greater attention to the detailed plan at briefing should be strongly emphasised throughout the Command.’20 Dickins was not pulling his

122 The Science of Bombing

punches; one wonders if he might have been not a little frustrated that a good workaround for getting the most out of H2S was being scuttled by lacklustre performance from the squadrons. At the end of June Dickins again offered his thoughts on the timed run and on the use of a master bomber to make good errors in placing the TIs: We believe that the ‘timed run’ method is worth a further trial but only as a means of eliminating incorrectly placed T.I’s. I think it will be just as difficult to mark the point from which the timed to run is to be made as to mark the target. The former has, however, the advantage that it will not get obscured by fires. I fear that the timed run may well lead to undershooting unless we can ensure that no evasive action is taken during the time period. We do not feel that the use of a controller keeping up a running commentary on the position of the target indicators is likely to be very successful. … It seems, however, a more practical solution than dropping markers to cancel erratically dropped or bogus T.I’s.21

While Dickins was not prepared to give his full support to the controller concept, some of the Ops staff were, and made mention to the SASO of 5 Group’s successes when using a controller during various small raids. They suggested that the technique should be transferable to bigger targets.22 In early July Dickins passed the D/CinC a review of H2S as a groundmarking aid. The actual report was not intended for wide distribution and a separate version was under development for transmission outside the Command. The internal version admitted that there had been problems with the number and serviceability of H2S systems and that, even when working, the equipment had limitations whenever the PFF aircraft had to manoeuvre fairly dramatically in doing their jobs. Some successes had been noted and it was felt that these were an indication of what was achievable. Perhaps most significantly, it was seen that the original technique of using H2S aircraft to conduct marking had not worked well, but that using these aircraft to illuminate the target area in advance of visual markers did have promise. Of the thirteen raids analysed, the six which used this latter technique averaged 44 per cent of photos within three miles of the aiming point compared to just 11 per cent that used H2S marking. While it was clear that H2S was not a target-finding device, the boffins were not about to let up on their well-founded criticism or on suggestions for finding a viable method to employ the system.23

New Concepts: March–June 1943 123

If by the summer Harris and Saundby had become frustrated by H2S’ failure to perform as a precision bombing device,24 the source of their frustration was the evidence so clearly distilled by the scientists. Their aviators’ concerns had been building over some time and in the midst of the battle, as the analysis of these issues was on-going, when Harris pulled his senior leadership together. On 23 April the CinC met with his senior staff and the Group commanders to review recent operations. He wanted to find ways of improving results and began the session by making three points with regard to existing problems: first, some crews were dropping their bombs before the designated aiming point; second, bad bombing seemed linked to bad navigation and to the lack of sufficient Mk XIV bombsights; and finally, many crews seemed incapable of appreciating the critical nature of the bomb aimer’s duties, while many bomb aimers themselves were of poor quality. Before getting to the these matters in detail, Harris first turned to concerns over losses, believing that these could be reduced if bombers crossed the fighter belts at extremely low altitudes, that is, below 1000 feet. Others in attendance, however, expressed concern with using this technique over Germany generally and also pointed out the negative effect on navigation. It was agreed that this tactic should only be used when conditions were exceptionally favourable. Saundby then presented the case for concentration, stressing the highest possible density on dark nights when such tactics would swamp the GCI defence. This was accepted, but it was also deemed necessary to review concentration on light nights when catseye fighters benefited from a densely packed main force. Moving on to the threat from anti-aircraft fire, Harris then spoke about evasive action against concentrated flak and expressed his concern that it did not much help. Dickins confirmed that evasive manoeuvres were in essence counterproductive and that the best defence was to keep up speed, make a good bombing run, and get out of the area of the flak. Dickins was directed to prepare a paper on the matter. Returning to questions of the PFF and marking, Harris said that he wanted to continue to perfect techniques, though there were admittedly problems. It was now that he approved the use of a timed run if this would keep crews from being led astray. To minimize the chances of a raid being confused by inexperienced crews, he also approved a concept which would see the least experienced crews bomb last so that ideally the PFF and seasoned crews would have accurately identified and bombed the aiming point. 25 Undershooting the target, which the boffins had often described in

124 The Science of Bombing

their assessment of Oboe and H2S bombing, continued to be a concern, and in June was the subject of a CNavO memorandum.26 The senior Navigator posited that there were two plausible causes: that a visual perspective error of the Mean Point of Impact (MPI) of Target Indicators led to early release; or that there was a problem determining the actual bearing from the aircraft to the aiming point. In the latter circumstance he said that even a vigilant bomb aimer could do the math incorrectly and end up with a 600-yard error. Since attacks took place in three sections, this meant that it would be possible for this 600-yard error to be renewed by both the second and third sections, leading to a cumulative error of 1800 yards. Asked to review the document, the boffins were convinced that the first source of error was likely to be the more significant, but they added three additional possibilities in putting the issue before A/V/M Oxland: underestimation of distances when viewing the ground from bombing altitude; seeing only nearer TIs and thus determining the MPI from these alone; and ‘anxiety to get rid of bombs.’27 Moreover, they concluded, there were so few cases of overshooting (dropping bombs beyond the aiming point) that, until any tendency to overshoot presented itself, the aiming point for a city should be adjusted to the far side of the target area and this distance undoubtedly needed to be greater than 600 yards. Acknowledging the importance of undershooting, Dickins indicated that the boffins would undertake a detailed study. Oxland agreed with Dickins’s comments and the proposed study, as did Saundby.28 More work was now done by the scientists to refine their ideas on short bombing. They concluded that there were both psychological and physical aspects to the problem. In the first case there existed two possibilities: a fear of going into defended areas that tended to produce ‘wild shots’ (which was less pronounced than it had been); and a rushing to get rid of bombs when over the target where crews simply wanted to drop their loads over the first visible TI and turn for home. Physiologically there could be problems seeing TIs due to smoke or obstruction from buildings. The scientists acknowledged the perspective problem, but did not consider it as significant as the CNavO had suggested. Rather, it was thought that crews were likely to see what they wanted to see and if something looked like the target, then that was good enough; this problem was to some degree associated with inexperience. While all of these errors occurred randomly throughout the crews assigned to a raid, the boffins thought that they did have a cumulative effect, which generated a systematic undershooting error.29

New Concepts: March–June 1943 125

These notes were shared with the Groups and a revised list prepared, which included several more possible sources of error. First, pilots often headed for the presumed target without taking account of wind drift, and this could play havoc with the tracking system in the bombsight. Many bomb aimers felt that lobbing their bombs anywhere among the TIs was good enough and that it was actually best to aim for a spot where there was no fire rather than contributing to a target already engaged. These errors the scientists called ‘blitz-consciousness,’ a term which typically meant that the crew would make only one run over the target and it had to be good; if the bomb aimer made a mess of it, there was no second chance. The Groups agreed with the suggested psychological errors, but also added that crews with over twenty-five trips, and thus approaching the end of their tour, tended to hang back. Details of the visual problems were generally accepted, but it was also felt that in general bomb aimers had difficulty in estimating the MPI and that instead they should be told to bomb a specific TI. Most important was the challenge of dealing with TIs that never spread the same way twice. Procedurally it was thought that having the track pass directly over the target was important and that the last turning point before and the first turning point after the target should be marked. Finally all groups wanted a timed leg to the aiming point as a means of a cross check against the bomb aimers’ identification of the TIs. This would help mitigate the effects of evasive manoeuvring and poor track-keeping which invariably contributed to undershooting.30 It seemed that in following up on the CNavO’s two theories the boffins had opened a real Pandora’s box, but, as was now the norm, Dickins’s staff had explored the issue in great detail and come up with comprehensive explanations and workable solutions. Indeed, having looked at all these issues and then the data from various raids, by the middle of August the boffins were confident that they could recommend an adjustment in PFF aiming points. For Oboe targets it was considered necessary to place the TIs 1¼ miles past the Command Aiming Point, and also adjust for cross winds, regardless of wind strength, by placing the TIs upwind by ½ mile. With regard to H2S marking they felt that undershooting would be at least two miles short of the TIs and so the PFF markers should be placed two miles beyond the Command Aiming Point.31 The basis for these conclusions was, as always, the data taken from bomb release photos. In June Dickins sent Harris and the senior leaders the Section’s views on the question of bombing photos. In view of the arguments which take place from time to time concerning

126 The Science of Bombing the reliability of night photographs as an indication of the fall of bombs, we have made an investigation into the relationship, the results of which are incorporated in the [enclosed] report which you may like to see. Taking into account the different types of bombs and the various alterations in course which aircraft make after bombing, it is estimated that in general over 60% of the bombs will fall within ½ a mile of the centre of the photograph taken under the present procedure and 85% within ¾ of a mile.

The actual report, which dealt in much more detail with the technical aspects of the question, stated that ‘within reasonable limits of tilt and bank … these percentages are reduced to about 46% and 75%.’ The two slightly different sets of values were not at odds, for it was accepted that while aircraft should be making as stable a bomb run as possible there would be occasions when they were manoeuvring for a variety of reasons.32 Soon thereafter Dickins provided this information to the Groups. At 5 Group, as usual, the news was received somewhat coolly. The Group’s SASO, in forwarding the document to the AOC, went so far to say that there was no need to read the report, the covering letter would do. In contrast, the Photo Officer signalled the existence of the debates which took place at squadron level around the question of what the bomb release photos actually depicted and recommended sending copies to the units. The SASO relented, directing that the documents be sent to station commanders in sufficient numbers for the squadrons.33 This tendency to ignore or refute the scientists’ conclusions is worth examining. Many factors could have been at play in these various incidents. For one thing it has been seen that a previous 5 Group AOC, Slessor, was not an unconditional fan of scientific conclusions. Perhaps those working in his former headquarters had picked up on his philosophy. There may have been Service politics at play too. We do not know, for instance, if these staff officers at 5 Group were friends, acquaintances, or rivals of Oxland, Saundby, or Harris. We do not know, either, if these officers were distrustful of the scientists and their work. Finally it may well be that like all professionals they had a strong dislike for having someone else, particularly someone from outside the profession no less, tell them how to do their jobs. It is interesting to note a more positive exchange of minutes which occurred some months later between Air1, a member of the Group’s Air Staff, and the ORS officer at 5 Group, Mr Curry. The latter seemed to go out of his way to offer a comprehensive response to a simple question and his comments appeared genuinely appreciated.34 Was he building

New Concepts: March–June 1943 127

new bridges or, as the Group’s own scientist, did he simply have the trust and confidence of one co-worker for another? It was not at all coincidental that Harris had begun his 23 April meeting on bombing accuracy by talking about losses. The two were always present in the thoughts of the commander and his staff. In early March, as the targets shifted towards the Ruhr, Dickins had been approached by the Ministry of Aircraft Production to provide loss data as it related to the direction of attack of enemy fighters.35 It was felt that knowing how German fighters attacked would be important in making decisions about the armouring of RAF bombers, a topic which, we have already seen, the scientists were also following. Dickins replied on 12 March saying that data was being collected to determine if attacks were being made from below, but that progress was slow, first, because damaged bombers often crash landed, making it hard to determine to what extent damage to the underside of the aircraft might have been caused by enemy action. Additionally, the engineering officers who inspected crashes were not trained in damage assessment – at least not in determining the causes. The ORS chief did add that the training and placement of ORS inspectors was soon to right that particular problem.36 The Scientific Research staff at MAP came back to Dickins in July with another hastener asking for the data. Harris was now saying that attacks were being made from 75 to 90° below the horizontal and the MAP wanted to see the data on which these statements were based. This time Dickins replied that he would send along what was available in short order.37 Short order turned out to be the end of August and the data and conclusions were anything but definitive. Dickins said that sources were questionable and that determining what amounted to attack from below was difficult, not just because of the damage from crash landings, but also because bullet and cannon holes could be scattered about the underside of the aircraft and because manoeuvring by the bomber as these strikes were occurring could make the direction of attack appear to be other than what it had actually been. In the end, Dickins provided data on just sixty-nine attacks, without stating the number of sorties from which these cases were taken. Of these, only fourteen appeared to be from attacks at less than 10° below the horizontal. This apparent obfuscation did not mean that the boffins were not committed to finding out what was going on. Dickins closed his letter stating: It is hoped in due course to obtain further information covering tactics of Enemy Aircraft which shoot down our Bombers from cases in which some

128 The Science of Bombing of the Bomber crew succeeded in returning, and we should then be in a better position to assess whether Bombers do frequently fail to return as result of steep attacks from below. In addition when the fire and explosion risk are reduced, considerably greater numbers of aircraft should get back and it should thus be possible to assess more accurately the usual directions of attack.38

This was after all operational research and not a tidy controlled laboratory exercise. Crew losses by and of themselves did not tell the whole story, and the boffins were always wanting to extract the maximum knowledge from the data. Thus in February the scientists had been intrigued by a letter from Curry at 5 Group. He had collected a range of human data which painted an interesting picture of how No 9 Squadron operated, including the characteristics of crews suffering losses. After some reflection, Dickins set his staff to gathering more information: ‘While I agree that the evidence on which the conclusions are based is very scanty, the investigations suggest to me desirability of looking at losses, crashes, abortive sorties and photographs squadron by squadron for different types of aircraft over selected periods. We may well find that some squadrons are worse than others and some useful lines of investigation might be revealed.’39 This decision was indicative of the scientists’ desire to be able to explain all facets of the conduct of the bombing. If their work was to have a positive impact on the human side of efficiency, then they needed to understand how units operated and what made for good and bad practices. Their work over the preceding weeks had been a reflection of this. Whether the issue was navigation and bombing accuracy, the vulnerability of certain aircraft systems, or the relative merit of one aircraft over another, their findings and conclusions were of central importance to the decision makers. For example, the analysis of the Ruhr raids showed a marked improvement in bombing effectiveness, but the reverse side of the balance sheet indicated an alarming rise in the casualty rates of the spring and summer of 1943. Loss rates of 4.5 per cent were reported and casualties (losses plus damaged aircraft) ranged from 20 to 25 per cent. In July, as the battle drew to a close, it was deemed necessary to take some action, as casualties were now approaching what was later called an ‘unbearable’ or ‘insupportable sustained’ rate.40 By that time it was estimated that 70 per cent of the losses were due to night fighters and that half of these were radar-controlled intercepts. Of the losses due to flak,

New Concepts: March–June 1943 129

approximately two-thirds were thought to be the result of radar-directed guns.41 In the opening phase of this first battle of the year, the scientists were still examining loss data from the previous year. It was only in April that ORS S 91 ‘Night-Bomber Losses on German Targets, 1942’ was ready for publication.42 The report had taken until late April to prepare and only in early May did Dickins send it forward for approval and release. Strangely, Saundby did not return it to the scientist until early July and then it was only actually released by Dickins in August.43 Why this apparently useful and important data was so slow in being distributed is unclear, as the review did offer a long list of conclusions. At the same time, however, the report urged much caution due both to limited data and to the indirect source of the information, the authors emphasizing that the conclusions were really ‘estimates.’ In interpreting their findings they had divided the country into three zones: western Germany – essentially the Ruhr; northern Germany including the cities of the Baltic and Hamburg; and southern Germany – the area stretching from Frankfurt to Munich. Analysis was extensive and in fact stretched back to the summer of 1941; however, the time frame reported on was only 1942. There were several conclusions which would have been of great value to the operations staff and many of the special staffs within the headquarters. Overall losses experienced during 1942 had shown a slow increase, averaging 2 per cent across the three zones. Aircraft reporting damage by either flak or fighters had also increased during the year. Losses to non-enemy causes were set at about 0.5 per cent of sorties, the greatest single cause being attributed to engine failure. While flak losses had remained generally constant at between 1.5 and 1.75 per cent of sorties, fighter losses had risen from approximately 1 per cent at the beginning of the year to about 3 per cent in the summer, with that number dropping to 2.5 per cent at the end of the year. The number of attacks made by fighters had also risen. Flak losses seemed to be much more concentrated around targets in the northern zone, where 80 per cent of losses seemed to occur over the target; the figure dropped to 75 per cent in the west and only about 40 per cent in the south. Losses to flak in coastal areas were deemed to be about 10 per cent for attacks in the west and 30 per cent for attacks against the south. It was also noted that, beginning in April 1942, fighter attacks against returning bombers had become much more commonplace; it was felt that this was to some degree due to the typically increased dispersion of returning bombers, but also to the fact that the enemy seemed unable to scramble its entire force prior to the actual

130 The Science of Bombing

attack. Finally it was felt that approximately half of losses due to enemy action could be attributed to aircraft catching fire after being hit. No recommendations were put forward by the scientists, and while it may have appeared that there were no major revelations in this document, it did bring together many of the most essential elements of data, allowing first the scientists and subsequently the planners and commanders to see at a glance what generally were the challenges facing the crews. It should be noted that this report came approximately halfway between the last monthly report on losses at the end of December 1942 and the resumption of that series in July 1943; as such, it filled a gap which might not yet have been fully perceived. But, it certainly permitted those in the Headquarters the opportunity to zero in on those issues which might make the campaign more efficient, while at the same time reducing the ever-worrisome casualties. Several weeks earlier, in April, as the result of many queries seeking explanations for rising Lancaster losses in December and January, the ORS completed a report dealing with these casualties. While at face value the figures in question seemed ‘disquieting since they imply that the Lancaster with a record well above the average until November has since that time descended to the common level,’ the boffins pointed out that the apparently dismal statistics for January (4.5 per cent) had been the result of sending almost all Lancaster sorties against tough German targets while other types had operated against French ports. Looking at Lancaster operations in isolation, comparisons between experienced and newly formed Lancaster units and between those flying Mark I and newer Mark III aircraft revealed no trends. The report indicated that more investigation was needed, but the authors felt that ‘the present results [should] allay any suspicion that the high performance of the Lancaster is seriously decreasing.’ Saundby gave Dickins the go-ahead to distribute the report, suggesting that it go directly to all AOCs whose groups were operating Lancasters.44 In putting together these figures and presenting their conclusions, the scientists clearly appreciated that the crews and indeed their commanders needed to be assured that they were not going into harm’s way with anything less than the best chance of coming home. Towards the end of April Dickins’s section head in ORS 2 (Bomber Losses), Dr R.J. Smeed,45 put forward a summary of the effect of Gee on casualties.46 While the data was somewhat dated, the conclusion that Gee did in fact reduce losses was thought to be significant and worth making known. Specifically, it had been demonstrated that the use of

New Concepts: March–June 1943 131

Gee had a noticeable effect in cutting loss and damage rates and that this could be attributed to Gee-equipped aircraft flying in higher concentrations than non-Gee bombers. Gee aircraft benefited from the concentration and particularly from the reduced chance of interception or flak engagement. On the other hand, non-Gee aircraft were likely to spend more time over enemy territory while carrying out their own navigation and target identification, with a resulting higher risk of engagement by the enemy. The report also concluded that more Gee aircraft managed to return to their home base, rather than to just any airfield, than did non-Gee bombers. While this point was not expanded upon, the positive impact on efficiency was obvious: aircraft and crews would be back where their supporting units were located and there would be no waste of staff effort in getting errant aircraft back to their squadrons when these same staff officers should be focused on organizing the next op. The draft copy of the report went forward to the D/CinC on 30 June, but Saundby approved it only on 16 July,47 so there seems to have been little impetus to get the information into the field. Indeed, the report acknowledged that the data was dated and that almost all aircraft now had Gee on board, but the scientists had thought that the evidence might be useful when considering the introduction of other new equipment and procedures. Here then was a second case of a report, the first being the 1942 summary, languishing in the D/CinC’s office. Why had it taken so long for these documents to be seen and sent on? One can only imagine how busy Saundby and others were with current matters. Trying to glean and apply insights from data that was in some cases well over a year old and based therefore on old aircraft, equipment and tactics might have seemed pointless or at best been relegated to those few moments when the phone was not ringing off the hook. In May the ORS completed a much more topical study on PFF losses after the pathfinders’ casualty rate had grown relative to the main force in the months of March and April. The investigation was characteristically thorough, with comparisons made between a number of variables. It was found that the loss rate of fully trained PFF crews was actually lower than the main force, while that of untrained pathfinders was ‘strikingly high over the whole period’ from October 1942 through April 1943, averaging 6.9 per cent to the main force’s 3.6 and the trained PFF crews’ 3.1 per cent. Untrained crews flying the Stirling had suffered fearful losses of 14.7 per cent, where trained crews on the same type had just 3.8 per cent casualties. The report concluded, however, that taken together the PFF losses were not alarming and in the case of Halifaxes were actually

132 The Science of Bombing

well below the norm, this attributed to the PFF crews’ greater experience on that type. The document was passed to the D/CinC but no indication of his response is recorded.48 Losses due to in-flight fires remained a constant concern. In June the Chief Engineering Officer (CEngO), A/Cdre Roach, and others were called to another demonstration at the Royal Aircraft Establishment (RAE). This time the subject was fires in engines, fuel tank fires having been investigated a year earlier. The displays were convincing and the RAE was asked to proceed with making firm proposals to fit Methyl Bromide extinguishers for each of the heavies, while at the same time Roach recommended that proposed modifications to fire-sensing systems go ahead.49 Specific data on weights of various configurations of fire suppressors reached High Wycombe in early August, and Saundby commented that he hoped they would do what they promised, as ‘there is no doubt that fire is the cause of a very high percentage of our losses.’50 Saundby’s thoughts were mirrored by data made available just a few weeks later. Smeed had provided a summary of reports from ‘evaders,’ crew members who had bailed out over enemy territory and subsequently made their way back to the UK.51 Of the thirty-two cases examined, fire had occurred in twenty-six (81 per cent), and it was thought that about half of the sorties could have made it home but for these fires. Of the twenty-six it was possible to state that fire had started in the engines in eleven cases and the fuel tanks in nine. Smeed underlined that this loss rate was far higher than what had been portrayed in their report S 74, ‘A Note on the Fire Risk in Bomber Aircraft,’ punished at the end of the previous year.52 In reviewing the proposed equipment additions Roach was, however, less enthused than the D/CinC and the boffins. He pointed out that at no time had the Command actually asked for the newly proposed extinguishers in the tank bays (the hollow area inside the wing which housed the fuel tanks). While he agreed to engine extinguishers, he felt that tank bay extinguishers were unnecessary, given the nitrogen systems which were to go inside the tanks; the new extras were just that in his view – extra weight. Some days later he indicated that if the Command was stuck with taking these tank extinguishers then they should first try out the internal nitrogen and see what the results were. Only then should they add the tank bay extinguishers and thus be able to tell how much more this equipment helped reduce losses. This approach did not sit well with D/SASO, who felt that it would be impossible to isolate the effectiveness of each system, given all the variables at play during

New Concepts: March–June 1943 133

sorties. Moreover, he said: ‘We should not, I think, withhold from our crews any equipment that adds to their safety.’ It was at this point that Dickins provided a reasoned compromise. Before the Methyl Bromide extinguishers were adopted, would it not be worthwhile to ask the Ministry to provide figures on their predicted effectiveness? In the meantime, he added, the Command could do some analysis on the probability of fires in tank bays. CEngO disagreed vehemently, stating that before long these systems would add a thousand pounds to an aircraft’s weight. He had ‘considerable doubts’ about the need for the tank bay extinguishers. All of these recommendations and opinions were put before Saundby, who agreed with the previously requested inert nitrogen for inside the tanks and with Methyl Bromide extinguishers for the engines; together these would weigh in at about five hundred pounds.53 Other studies were also completed at the mid-year point. In one case, the scientists had been approached to provide figures on engines and vulnerability to enemy action. In a June letter responding to the Director of Servicing and Maintenance at the Ministry, the boffins put forward data on engine casualties between December 1942 and February 1943. During that period, of the 5712 sorties by liquid-cooled-engined heavies (such as the Lancaster Mks I and III), only 39 had had hits to the engine and of these only 21 had sustained damage; data for radial engines (such as used on the Stirling and Lancaster Mk II) were 1194, 14, and 10 respectively. The boffins had concluded that there was little difference between the two engine types. The letter closed by indicating that a more detailed study was contemplated. 54 At the end of June the Section completed a wholly different investigation, looking into the question of damage caused to bombers by incendiary bombs dropped from other RAF aircraft. No strikes had been reported in January, but as sorties and concentration started to increase, the figures for May showed that twenty-four aircraft (0.53 per cent of the sorties for the month) had been damaged in this fashion. The size of the raid, they said, had little to do with the phenomenon; rather it was the concentration, and mathematically it could be proven that for the currently used densities of between eight and eighteen aircraft per minute the rate of strikes should range between 0.38 and about 0.84 per cent. This was an acceptable and minimal risk, they felt, as the type of damage incurred probably meant that only a few aircraft each month were actually lost to this cause. It was also predicted that the introduction of cluster bomb units (which would keep the incendiaries from spreading until well below the operating altitude of the raiders) would essentially

134 The Science of Bombing

eliminate the problem. For the moment the only solution was to reduce concentration and this was not recommended due to increased risk from other sources.55 It was also in June that Dickins put forward more evidence of the problems with the Halifax, this based on data brought back by his staff from a visit to 4 Group. The effect of experience, or more precisely inexperience, on losses was particularly heavy, with some 47 per cent of those crews gone missing between January and May having completed less than four sorties. Significantly, loss rates were appreciably lower in those squadrons which had done fighter affiliation training, with the chance to practise evasive manoeuvres against a real fighter. The boffins pointed out that Halifax losses on easy targets, those where little manoeuvring was required, were about the same as other aircraft types. This said, it was reiterated that the extreme losses for inexperienced captains and the comparatively lower loss rate for crews who had had the benefit of training made it clear that more training would make a difference. Dickins underscored the fact that the AOC, A/V/M Carr, was looking for additional fighter support for training. Dickins also took this opportunity to recommend the reintroduction of the policy of sending new crews against relatively safe ‘fresher’ targets, where the risk of fighters and flak was minimal and new crews thus had a chance to gain experience with the full complexity of operations. He argued that while sending junior crews against easy targets would detract from the weight of attack on main efforts, photo evidence showed that these new crews rarely came within three miles of the aiming point in any case. His message was clear: it would be better to prepare new crews to effectively fly these demanding aircraft rather than waste their bombs, and frequently their aircraft and lives, before they were ready for a high threat environment. Saundby fully agreed with the suggestions and asked the air and training staffs to make the additional training a priority.56 The Halifax data highlighted the inescapable link between losses and defensive practices, in this case proper training in aircraft handling to acquire the skill necessary for evasive manoeuvres. Indeed, it was only weeks since the question of teaching crews the current evasive tactics had been debated. At that time the D/SASO referred to a recent review by the Tactics and ORS staffs: ‘the Air Staff view is, in brief, that there is little to be gained by taking violent evasive action to avoid being shot down by fighters or flak.’57 The associated Air Staff memorandum to the Groups opened by stating that ‘the Operational Research Section of this Headquarters are of the opinion that over a very hotly defended target evasive

New Concepts: March–June 1943 135

action for the avoidance of flak, is meaningless, especially when a high concentration of aircraft is achieved.’58 The letter continued, explaining the negative results of evasive manoeuvring: The collision risk is seriously increased. It results in no saving of aircraft. Attempts by turning away to avoid flak bursting ahead are just as likely to lead to a hit from other bursts off the original track. Violent evasive action makes it impossible for gunners either to see or to hit attacking fighters. Finally, evasive action in the target area makes accurate bombing impossible and necessitates, therefore, repeat attacks; these in turn lead to an overall higher total of casualties in achieving a given object.

The communication went on to cite a 5 Group attack against Dortmund on the night of 4 May. The Group had directed that no evasive action be taken on the run in to the target. Accurate bombing and speed were to be paramount. Of the 125 sorties involved only one had been lost. All AOCs were asked to consider this example, although it was admitted that the evidence could give a false impression. The letter concluded with a personal statement: ‘I need hardly point out the vastly improved bombing which could result if in fact we find that evasive action does not pay and that a straight run across the target exposes the aircraft to less risk than the longer run caused by weaving or violent evasion.’ While the example of the letter on file is not signed, the fact that it was addressed to the AOCs and not just the Group headquarters as well as the use of the first-person style of the last paragraph suggests strongly that it was intended for either Harris’s or Saundby’s signature. This was not just staff musings, but rather, commander’s firm direction and was based without question on ORS advice and ORS Air Staff collaboration. Later in the month the ORS produced a tactical memorandum that reviewed the question of evading flak in the target area and provided the details that crews would need to understand the principles behind the interdiction against evasive action.59 In its introduction the paper made a number of straightforward statements about how crews were reacting and should react to the increasing intensity of fire from the ground. Much evasive action is normally taken with a view to minimizing the effectiveness of this fire and bomb-aiming is in consequence rendered considerably less accurate and many bombs are wasted. The enemy is, therefore, achieving his purpose to a great extent. … a large part of the evasive action at present being carried out is completely useless against any form of

136 The Science of Bombing A.A. fire and much of the rest is very doubtful. No form of evasive action can possibly be of any value against deterrent or barrage fire and with the low accuracy of A.A. fire against unseen targets and the high concentration of aircraft we now achieve over enemy targets, even the best forms of evasive action are not of so much value against predicted fire at heights above 12,000 feet.

Compounding the problem, evasive action had many unintended consequences: it increased the time needed to get through the target area and with it the time exposed to flak; it could cause serious structural strain to the aircraft (and occasional structural failure); it ‘materially increased the collision risk’; it made seeing and engaging enemy fighters almost impossible, while not seriously affecting the fighter. Evasive action also made accurate bombing impossible and meant the crew would have to make at least one additional bomb run, thereby increasing the chance of damage or loss; otherwise the sortie would be a waste. The boffins recommended using speed and trading height for speed when able, the idea being to pass through the target area as quickly as possible.60 If serious close flak was encountered they urged changing course significantly, at least forty-five degrees, for at least fifteen seconds, and, in the worst case, exiting the target area to set up for another run. The body of the paper explained the technology of flak and demonstrated that flak bursting more than four hundred feet from an aircraft had little chance of damaging anything vital; it more or less took a direct hit for flak to bring down a bomber. The scientists included diagrams of actual aircraft tracks showing how fairly steep and abrupt banking did little to actually change the path of the aircraft. Additionally they included sketches which depicted good and bad methods of avoiding flak. Compared to their views from the year before that evasive action was a valid technique the scientists were now firmly committed to this alternative policy. Their research had convinced them as completely as it had the air staffs and commanders. As always, discussion regarding concentration was frequent among the Air Staff, but the scientists were not as a rule drawn into these conversations, despite the fact that they had developed the science to support the concept. Nevertheless, in June Dickins had recommended to the DCinC a short letter for distribution to the Groups that re-emphasized the need for concentration not only generally but in particular after leaving the target. Included were two Gee plots which showed how returning bombers had become dispersed. While the letter, signed by Saundby, did not

New Concepts: March–June 1943 137

mention it, likely for security reasons, Dickins’s minute stressed the need for concentration in view of ‘certain countermeasures’ which were about to be introduced. The letter did, however, indicate that beginning in mid-July a specific return route from the target to a point fifty miles off the German coast would be directed for each raid. 61 Indeed, as the Battle of the Ruhr concluded, the Command had already started to employ feints and frequency jamming to throw off the intercept and control practices of the Germans. What lay ready for use as the Battle of Hamburg opened was the ‘most promising’ countermeasure – Window. The device had been available for use since the spring of the previous year but had not been introduced because of fears that the Germans would be able to copy it quickly and use it to re-energize their own night bombing campaign over England. Now, the decision to introduce Window was made by the CAS on 2 April 1943. Despite problems with supply and the lack of an automatic dispenser, it was decided to employ Window for the first time on 1 May. Then the Chiefs of Staff Committee delayed introduction until after the invasion of Sicily: first use was now set for 23 July, the opening raid of the Battle of Hamburg. While Window and the other deception tactics employed by the Command could not destroy enemy fighters and guns, they could neutralize them, and this, for the time being, was the best that could be hoped for.62 The preceding months had been eventful, if not fully satisfying for the scientists and for Harris. Given new technologies with which to focus their striking power, the potential to strike Germany hard had been turned into reality on many occasions. Still there had been some missteps and it was the scientists who had zeroed in on these in many cases. They had continued to provide the quick feedback of the Interim Raids Reports and to do the comparative analyses which uncovered trends, concerns, and often solutions. Among their particular successes was the confirmation of viability of the technique which they had proposed for Oboe during the previous year. Recognizing that H2S was the weaker cousin, they had carefully monitored its use by the PFF and had not been reluctant to voice their concerns. Gathering all this data, they had come up with methods for dealing with poor weather over the target. Navigation remained a concern and here too the boffins had been vigilant to the imprecisions of both the main force and PFF navigators, providing details of the problems to both the Navigation staff and the senior leadership. Always cognizant of the impact of losses, the scientists had probed a variety of questions, focusing on in-flight fires and bringing the importance

138 The Science of Bombing

of fire-extinguishing systems to a sometimes hesitant staff. At a more general level they had also been ready to offer comparisons between crews of various experience levels and organizations, all in the hope of getting those most in need of additional training or encouragement to fly as effectively as possible the assistance they needed. They were even prepared to change their views and recommend abandoning evasive manoeuvres against flak when it was seen that such tactics were counterproductive and actually increased the risk to the flyers. Data had begun to emerge pointing to the limitations of the Halifax; even its engine exhausts had been a major concern during these months. While these lessons and developments were all significant and required considerable attention, they had to be analysed and acted upon ‘on the fly.’ There could be no pause, as the campaign shifted to targets deeper in Germany, targets with their own unique challenges.

9 Hamburg and New Science: July–October 1943

By late July 1943 the Command had been making use of the most up-todate bombing aids in large-scale operations for seven months. Results had been better than in the past, but not without disappointment. While Oboe generally worked well for target marking, the introduction of H2S had not been so successful. Concentration offered some protection for the raiders, but the Luftwaffe had adapted to it and the Command was in need of a new form of protection to reduce losses and at the same time give the attackers an opportunity to bomb more accurately. It was hoped that Window would provide this additional protection, and that improvements to H2S and its employment would make the entire campaign more efficient and effective. While, as shall be seen, results were uneven and losses persisted, the Battle of Hamburg was a time of relative efficiency. The Battle of Hamburg, like the Battle of the Ruhr that preceded it, included raids that ranged widely across Germany, with forays into France and Italy as well, but was named for the most remarkable aspect of this phase of the campaign, the destruction of the north German port city of Hamburg. During this part of the bombing offensive, which lasted from 23 July until 18 November, 33 major raids were conducted; 14 of these included more than 600 sorties while only 4 had less than 300. Of a total of 17,021 sorties dispatched there was a loss rate of 4.1 per cent and a total casualty rate of 10.7 per cent. During the preceding Battle of the Ruhr the corresponding figures had been 4.7 and 16.2 per cent. Credit for these improvements went to Window, which ‘obviously played an important part,’ but there were other differences as well: more dispersion between targets; lighter defences; deeper penetrations; lower damage rate. This last change was particularly attributed to Window, which rendered the flak (the major cause of damage) less effective. Window

140 The Science of Bombing

also prevented the night fighters from getting at the bombers during the entirety of the deep raids, thereby reducing losses.1 The first attack on Hamburg took place on the night of 24 July 1943. Seven hundred and ninety-one sorties were involved made up of 347 Lancasters, 246 Halifaxes, 125 Stirlings, and 73 Wellingtons. The pathfinders, using H2S, marked the coastline with yellow TIs; they also sent back wind observations so that an updated wind value could then be broadcast to the main force, which was operating without any navigation aid. The entire force was instructed to start dropping Window when crossing 7°30s East and to continue to do so until passing 7° East on the homeward leg. Two H2S-equipped pathfinders were to mark and illuminate the target area blindly so that a second wave of pathfinders could visually mark the aiming point with red TIs at Zero hour. Some 50 backers up were to use green TIs from Zero plus 2 until Zero plus 48 minutes, at which time all main force aircraft were have to completed their attacks. All backers up were instructed to overshoot the aiming point by two seconds to eliminate the creepback which had been noticed in earlier raids. ‘Thus,’ it was officially recorded, ‘were the lessons of the past applied.’2 The bombing results for a raid beyond Oboe range were good, although there was significant creepback and some difficulties with the visual marking. In all, 306 sorties were plotted within a three-mile radius of the aiming point based on their bomb release photos. Losses were very light, with just 12 aircraft missing and 31 damaged. As a result of Window blinding enemy radar, fighters had been ineffective and searchlights had been seen to wander the skies without direction. Bomber Command returned to the city on 27 and 29 July and 2 August. The results on these raids varied, owing to problems with weather, equipment, and tactics, but in the end the city was to for all intents and purposes destroyed. The second of these raids, no different in concept or conduct from the others, led, because of weather conditions, to a firestorm which killed tens of thousands and destroyed a considerable portion of the city.3 Yet even as the raids continued, the Germans reacted quickly and made steady progress adapting to the jamming by Window. By 29 July they were starting to use a ‘running commentary’ to broadcast to the fighters the general progress of the bombers that could still be detected despite the degradation of ground-based radar. As a result, loss rates crept up towards the 4 per cent level. Interspersed with these raids, other targets had been hit. Essen was attacked on 25 July. Window neutralized the defenders and this, plus highly effective Oboe marking, allowed 368 of the 604 crews who claimed attack to bomb within three miles of the aiming point. The destruction

Hamburg: July–October 1943 141

was severe for the cost of 23 lost aircraft and another 63 damaged. Five days later, on 30 July, a force of just 273 heavies attacked Remscheid with similar success. The Command had never before operated with ‘greater efficiency ... and the result was not unlike that which might have been expected to follow the despatch of five or six hundred bombers.’4 While not precision bombing, it showed that effective targeting was not just a dream. Yet, of the many attacks to follow, there were only four which could be termed complete successes. On these few occasions, such as the attack on Kassel on 22 October, the H2S blind marking and illumination was effective, allowing precise visual marking and very accurate bombing. At the other end of the spectrum three major raids against Berlin, mounted at the end of August, showed the limitations of tactics and technology. Of 1719 sorties only 27 could be plotted within three miles of the aiming points. H2S had proven ineffective against the mass of the capital; radar screens had shown solid white returns. And the city’s defences had contributed to a loss rate of 7.2 per cent of the raiders with another 6.8 per cent damaged.5 In early July the scientists were already making the point that to get the most out of Window it would be necessary for raids to remain concentrated, and for this reason at the end of August they began a series of studies of selected raids for the purpose of analysing navigation and time keeping ‘in order to provide information as to the general degree of correspondence between the general plan and its practical fulfillment.’6 Each review would contribute to the body of knowledge that could and would make subsequent raids and battles more efficient and effective. The first raid of the series was a 575-sortie attack on Nuremberg on 27/28 August. A detailed study of navigators’ logs indicated that the raid had gone particularly well from a navigation point of view, this the result of a very accurate forecast of winds, with 87 per cent of the force remaining within 10 miles of track on the outward route and the same percentage again plotted within 10 miles of the planned route for the return. The scientists found this interesting, as the ‘scatter’ on the homeward leg was usually greater, but concluded that this concentration had resulted from the use of route markers dropped by the pathfinders, these aiding in setting course for home. They were, however, somewhat puzzled by the fact that on both legs the majority of navigators’ pinpoints were about five miles north of track. For the homeward leg they believed that the reason might lie in pilots not keeping to the planned turn west towards England, given that they were just exiting the target area and would have required a couple of minutes to get established on the return heading. Time on target had not been particularly accurate. Only 60 per

142 The Science of Bombing

cent of the raid had bombed in the allocated time slots, although 90 per cent of the first wave had been on time. On the homeward route time management was well off, and the five waves became thoroughly mixed. To address these issues the boffins recommended a short initial leg at 90° to the homeward track and the maintenance of a specific speed so that the natural tendency to get away from the enemy would not lead to later bunching up over landing fields.7 A late September raid against Hannover was the subject of the next detailed review. This four-wave 698-sortie attack had not gone as well. Navigation was off with only about 70 per cent of the aircraft staying within 10 miles of track. This lapse was attributed to a decision not to use route markers and a significant difference (20–30 mph) between the forecast and actual winds. The latter problem was compounded because there was no common wind value used by all groups and because there was no common procedure for adjusting navigation if the actual wind values were found to be different from those forecast. Despite these problems, bombing, which was also affected by the wind, was reasonably accurate in terms of time. The recommendations, which had been worked up in conjunction with the Navigation staff, underlined the need to develop a common procedure of adjusting for winds. The scientists believed ‘it is most important that Met. should make only one W/V [wind direction and velocity] forecast for the whole Command.’8 Unfathomable as these problems seem, they are very plausible. Often squadrons in the same group can have different ways of dealing with similar conditions. Each practice is effective, but not at all the same. However, the failure to publish and use a common wind, when trying to get 700 aircraft to the same place at the same time, does, on the other hand, defy reason. The third raid to be studied was the 22/23 October attack on Kassel involving 559 sorties, and the scientists once more zeroed in on the question of winds: In addition to the usual general objectives, this analysis was undertaken with a particular view to ascertaining the relative accuracy of winds forecast, found and used, and the proportion of occasions on which forecast, found and compromise winds were used, in order to assist Command in deciding what steps should be taken to improve wind forecasting arrangements. The Kassel raid was chosen for this investigation because, being a highly successful and concentrated attack, it was in contrast to the previous raid considered, Hannover 22/23 September, which was scattered.

Hamburg: July–October 1943 143

While the outward leg had suffered from poor weather, with only 80 per cent of sorties being plotted within 10 miles of track, the return journey had been excellent. This was attributed to four factors: the accuracy of bombing, which meant that almost all aircraft started for home from the same location; the use of route markers to indicate the first turning point of the return journey; the good weather between the target and that first turn; and the accuracy of the Met prediction of winds for this leg of the operation. The boffins noted that three wind values had been used: the ‘found’ or actual winds, the ‘Met’ or forecast winds, and a ‘compromise’ or average wind. Navigators were about evenly divided in their use of the three, but there had been a tendency for reliance on the Met winds as the raid developed. Analysis showed that the Met winds had been more accurate than the found winds, meaning that the navigation got better as the raid progressed; moreover, ‘this result is the reverse of that noted on the Hannover raid, the contrast being due to the accuracy of the Met. W/V on this occasion compared with the comparatively large error on that.’ In terms of bombing itself, the report noted that 90 per cent of the PFF sorties had been within two minutes of planned time and 80 per cent of the main force met the same criterion; the four waves had arrived relatively intact and so the rate of attack remained generally at the desired thirty aircraft per minute, unlike the two previous raids.9 On the day of the Kassel operation Dickins’s staff forwarded to him a report covering a PFF raid against Düren for which the bombing aid G-H had been used.10 The new system, which had the precision of Oboe and also the ability to be used by approximately one hundred aircraft at a time, had achieved an accuracy of 600 yards using a ‘release point’ method which employed the Met forecast wind. These were very good results, but it was noted that an error of just 350 yards could have been achieved using a ‘warning point’ method which was based on the actual or ‘found’ winds at the target. It had always been intended, the note reminded Dickins, that in future operations the main force would employ the second method. The implications were obvious and the need for accurate wind data was reinforced. The first large-scale raid using G-H took place just days later and Dickins passed the analysis to Saundby on 7 November. The accuracy had been very good and Dickins said that, once a few minor bugs were worked out, ‘we can expect practically all aircraft to drop their bombs within one mile of the A.P. At least 50% should be within ½ mile. G-H attacks on all targets within coverage, irrespective of the method used by the main force would appear to be justifiable…’ The

144 The Science of Bombing

report itself said that G-H was without question more accurate than H2S and Oboe skymarking and that it was more effective than Oboe groundmarking attacks except where the latter experienced no creepback.11 Finding and hitting the target with the optimal load was certainly a major focus of the Headquarters and the ORS during these months, but the period also saw first a decline and then a sharp increase in losses. While it might be acceptable to live with bombing accuracies of 50 per cent, or sometimes less, to lose 4 or 5 per cent or more of the force each time out was a statistic that no commander was prepared to accept. Harris and the staff therefore continued to seek ways of safeguarding the very crews and aircraft without which Bomber Command was powerless. Window, long awaited, was one answer.12 The first combined analysis of the Hamburg raids was contained in the Monthly Losses and Interceptions Report for July 1943. As mentioned earlier, the boffins’ monthly summaries had been discontinued at the end of 1942, but it had become apparent that these documents were of sufficient use and interest to reinstate them. This particular report was unique, as it included the period during which Window had first been used and also referred to the devastating raids on Hamburg. However, because of the gap since the previous year, the authors also offered a synopsis for the first half of 1943. In reviewing losses by aircraft type they had found among other things that the new Lancaster II had, at least for the initial period, a lower loss rate than other variants of the aircraft. More significant was the fact that losses for the Halifax averaged 5.0 per cent between April and July while losses for all types of Lancaster were much closer to 3.5 per cent. The main discussion of the report centred on a review of night fighter opposition with regard to the introduction of Window: ‘over the whole month the percentage of interceptions reported was only 8.4%, a very substantial decrease from the 12.5% in June, while attacks also fell suddenly from 3.2% to 1.9%. In July of last year total interceptions were 9.6% and attacks 2.9%.’ The scientists also noted that while most attacks continued to be made during the homeward journey, it was thought that there had been several instances of considerable fighter activity over the target. A return to fighter-searchlight cooperation at the end of July was seen to be the result of the introduction of Window, and, finally, it was noted that roughly 70 per cent of all first attacks by fighters had come from either astern or the rear quarters. The ORS believed that Window had had a very noticeable effect on flak defences. Flak damage had been set at 15.1 per cent on several earlier raids during the year, but after 24

Hamburg: July–October 1943 145

July flak-damaged aircraft made up only 4.9 per cent of sorties. The scientists noted that for the Hamburg attacks ‘of 2360 aircraft which attacked only 78 (3.3%) were hit by flak.’ Additionally, during the first attacks on Hamburg only barrage flak, a form of flak engagement normally associated with an absence of radar direction, had been experienced, and even this fire had dropped off as the raid progressed. Searchlights, normally radar-directed, were judged to have been largely ineffective over the city. The attack the next night on Essen had seen similar experiences with flak and searchlights. Overall, concluded the boffins: ‘in bombing raids on German targets in July approximately two thirds of the casualties were caused by fighters and one third by flak.’13 Thus a valuable ORS information tool returned at the very moment when there were important things to report. It had been particularly vital to describe the success of Window, but no less critical to signal the divergence of loss rates between Halifaxes and Lancasters. On 20 August Dickins passed Saundby a separate detailed report on Window’s first eight uses.14 From a review of the four attacks against Hamburg it was possible to see improvements in all categories being compared: missing (2.8 vice 6.1 per cent); fighter attacks (1.3 vice 2.6 per cent); flak damage (3.9 vice 8.4 per cent); and fighter damage (0.5 vice 1.3 per cent). Losses in the Ruhr were also down. There was only one exception, and here it was thought that Window had not extended over one wave of Stirlings that had suffered heavily as a result. Not only was flak damage down generally, but the extent of the damage was less – it was again believed that the enemy was using the less effective barrage fire. As importantly, the report included data on the German reaction to the countermeasure. It was noted that while during the first raid the enemy had more or less given up, evidence from the second and subsequent raids indicated their controllers had resorted to a running commentary of the progress of the bomber stream. It also seemed from intelligence intercepts that airborne (AI) radars in enemy fighters were still working fairly well. In drawing some general lessons the boffins concluded that losses and damage from enemy action were down as much as one-third and that the enemy air defence system had been degraded. In terms of exploiting Window coverage it was recommended that no aircraft fly above the maximum set height and that Stirling aircraft not follow immediately behind Lancaster waves as there would be ineffective cover for the former because of their lower altitude. The organization and sequencing of waves would need to be reviewed to optimize Window coverage.15 While it was clear that Window was working as intended, by

146 The Science of Bombing

pointing out German reaction the scientists were offering the Air staff key insights for the future. The Luftwaffe was indeed adapting quickly and August raids against Berlin indicated two alarming trends: the bulk of the losses had been against the Stirling and Halifax sorties; and fighters had been by far the chief threat. The night of 23 August gave an indication of the concern over losses. While 5.4 per cent of the Lancasters had been shot down, Halifax losses were 8.8 per cent and Stirling missing rate was set at 12.9 per cent. Then on 31 August 3.3 per cent of the Lancasters failed to return while the Halifax and Stirling figures were 11.4 and 11.3 per cent respectively. Halifax casualties had been on the boffins’ list of action items for some time. Harris himself had been particularly concerned about the performance of No 6 (RCAF) Group, formed at the beginning of 1943, which was flying the type; in the late spring he directed the ORS to examine the matter. Their first report indicated that the Canadians were not doing as well as their sister Group, No 4, which also flew Halifaxes. They pointed to problems with tactics, maintenance, training, and morale, but did not provide conclusive evidence. The Canadian AOC, A/V/M Brookes, maintained that the problems stemmed from lack of experience, since almost all of his crews had done fewer than ten trips. When Harris doubted the results of the first study, he asked for two more, in August and October, both of which ‘largely confirmed’ the Canadian view. A parallel study conducted by a newly formed ORS in RCAF Overseas HQ produced similar conclusions in the fall.16 In fact it was on 19 July, before the battle of Hamburg had begun, that Dickins had provided Harris with details about 6 Group performance. While the available data did not permit ‘firm conclusions,’ it did point to some questions for further investigation. When compared to 4 Group, 6 Group Halifax losses had been higher throughout the year (the averages for February to May were 5.0 and 6.5 per cent respectively) and now that most raids were against German targets Canadian losses continued to climb. It seemed that part of the problem resulted from missed opportunities for training sorties against relatively lightly defended targets in France. It was also noticed that 6 Group experienced more fighter attacks than 4 Group (3.9 to 2.4 per cent) and this might be due to faulty tactics or the fact that 6 Group’s aircraft spent less time in the main stream due to heading to and from the target directly from their northern bases. Growing losses were accompanied by growing aborted sorties. Both might be due to a ‘lower standard of training.’ Indeed, the lack of fresher trips against easier French targets and a failure to realize the necessity of extensive training might explain the difference in losses

Hamburg: July–October 1943 147

compared to 4 Group. Harris and Saundby approved the report and the D/CinC asked Dickins to produce a covering letter for transmission to 6 Group. The salient elements of the study were stated in the letter and both were on their way to the Canadians on 17 July.17 By 26 July Dickins was able to report that he had personally visited 6 Group and was sending them one of his senior staff to assist in additional analysis (in a separate letter to the AOC, A/V/M Brookes, he confirmed his hope to shortly thereafter attach a scientist permanently to the Group).18 Dickins noted that it did appear the Group’s fresher crews were suffering about twice the loss rate of those which had completed more than five sorties. This seemed related to the Canadians’ misunderstanding of the Command’s instructions for new crews, which Dickins confirmed that Saundby had now discussed with the AOC. Unlike 5 Group, it seemed the Canadians were far more happy to have advice from the scientists. While the Canadians were facing particular issues, the Halifax itself remained a concern, and as part of his section’s review of the aircraft, Dickins had made a personal visit to the manufacturer, Handley Page, on 1 July. Four issues had been raised with Sir Frederick Handley Page: the Lancaster was outperforming the Halifax in speed and operational ceiling; the Halifax had a reputation of being more difficult to fly than the Lancaster; the ‘known rudder-stall problem of the Halifax had possibly caused some casualties directly and others indirectly owing to induced lack of confidence in the machine’; and the Halifax suffered from poor exhaust flame dampening. Dickins reported on the director’s rebuttals, most of which referred to the improved performance that could be expected from the new Halifax III, expected to enter service in the fall. Dickins concluded his memo with two points. I was left with a feeling that the Handley Page organization has said its last word in the Halifax III. We should therefore press on with consideration of the best results obtained with this type in order to see if they promise that the machine will meet future operational requirements. If they do not, then we know we have little more to expect from the Halifax and must act accordingly. With regard to the new fin and rudder I feel that we should immediately attempt to get the trial installation machine at present under test at Boscombe Down flight tested in an operational squadron. If the tests promise well, the greatest possible pressure should be brought to bear in order to speed up the introduction of this modification. Otherwise I fear that the promised introduction in August may be somewhat deferred.19

148 The Science of Bombing

Dickins had sent his comments to the D/CinC and the CEngO, A/ Cdre Roach, who for once was in general agreement with Dickins. The latter was able to expand on some of the scientist’s observations from his own visit to the firm, but overall felt that the boffin had been given ‘exactly the sort of “guff” that one would expect a stranger to the Handley Page firm to get from that firm, and I don’t think I can answer the minute better than by using the C.-in-C.’s own words: “I don’t believe one word of the firm’s statement, and I have no faith in any promises made.”’ Roach was able to add that a prototype of the Halifax III was being flown by operational crews on an evaluation basis, but did not have any information on handling and performance. On the matter of modification kits for the fin and rudder, apparently five hundred sets had been ordered, but these would take approximately eighty man-hours each to install and this would put a huge engineering bill on the units. He confirmed that Halifax IIIs would enter production in October at a rate of about seventy aircraft per month and that by April 1944 only this variant would be produced by the firm and its subcontractors. Roach closed his memorandum on a bit of a tangent, suggesting that the Lancaster II was by far the superior aircraft and that Bomber Command should push hard to acquire these models.20 At the same time that he minuted Roach, Dickins had taken the opportunity to relate his misgivings to Air Marshal R.S. Sorley, the Chief of Research and Development (CRD) at the MAP, who had organized Dickins’s visit to Handley Page in first place. The scientist stated many of the same concerns that he had expressed to the company. While he made the point that better overall performance was necessary, so that the Halifax would not be the lowest and slowest of the raiders, he was most insistent with regard to the tail modifications. Dickins pressed for a ‘crash programme’ of two squadrons’ worth of kits without delay so that there would be a large enough sample available to confirm the predicted improvements in handling before considerable time and energy was expended in producing (and installing) the hundreds of kits actually needed across the Command. As he expected that there might be resistance from the firm, he asked Sorley for help: ‘In any case, the greatest pressure which you can put on Handley Page to introduce the new rudder and to provide a number of sets for retrospective fitments will be of the greatest value to us.’21 Clearly Dickins understood the root of the problem and equally clearly he had taken responsibility for pushing the bureaucracy for a solution. Several weeks later Dickins, still no doubt thinking about the Hali-

Hamburg: July–October 1943 149

fax, sent a short communication to Saundby expressing his thoughts on the related ‘Bomber Command Expansion Programme.’ While agreeing generally with the concept of a more homogeneous force based on the Lancaster, the scientists felt that the Air Staff were basing some of their conclusions on a false interpretation of the data. For example, the boffins noted that the difference in efficiency between Lancasters and Halifaxes was based on data from just August and was therefore ‘not at first sight truly representative.’ They wondered rhetorically why the ratio of bombs dropped per aircraft lost had been so much higher in August than in the preceding several months (3.4 versus 2.4). The answer, they opined, was that, in August, eleven of the fourteen main ops had been long-range missions where the Lancaster profited from its higher bomb load. The ORS had done a brief study of data when the Lancasters operated alone and had found the losses to be much higher than in split raids, although admittedly these statistics were based on small attacks. 22 It is important here to recognize the scientists’ desire to give the senior leadership the best advice available. While they too had reservations about the Halifax, their concern in preventing the skewed results of August from forming the basis for a significant long-term decision was indicative of their understanding of the central issues within the Command and their commitment to make a meaningful contribution to the success of the campaign. A number of long-term studies were done by central agencies over the winter of 1943/1944 in an attempt to quantify the efficiency of the Mosquito in comparison to the Lancaster and other heavies. Perhaps not surprisingly, the smaller aircraft was relatively economical to operate. But the statistics provided in October, shown below, demonstrated that, for operations over Germany during the summer (1 June to 15 September), the Lancaster stood out as the clear leader.23 Aircraft

Casualty rate

Bomb load/sortie

Bombs dropped/missing aircraft

Lancaster Halifax Mosquito

3.5% 5.4% 2.3%

3.95 tons 2.20 tons 0.68 tons

112.6 tons 45.4 tons 29.8 tons

A further study was conducted by the Air Ministry and published in April 1944. It showed that the Lancaster ‘cost’ twenty man months per ton of bombs dropped, the Halifax sixty and the Mosquito (with a 4000 lb bomb load) about twenty. While all of these figures were intended to show the relative value of the Mosquito, they could not but have helped to convince the leaders of the futility of maintaining the Halifax as a front-line bomber.24

150 The Science of Bombing

Related to aircraft losses was the question of personnel losses, and in September Harris passed casualty statistics from 1 Group to Dickins asking for comments. By the end of the month Dickins was able to report back that the file had been examined and that two significant points had emerged. Results had been compared against a similar period from 1942 and a much lower rate of missing (crews who had become prisoners) was now apparent. More significantly, this change seemed to be associated with the Group’s conversion to Lancasters. This phenomenon had now led the scientists to begin collecting data on the ‘fate’ of missing crew members both by type of aircraft and crew position. The hope was that this information might assist in pointing to possible improvements in crew escape arrangements. The second point that Dickins wished to underline was that the number of escapees seemed to be on the decline. He posited that this could be the result of the German occupation of the south of France and the fact that the Command was flying more ops deep into Germany. 25 While this second trend might have been interesting, the first was critical, and Harris immediately asked Saundby to ‘write officially drawing attention to urgent need of improvement in escape hatches etc. Far too few crews get away with it and bad hatches etc. are the reason.’ Saundby in turn asked Dickins for ‘some details – e.g. what [aircraft] requires attention and what suggestions we have for improvement etc.’ Dickins indicated that he would accelerate work on these issues but that it would be ‘2 or 3’ weeks before he had the answers.26 This began an intensive investigation into the reasons for extremely high crew losses in the Lancaster and similar though less critical issues in the Halifax. Historiographically this activity is significant as it clearly refutes the contention of OR scientist Freeman Dyson, discussed earlier, that Harris, Dickins, and the leadership had attempted to ignore this significant problem.27 The two or three weeks that the scientists had said they needed became close to four months and it was not until the end of January 1944 that the issues had been defined. In submitting the detailed study, the scientists reported that there was ‘no doubt that the chances of survival in a Lancaster are significantly lower than in a Halifax and this is thought to be due primarily to the more restricted space within the aircraft and to the poor rear escape hatch.’ Work needed to be done as soon as possible to correct problems with jamming of all escape hatches. Equally, bail-out training needed improvement. In truth the statistics were not pleasant. Survival rates for the Lancaster were just 10.9 per cent; the figure for the Halifax was 29.4 per cent. Pilots and gunners had the least chance of getting out. The report also pointed out that the Lancaster, compared to

Hamburg: July–October 1943 151

the more strongly constructed Halifax, had a greater tendency to break up in flight, and this in turn contributed to the fact that in 68 per cent of the cases of Lancasters being shot down there were no survivors.28 The report precipitated staff action in several areas. First off the mark was the G/C T1 [Training 1], A/Cdre Capel’s chief assistant, who sent out new directives on training for abandoning aircraft. He thought, however, that the scientists’ full recommendations which included practice parachute jumps were more than the training units could absorb in time and effort, but did agree that crews needed to be more adept at hooking on their parachutes and then bailing out. He was much in favour of the boffins’ suggestion that pilots needed a parachute that they could wear at all times and that a seat- or back-pack model should be instituted. In sending these comments forward to the D/CinC, Capel underscored the significant problem faced in the air, which was simply that when bailing out during a raid crews were not exiting an aircraft that was flying ‘on an even keel.’ Moreover, there was normally some major problem like fire or damage to make getting through the restricted space and tricky escape hatches that much more awkward. He did accept that more training was needed, but thought that controlled practice jumps from aircraft or balloons had little to offer. In fact he said the ORS data showed that crews already had good success when they abandoned aircraft that were still under control as happened for incidents of fuel shortage or undercarriage failure. After further staff discussion the D/SASO, A/Cdre H.A. Constantine, indicated that he would issue direction to the Groups on the matter of parachute drills.29 At the same time the CEngO entered the discussion with some thoughts on in-flight fires and escape hatches. Regarding the former, he felt that changes then being made to include fire warning lights in the engine instruments would give crews a better chance of abandoning the aircraft before a fire got out of hand. As far as escape hatches were concerned, he agreed strongly that hatches needed to open outwards, but felt that making modifications to existing aircraft was problematic. If there was a real desire to adopt outward-opening hatches then these changes should be instituted at once for the Lancaster and Halifax Mk VIs which were currently under development.30 It seems apparent from Harris’s initial question to the boffins and his reaction to Dickins’s first response that he knew something was not right, and it is clear too that the scientists had zeroed in on a number of key issues and catalysed staff action to sort out a significant loss of personnel, regardless of whether they could be saved for future ops or just kept

152 The Science of Bombing

from perishing. It is true that this work took several months to gel and that there were no easy solutions, given operational and technical constraints, but, whatever obfuscation Dyson recounted in his writing, there was little trace of it in these extensive investigations and actions. Many aircraft did survive enemy action, but not without damage, and the issue of getting these aircraft home had been a theme of importance. At the end of September, Dickins and Mr E.A. Lovell (responsible for vulnerability research) met with staff from the Structural and Mechanical Engineering Department of the Royal Aircraft Establishment to review self-sealing fuel and oil tanks for the Lancaster. One hundred of the bombers were due to undergo a full-scale trial using the ‘Bransom’ tank system, which would add a thousand pounds to the weight of the aircraft. The scientists immediately advised the RAE staff that this was an unacceptable weight penalty, but discussion continued and soon the real issue emerged. It turned out that the main wing root tank (the tank located where the wing and fuselage join together) of the Lancaster was much larger than similar tanks for the other heavies. It was ‘obvious,’ said the minutes of the meeting, that any fuel storage system should be such that the loss of one or more tanks would not make it impossible for an aircraft to get home. ‘It appeared that, under present conditions, the Lancaster root tanks are very much too large.’ It was also important that the bi-directional valves between tanks be such that fuel would not be lost from a good tank because the fuel could continue to flow towards one which had been damaged. More work would be needed and both organizations would attempt to determine how best the tanks should be subdivided. Damage to oil tanks was also discussed. The statistics which might support the need for self-sealing covers on the oil reservoirs were so insignificant that it was decided that these covers could be replaced with armour plating, which was deemed more efficacious. At the end of the year Lovell was involved in similar discussions regarding the use of self-sealing fuel lines. Here too the scientist provided data which convinced staff from the Directorate of Operational Requirements at the Air Ministry that these lines could be deleted without any adverse effects. 31 While work on specific causes of losses occupied much effort, the boffins also tracked broad trends and thus as the Battle of Hamburg drew to a close in November the ORS 2 staff under Dr Smeed were able to produce an analysis of losses by squadron for the period May–September 1943. It had taken some time to do this analysis because of the statistical challenge when working with relatively small sortie numbers which could be unduly skewed by chance events. In passing the file to Dick-

Hamburg: July–October 1943 153

ins, Smeed made a number of observations, including the fact that 434 Squadron (RCAF) seemed to have had an ‘outstandingly high loss rate [but] that the greater part of the other fluctuations can be explained by random chance.’ It was also was noted that ‘non random influences are conspicuously high among Halifax sqdns.’32 Whatever the direct technical causes, most losses were now attributable to fighters, and this trend suggested that the German defenders, in using a running commentary to direct the fighters to the bomber stream, had found a suitable tactic to counter Window. It was evident that there was no place for satisfaction. Indeed, in August the ORC had issued a revised version of their ‘Tactical Countermeasures to Combat Enemy Night Fighter, A.A. Searchlight and Gun Defences Report.’ The new document, like previous versions, provided an extensive description of the enemy’s existing defences, while reminding the reader that the Luftwaffe’s defensive tactics and equipment were constantly evolving. Concentration was again recommended as the best preventative against all types of defence. A density of ‘at least 1000’ aircraft per hour was suggested – with the duration of an average attack recommended to last no more than twenty minutes where possible. Aircraft should approach on a narrow front, of not more than four to five miles, to counter the more prevalent risk of GCI fighters, but should at the same time be distributed vertically over about 5000 feet. Proposed action by individual aircraft included using high speed and altitude to pass through the target area as quickly as possible. Evasive action to escape flak was to take the form of a deliberate turn away from the threat. The ‘extended corkscrew’ was the recommended defence against fighters and was intended as an active ‘combat manoeuvre’ where the gunners would lay down fire in conjunction with the pilot’s manoeuvring. The corkscrew is often referred to in histories of Bomber Command, but not often described. The report provided a detailed explanation of the process: ‘The extended corkscrew commences with a straight dive which is converted into a turn of about 60°, losing height by about 1800 ft. This is followed by pulling out of the dive and climbing as sharply as possible in the opposite direction with full power, regaining as much height as possible. The whole manoeuvre is then repeated exactly as before.’ The tactic, it was stated, made deflection shooting by the fighter difficult and it was also likely that the attacker would overtake the bomber as it slowed down in its climb. Alternatively the bomber could conduct a steep diving turn towards the darkest part of the sky. Weaving was not recommended as it did little to throw off the interception.33 Many of the defensive tactics already proposed by

154 The Science of Bombing

the scientists at High Wycombe were thus recorded for broad distribution and reference. While the boffins reported major trends they also kept up their study of individual raids and there can be no doubt that their analyses of the losses suffered during each operation were seen and studied by the seniors. Their review of the Berlin raid of 23/24 August 1943 is a prime example. The scientists had completed their initial report within fourtyeight hours and forwarded it to Saundby on 27 August. Dickins drew attention to the success of the German fighters and to comments from Fighter Command which suggested that the German tactics were similar to the British ‘fighter night’ interceptions using moonlight. He also commented on the question of enroute dispersion and on the use of Window. In the report itself, losses to flak were linked to those aircraft which had strayed over major anti-aircraft defended areas around cities. Of significant importance, the authors pointed to ‘the evident determination to have fighters to protect Berlin at the earliest indication that the attack was heading for the city. It may be possible to make use of this when attacks are made on targets on or near possible routes to Berlin.’34 Both Saundby and Harris signed off on the document within seventy-two hours,35 and while they made no particular comments it would not have escaped their attention that, while the raid had been costly, the boffins had extracted an important lesson which might well save lives and aircraft by deceiving the Luftwaffe in the future. Continuing to track the effectiveness of German defences in September, the ORS provided an update on the Luftwaffe’s reaction to Window and the effect it had had on their ground-controlled interception (GCI) operations throughout August. It was found that while the number of fighters scrambled was about the same, the loss rate to the raiders was down significantly. GCI sorties had been reduced in number but had not been done away with altogether. At the same time the running commentary and free-lance fighter tactics were evolving and it was now possible to observe that the controllers were attempting to send the fighters towards the presumed target rather than simply into the stream. It also appeared that fighters using AI radar were having some success against Window, particularly at lower altitudes. The report concluded by recommending increased concentration in time over the target area as the best means to overload the free-lance fighters that were now congregating there. An increased use of Window in the target area would also be advantageous in degrading AI radars. 36 If Window could not provide comprehensive safety for the entire raid,

Hamburg: July–October 1943 155

then there were other systems which could help. In August a meeting, attended by Dickins, Smeed, and Mr L.F. Lammerton, the ORS scientist responsible for radio countermeasures, was held to work out the introduction and procedures to be used for Airborne Cigar, an aircraftmounted system designed to jam certain frequency bands used for fighter direction. Smeed put forward considerations for the number of aircraft required to provide effective coverage of the stream and after discussion a number of decisions were made to ensure that there would be two aircraft at the front and rear of the raid and another spaced along every ten minutes of raid mileage. Lammerton recommended that prior to first use of the system a Cigar aircraft, without the highly classified jammers, be sent over enemy territory in order to develop a sense of the numbers and types of signals that the aircraft would actually face on ops. This generated considerable discussion, but the value of making this investigation was recognized and accepted, details of the flight were established, and it was further decided that the test run would not take place more than two weeks before the expected operational introduction. Continuing on with the need to gather information, Lammerton then asked that, once the system entered service, the operators keep logs of what frequencies were jammed and at what times. This information would prove useful, he said, for future operations. There was some concern that this task would prove a distraction, but it was agreed that the operator should keep a log so long as it did not detract from the actual task of jamming.37 Of course Window and Cigar were designed to protect the entire raid and could only do so much. It was still necessary to give each aircraft some self-protection devices. The ORS provided a comprehensive review of the operational effectiveness of AI Boozer (a warning receiver which alerted the crew that the aircraft was being tracked by an AI radar-equipped fighter) on 14 September. Because the first sets had been put into service before the introduction of Window, it was possible to compare effectiveness both with and without the effect of the jamming aid. It was generally concluded that Boozer-equipped aircraft had not had much lower losses than those without the warning device, but it was thought that this might be associated with the ‘bright night’ conditions during the summer months where visual intercepts could be conducted from outside the range of Boozer. If the enemy switched off his AI set after visual acquisition this could explain the paucity of Boozer indications. Alternatively it was considered possible that Boozer was not set to the precise frequencies of the German AI radars and it was decided to

156 The Science of Bombing

enlarge the frequency band of the set. Subsequent to the introduction of Window it had now been found that there were ‘promising’ improvements in the sets’ ability to pick up AI signals; this had been particularly apparent for Lancaster sorties over the Ruhr and southern Germany. More significantly, it was thought, there had been no reports of attacks on Boozer-equipped aircraft where a Boozer warning had been received; the implication was that Boozer was doing its job in warning crews of potential AI attacks and thus allowing them to take effective countermeasures. The report concluded by suggesting that the planned increase in Boozer range from five to eight thousand feet and the return of dark nights would make the receiver relatively more effective during a period of the year when fighters would be more reliant on radar interception. Equipping the force with Boozer should be a priority, the scientists said, particularly as it was known that Window was degrading GCI radars and making the enemy more reliant on AI.38 The boffins had been equally interested in getting Monica into service and, as discussed in the previous chapter, had been studying the system for some months. First uses of Monica had led to a restriction on its operation; it was directed that the set was to remain switched off until approaching the target, as there had been numerous false signals from friendly aircraft. At the end of June the ORS and Ops staff visited a number of 4 Group squadrons using the device. Based on crew opinion that the set should be used throughout a sortie, Dickins sought approval for lifting the restriction. His logic was deemed sound and Saundby, who had put the limitation in place, cancelled it immediately.39 Approximately a month later, Dickins again sought approval to modify Monica procedures. The boffins noted that Monica had an adjustable range scale and that it would be worth surveying crews to determine the most effective settings under different conditions. The SASO, A/V/M Oxland, gave the go-ahead and the scientists began collecting the data.40 Gathering data had been the scientists’ chief occupation during the preceding months, and while the impact of Window had been one of the great improvements it had not been a miracle cure. More than jamming and warning devices would be needed to safeguard the bombers as targets shifted further east to Berlin. Losses, coupled with the relative limitations of H2S and the extremes of winter flying, would soon put the efficacy of the bombing campaign in question.

10 Science versus Attrition – Berlin: November 1943–March 1944

In mid-November the Command’s focus shifted to its next major target – Berlin. Between 18 November and 24 March 1944, 16 raids, totalling 9111 sorties (7256 Lancasters, 1643 Halifaxes, 162 Mosquitoes, and 50 Stirlings), were dispatched against the city. The German capital was not the sole target during this Battle and some 11,000 additional sorties were sent out against 19 other cities. Harris had made it known that if given 15,000 Lancaster sorties for use against just Berlin he could knock out the city by 1 April 1944 and with that victory would start the collapse of Germany. By the end of the battle the number of Lancaster sorties employed against all targets was not far short of that total, not to mention Halifax and Mosquito sorties, but Germany was in no sense ready to seek peace. In fact, ‘in the operational sense, the Battle of Berlin was more than a failure. It was a defeat.’ The final attack against Nuremberg on 30 March resulted in an 11.8 per cent missing rate, and this ‘brought the Bomber Command tactics of massed and concentrated attack against major targets to a dead stop and they were not again resumed until the entire air situation over Germany had been radically altered.’ None of the innovations and techniques that had been adopted were capable of countering the fighters, which seemed to dominate the night. It was later pointed out that ‘The two most important factors in the comparative failure of the Battle of Berlin were the high casualty rate sustained and the relative inefficiency of attack achieved.’ When both criteria were measured, the results for Berlin ‘compared unfavourably’ with what the Command had achieved in the preceding months. Loss rates had climbed to 5.2 per cent compared to 4.7 in the Ruhr and 4.1 for the Hamburg attacks. Concentration and accuracy of bombing had suffered during this final series of raids, comparing ‘most unfavourably’ with the results from earlier in the year. The underlying causes were seen to be ‘geogra-

158 The Science of Bombing

phy, the German defences and the weather.’1 In hindsight even Harris was prepared to admit to the problems. ‘Judged by the standards of our attacks on Hamburg, the Battle of Berlin did not appear to be an overwhelming success.’2 While the CinC had had some concerns about the ability of H2S to get the bombers to the target, he was still prepared to give the system the benefit of the doubt, and indeed H2S was to be the dominant aid to navigation and bombing during these months.3 Analyses of the device had continued throughout the summer and fall in order to work out the bugs. As the year drew to a close the boffins completed two additional studies in collaboration with the Nav and Radar staffs. The first paper reviewed the blind-bombing capabilities of the system, drawing on all evidence gathered since its introduction, including flight trials and operational use by the PFF. Findings indicated that the probability of error, given the system’s limitations, was 50 per cent. In addition, the boffins concluded that H2S worked better with some targets than others and that towns of similar shapes could be easily confused.4 The second report, dealing with navigation, found that if used properly the equipment was quite effective: 93 per cent of coastal fixes and 80 per cent of fixes made on inland towns were accurate. Inaccuracies on the homeward leg were found to be twice those on the outward leg, but the scientists felt that this could be explained by the difficulties in setting course after the target and also by lapses in navigation. This last point led to a recommendation that it be impressed on navigators that success with H2S was linked to careful dead reckoning, more so than for other navigation aids.5 It was evident that H2S could not be considered in any way as a cure-all for navigation or target-finding problems for targets deep in Germany. Sorting out the accuracy of wind data and how navigators employed the information was also a major challenge to which the boffins continued to give much thought. Some progress had been made by early 1944; wind-finding aircraft were now being tasked to transmit the actual or ‘found’ enroute winds back to Command so that an average found wind value could be calculated and then broadcast to the raid. Such was the concept, but analysis of the raid against Madeburg on 21/22 January revealed that changes were needed. Using information and analysis provided by the ORS, the CNavO issued a gentle rebuke and a detailed procedural instruction on 14 February. Analysis indicated that wind reports had ‘sometimes been outside limits’ and not submitted at specified times. There might also be ‘a lack of appreciation on the part of selected navigators of the importance of individual accuracy and reliability. There

Science versus Attrition: November 1943–March 1944 159

is also evidence to show that some navigators are having difficulty in the speedy application of the past and future winds. ... Any improvement that can be obtained in this respect is most desirable.’ This said, the CNavO went on to define in precise detail how and when winds would be passed from the wind finders to Command HQ and then retransmitted to the main force once an average value had been determined.6 This did not tell the whole story, and the finalized report, ORS B 201 ‘Note on the Standard of Log and Chart keeping and Accuracy in Plotting and Calculations – Madeburg, January 21/22nd 1944,’ was therefore distributed the following month under cover of the CNavO’s signature. In this letter he reminded Groups that attention to navigation and wind finding was only part of the search for optimized concentration. If reliable winds could be found and applied, this would ‘assist in exposing other errors that are preventing the achievement of the standard of concentration that is an outstanding requirement.’ With this as background, the letter went on to identify the key problems that were in need of attention. Log and chart keeping were frequently inaccurate, leading to other errors. Navigation supervisors at the units needed to be diligent in checking the work of navigators and in bringing errors to their attention. Precision in calculations and plotting was similarly lacking and ‘there are few navigators whose work has sufficient accuracy to be classed as reliable navigation and the percentage of errors [shown in the ORS data] can only be due to lack of care.’ With regard to time of arrival over the target the letter stated that on average it took about twice as long to complete bombing as was planned. With better wind data, track keeping had now improved and it followed that ‘time over target’ should also be more reliable. If this was not happening then it seemed that accurate planning before takeoff was simply not being done. Along these lines, while it was possible to adjust the timing of a raid to minimize the time over enemy territory, this could only be done effectively if the whole force was on time. In concluding, the letter reminded Groups that better equipment and processes meant little if those responsible for the actual navigation were failing in their attention to detail. 7 Dickins would later write that these ‘faults proved to be extremely difficult to eradicate, despite the extensive O.R.S. and Navigation Branch campaign directed against them for many months.’8 Following the promulgation of this second letter, some of the scientists apparently felt short-changed. Mr Roberts, the ORS second in command and head of ORS 1, minuted Dickins saying that his staff had intended to put out their own communication, but now saw little need given this

160 The Science of Bombing

new instruction from the Navigation staff. On the other hand, Roberts thought that it might still be worthwhile to produce a modified document if only to ‘represent a record of the work we did and [that] as a result of this work important changes were made.’9 This is the only case found of the scientists descending into office politics and is therefore striking by its uniqueness. In a similar vein, the scientists did discuss issues among themselves, and in early January the topics were bombing accuracy and erratic bombing. New data had been extracted from the Kassel raid of the previous October, as well as from similar experiences at Hannover and Düsseldorf (8/9 October and 3/4 November), since the bombing on all three had been so concentrated and accurate. A report analysing 4 Group’s Halifax sorties showed that almost invariably those aircraft which could not be plotted or had not been plotted within three miles of the aiming point were flown by crews with less than three trips in their logbooks.10 Statements from these same crews, however, all made claims to have bombed on TIs. The simple conclusion of the document was that inaccuracy was a result of inexperience. Mr F.J. Lloyd, the 4 Group ORS officer, and author of the document, recommended more practice with high altitude night bombing and additional analysis of raids to confirm the trend which he had observed. At High Wycombe the report elicited comments from Mr Roberts: I doubt whether freshmen can be expected to do much better. I spoke to SASO 1 Group on this and he said that on his first two trips (Spring 1943) he was completely [taxed] as to what was happening in the target area. In view of the fact that we cannot, however much we try, reproduce T.A. [target area] conditions in this country, I think we must expect bad results from aircrews on their first 3 ops and should concentrate on analysing results of the ‘over 3’s.’ I think SASO and A/Cdre Ops would be interested in Lloyd’s report. ...11

In the same month the Section began a wide-ranging year-long review of night vision and bomb aimers’ capabilities. Their preliminary paper noted that TIs had been generally effective since their introduction, but that there were several problems that remained, some of which should be resolvable using scientific method.12 The report paid much attention to the characteristics of TIs and to the conditions (fire, smoke, dust, and haze) normally found in the target area that often obscured much of the scintillation and colour characteristics of TIs. The authors concluded

Science versus Attrition: November 1943–March 1944 161

that of these two visual characteristics colour was the most noticeable recognition feature. The study also pointed out that bomb aimers generally had difficulty explaining how they would pick out a good TI on which to base their aim; they just did. In the midst of these investigations the boffins came to the startling conclusion that many main force crews simply did not understand how the pathfinding system worked. The Section produced a ‘somewhat popularized account’ of the procedures called ‘Methods of Target Marking’ to at least partially address the problem.13 The authors of the report thus felt that ‘it appears that some knowledge of the airbombers’ psychological and physiological potentialities and performance at discriminating target indicators from the rest of the picture and at assessing the centroid of them taking into account the large inequalities in their individual brightness is vital.’ First, it was suggested that the colour vision and visual acuity of bomb aimers be reviewed. While personnel were currently accepted for training as bomb aimers so long as their colour vision was graded as ‘defective safe,’ it would perhaps be more effective should they be screened for the general standard; regrettably neither term was defined. As well, it was recommended that bomb aimers and in fact all crew members be given a fullscale practical demonstration of TIs from the air before going on ops and that, in addition, crews should be shown slides of target areas with their full panoply of pyrotechnics. Follow-up action resulted in copies of the report being sent to Messieurs Jukes (now with the PFF), Lloyd, and Curry for review by their Groups.14 As well, towards the end of February a meeting was held so that various research tasks could be organized. Of particular relevance were studies of TI effectiveness and of the psychological aspects of picking out the mean point of impact (MPI) from a centroid of burning TIs; the latter work was to be done at Cambridge University.15 By April 1944 the Air Staff were working hard to make use of the scientists’ conclusions and recommendations. A short paper ‘Notes on the Visibility of Target Markers,’ had been prepared based on the ORS document and was ready for issue to OTU instructors and squadron Bombing Leaders, the latter responsible for bomb aimers’ training and performance once they reached their operational squadrons. Using the material, students would ‘have some appreciation of what happens at a target and be impressed with the need for great care in estimating their correct point of aim, while still in the O.T.U. stage.’ It was ‘hoped that additional research will provide material for further notes of this kind.’ 16 The sci-

162 The Science of Bombing

entists had focused on yet another significant problem facing the crews, and the Air staff were happy to make use of their conclusions and recommendations. S/Ldr Booth, of the Training staff, had further ideas in mind by the end of the month. He wished to embark on a major study of bombing error, focusing attention on what was termed the ‘missing link’ between the expected error of about 300 yards from 20,000 feet and the actual errors which ranged between 600 and 1000 yards. Booth wrote that according to the scientists this link appeared to have two elements’: ‘incorrect selection of the aiming point and target not correctly in line of sight,’ which were the result of target area conditions and crew experience and attitude. From this he had distilled two training requirements: instruction to teach how to assess the spread of TIs in order to determine the MPI, and general training in bomb aiming in conjunction with investigation on various sources of error. Much corrective action, such as the ‘Notes on the Visibility of Target Markers,’ was acknowledged, but additional analysis would be required along with additional training in ‘cross country bombing’ and ‘Flashlight’ bombing training exercises. Now Booth sought approval for seconding additional ORS personnel to work specifically on this next phase of analysis. His superior, the number two man in the Training staff, indicated that he was ‘entirely in agreement’ with the whole initiative.17 The inertia for the project was maintained with the release of ORS Report 99, ‘The Visibility and Recognition of Target Indicators,’ at the beginning of July, which called for additional night vision training for bomb aimers with a focus on colour recognition as well as additional intensive training on estimating the MPI.18 Other recommendations in the document requested that the medics review the definitions of normal and defective-safe colour vision to ensure that there were no wide variations between these categories. As well, the MAP should be asked to undertake a study of TIs in order to provide precise information on the characteristics and conditions of their visibility. Underscoring the importance of the recommendations, as a general series report, the document was given wide distribution across the RAF. As the summer of 1944 passed, work on training devices and night vision progressed, particularly at Cambridge, where work was done to develop a series of transparencies which depicted TIs in a target area as a raid developed.19 Further evidence of the extent of the challenge facing the Command in making bomb aimers more effective was seen in the results of the work done in Cambridge during the remainder of 1944. In a short report on

Science versus Attrition: November 1943–March 1944 163

the findings of testing with operational bomb aimers it was stated that there was ‘capricious disregard of certain markers’ and, perhaps worse, that there was no correlation between which markers the various subjects ignored.20 Even when conducting these experiments while on the ground it was found that some subjects made consistent errors of 1000 yards. Overall, the results were deemed sufficient to validate concerns. As a simple and readily implemented solution it was suggested that bomb aimers simply be told ‘to aim for a point such that there are as many coloured markers beyond it, as in front of it, and as many to the left as to the right of it.’ Dickins passed these thoughts to the A/Cdre Ops on 28 November and received the green light to work in collaboration with the Training staff to develop the recommended guide.21 Finally, in January 1945, the training package, developed in large part by the Air Ministry Psychology Unit at Cambridge, was ready for distribution. 22 As the work drew to a close it was impressive to see how the boffins and trainers had worked so closely on this fundamentally important issue during the course of a very critical year. As significantly, what had begun as an internal ORS discussion had resulted in some fundamental changes to bombing training that could go far to improving the efficacy of any and all raids. While this extended research was going on, the scientists had none the less kept the performance of the PFF squarely in view, and at the end of March 1944, in fact on the last day of the Battle of Berlin, had produced an assessment of PFF effectiveness for the proceeding months.23 The report actually focused on the ineffectiveness of H2S in the PFF’s hands and several damning findings were reported. The chief cause of the lack of success of groundmarking attacks is that the M.P.I of the pattern of T.I’s laid by the Blind Markers has not been centred on the A.P. It is thought that the main reason for this is that the H2S crews, lacking confidence in their ability to interpret the P.P.I. response [on the navigator’s H2S set], are being influenced by what they see on the ground, thus replacing the random errors of blind marking by systematic errors, which inevitably result in the displacement of the M.P.I.

Several other problems had been noted. The crews were not taking advantage of the increased definition available in H2S Mk III, which would allow them to refine their navigation. Much time had been wasted in planning for visual marking against Berlin despite the fact that there was little chance of fair weather. As a result the number of secondary blind markers available was reduced, thus compromising the chances

164 The Science of Bombing

of success of a blind attack. Visual markers were releasing TIs without identification of the AP. It was noted that backers up were using the same colour markers as the primary markers and that their bombing was of a low standard. Routes were being planned to avoid cross-wind approaches to the target even when these routes would make H2S accuracy problematic. Finally, the main force was not aiming at the most accurately placed TIs (although their ability to distinguish these particular indicators was doubtful, as has just been seen). A number of recommendations followed. The remedy for most of the faults mentioned above lies in a much more intensive programme of training for Pathfinder crews, particularly in the use of Mk. III H2S, and it is considered that this should be carried through even at the expense of a temporary reduction in operational strength of the Group. Unless a very marked improvement in the general standard of performance of Pathfinder crews can be brought about, there seems little prospect of future attacks being any more successful than those of the past three months. Now that Pathfinders are almost completely equipped with H2S Mk. III equipment, such training programmes should obviously be designed to exploit the peculiar advantage of this equipment. All necessary steps should be taken to ensure the P.F.F. is supplied with the very best crews available in the Command, and inter-Group rivalry should under no circumstances be allowed to interfere with this requirement.

These were in no way tentative observations and as far as military verbiage goes their tone and word choice were about as pointed as one might be likely to see. The scientists were not content with PFF performance and had said so – with no punches pulled. Regardless of the bombing aids and techniques used, part of an overall equation of effectiveness included optimizing the fuel and bomb loads carried. The scientists had tracked bomb load data throughout the fall and at the beginning of November 1943 they prepared a short report on what the Groups were carrying.24 Some anomalies were noted, with different Groups specifying different weights of bombs on the same ops. Investigation pointed to several factors that would influence this situation. For example, the useable load for 3 and 6 Group Lancaster IIs was 1500 lbs less than for the Lancaster Is of 1 and 5 Groups. The simple explanation was that the ‘empty’ or basic weight of a Mark II was 1500 lbs more than a Mark I, and, since the maximum weight was the

Science versus Attrition: November 1943–March 1944 165

same regardless of variant, the difference was accounted for in the lower bomb/fuel load. In addition the Mark II had a higher fuel consumption and (a point not made by the scientists) the 6 Group bases were a greater distance from Germany than those of the other groups, thus requiring a greater fuel load. Examining distances in more detail the boffins observed: In order to estimate the load which an aircraft can carry to a given target it would be desirable to know the air miles travelled [which would include such additional distances as diversions, evasive action and flying around weather] but there is no reliable record of this. Engineers’ log looks are not very reliable. The usual practice has been followed therefore of using track miles (measured up for particular raids from the route map) assuming still air [i.e. no wind] and most economic speed, then finally applying an operational factor to allow for these assumptions and to give a safety reserve of fuel. This factor consisted of reducing the range for a given bomb load by 25%. Discussions with various experts indicate that the bomb load obtained in this manner should constitute a minimum.

The authors had also found that 1 and 5 Groups aircraft were ‘occasionally’ overloaded and that 6 Group aircraft were often overloaded on longer raids where the high fuel requirements would otherwise reduce the bomb load significantly and that this overloading was done with express approval of the Groups’ headquarters. On the other hand many aircraft often flew with less than a full load, especially on short distance ops, because there were insufficient bomb racks to carry the full payload that the weight calculations would permit. Overall it seemed that any deviation from the expected values could be rationally explained, but Dickins and Roberts still felt the need to pass a draft of the report to the Groups before sending it to the D/CinC.25 As a result of this general review, a specific examination of two raids was directed.26 In looking at the 18/19 November attacks against Berlin and Mannheim the boffins now compared the theoretical ‘Command’ values for fuel and bombs against what had actually been carried. They stated that these ‘average’ values could be subject to modification for many reasons: variations in the load compared with aircraft maximum gross weight; uncertainties about the weather where, for example, a lighter weight might allow an aircraft to climb above icing; inexperience of the crew where a new pilot might want a very large fuel reserve; wind direction and length of runway available for takeoff that would limit the

166 The Science of Bombing

safe takeoff of a heavily loaded bomber, as would engine power available for takeoff. While these factors were not thought likely to cause large differences, the actual findings were different. The Command bomb load for the raids had been estimated at 1700 tons but only 1500 tons had been carried. Lesser loads than expected were associated with 1 Group and with units operating Lancaster I and Stirling aircraft. The scientists felt that more work needed to be done to see why individual units and crews were deviating so significantly from the norm. Suggesting future investigations, the report closed noting that a study of a more recent Berlin raid (22/23 November) had been initiated and the preliminary figures seemed more in line with the expected values. Perhaps all that was needed to get the shirkers out of their lethargy was to have the scientists ask a few pointed questions. In early December the Section had the early results of the 22/23 November evaluation.27 With the exception of the Stirlings from 3 Group, normal bomb loads had been carried. The boffins did note that, in some cases, 4 and 6 Group aircraft had had spare capacity which had not been used. It was also noted that as a general trend there seemed to be a tendency to carry more fuel than would be expected for a given distance. Conceivably, they said, the aircraft load and weight tables needed revision. As a result of this suggestion G/C Ops now asked for a detailed investigation of bomb loads.28 A letter was immediately prepared and distributed to the Groups on 1 January 1944 advising them that in an effort to develop new Command load/ range charts the ORS had been tasked to review the matter.29 The Groups were advised that Mr Roberts and his staff would be collecting the necessary data and would be visiting them in the coming days. The extensive report prepared by Roberts’s section (discussion of which covers the following three pages) was ready on 3 February 1944.30 It dealt first with fuel and then bomb loads. Starting from a recognition of a wide variance between the Groups it worked through weight questions from first principles, asking questions like how many rounds per gun should be carried and whether oil tanks needed to be full if the actual technical requirement was something less. It was noted that aircraft of the same type could and did have different empty weights. In terms of fuel carried the boffins underlined the fact that while Group standards varied, all were united in using 200 gallons more for reserves than the Command value. The report stated: ‘The principal allowances which are not covered in the existing Command graphs are those for possible diversion, landing congestion, extra consumption over very heavily defended targets and strong winds.’ Once these factors were integrated it

Science versus Attrition: November 1943–March 1944 167

meant that the actual miles flown per gallon dropped off from 1.0 to as little as 0.78 for the Halifax III. In addition to this revised ‘fuel required’ equation the boffins accepted that there was an additional need for a safety allowance of 200 gallons to cover such likely eventualities as fuel loss due to enemy action, getting lost, and bad engine management where throttles were set incorrectly resulting in high fuel consumption. Because adequate fuel was needed before the bomber could go anywhere and then get back, calculating the available weight for bomb load was left to the end. Here too it had been found that there were no standard values used for various bombs and the ORS said that this needed to be addressed by the CArmO. Shortages of Lancaster small bomb containers (SBCs) used to carry incendiary bombs had also been noted; the problem was such that the proportion of incendiaries on most raids was lower than the specified 70 per cent. To address this issue it was recommended that on raids where Halifaxes accompanied the Lancasters, the former carry only incendiaries, as this was their most efficient bomb load in any case. The report acknowledged that there would always be some inconsistencies but that these recommendations would help address the issues. More follow-up was recommended. Roberts’s report acknowledged the excellent support from the AOCs and their staff but confirmed the inconsistent methods and figures employed by the Groups. While stating that there was room for additional investigation, he believed that sufficient work had been done to propose new load/range figures and make associated recommendations. In general his scientists felt that the existing Command figures did not take into account a number of operational considerations already mentioned – ‘diversions, landing congestion, extra consumption over heavily defended targets and strong winds.’ They therefore recommended that calculations be based on ‘Operational Track Miles per Gallon’ (rather than Air Miles per Gallon or Track Miles per Gallon), since the former included fuel for takeoff, climb, and defensive manoeuvres, as well as the ‘normal difference between air miles and track miles.’ In other words the fuel carried should be sufficient for all eventualities. No specific formula was offered for calculating this value; rather it was thought that the ORS should continue to monitor Group methodologies for the time being. The Section was much more categorical about the Safety Allowance, recommending a standard 200 gallons across the Command (300 for Stirlings), stating that this would clearly differentiate the safety reserve as something other than the variable reserve. Finally, it was recommended that the Minimum Fuel published for a raid should be based on the

168 The Science of Bombing

formula included in the report and that new fuel tables should be developed for each aircraft type. Turning to bomb loads, Roberts noted that the Groups used no standard weights for various bombs and that the Armament staff needed to publish these figures. He noted, too, that given the problems with obtaining sufficient ‘deep SBCs’ many aircraft were taking off with less than maximum bomb loads. As a temporary remedy he ‘recommended that on the first night after a blank night [i.e. a night without ops] Lancasters should carry as many deep S.B.C’s as are needed to make up the permitted weight and that they should not be withheld to cover the possibility of another operation the following night, since the number of occasions on which operations occur on consecutive nights is not sufficiently high to justify any reduction in the efficiency on the first night.’ Finally, he said that where weight became an issue for the Halifaxes they should carry a full load of incendiaries in SBCs rather than leaving incendiaries behind to take a 2000 lb H.C. bomb as was currently stipulated in the ‘Usual Bomb Load.’ Such a revision, he said, would increase the percentage of incendiaries which was always a concern in the Command. While there is no evidence of the application of these recommendations one can see that there was considerable value in the proposals. The Ops staff had been correct in asking the boffins to attempt to normalize both the fuel and bomb carrying policies. In the first case, use of the scientists’ methodology would allow the crews to know that they had sufficient fuel for virtually all circumstances, which would undoubtedly allay some of the nervousness associated with any raid, and, in the second, Roberts’s recommendations would help maximize the tonnage carried on any given night, contributing, as he stated, to the efficiency of the raid and the campaign. November had marked a revision to the format of ORS reports, but this was more than a procedural change. While the scientists had not been averse to providing conclusions and recommendations, Dickins now took his lead from an ORC letter asking that all OR sections ensure that reports included specific conclusions and recommendations. Dickins asked that his officers now write reports with a ‘Summary and Conclusions’ section as the introduction to the report, this distillation of the key material to be followed by detailed discussion.31 In this he was clearly thinking of how the boffins could minimize the task of busy staff officers who would normally be happy to have the answers to problems without having to wade through the scientific proofs. The approach also spoke to the fact that the boffins had the confidence of the flyers to

Science versus Attrition: November 1943–March 1944 169

make their thoughts known, these based on science and logical thinking. In the same month Dickins submitted to Harris and Saundby a summary of the work then ongoing in the Section.32 The table covered fifty-six activities ranging from such things as the analysis of losses and use of Window to the production of training films and simulators. In this they were fulfilling the broad mandate first laid down by Peirse some twentysix months earlier. At the end of 1943 the boffins were ready to comment on the effect of the Halifax’s modified tail assembly, an adaptation needed to improve the handling of the aircraft under combat conditions. A two-page draft study, which concluded that the modified tail had had no reduction on losses, was circulated to the Engineering staff for comment. It precipitated a long and fairly critical response from Wing Commander Smith, Eng 3.33 While attacking the sample size and disproportionate number of unmodified aircraft used in the comparison, his observations included significant thoughts on the flying characteristics of new as compared to ‘seasoned’ aircraft. The former he said were often considered to be more likely to be lost: there were unidentified production defects which led to many teething problems. Engine performance was at its best, but conversely the newly assembled controls were ‘stiffer so manoeuvrability may not be so good’; and it took time for the crew to get a feel for the aircraft, since each had its own peculiarities.34 Smith concluded his observations by asking if it was known whether all modified aircraft had been given to new crews as they were certainly the most in need of a more easily controlled bomber. Finally, he asked if the data could be sorted in order to isolate one or two squadrons so that a comparison could be made based on results obtained from similar operational conditions. Smeed and Dickins reviewed these points and, while Smeed was generally satisfied with the analysis that his staff had done, Dickins was open to Smith’s requests.35 He asked that the report be rewritten to attempt to address the questions as completely as possible. The revamped document was subsequently published as ORS S 114 ‘A Note on the Losses of Halifax Aircraft with Modified Rudders’ on 24 February 1944.36 In it the boffins concluded that the analysis of the data had not pointed to any appreciable reduction in losses. As well they included a qualitative comment to the effect that there was some benefit for inexperienced pilots using the modified aircraft, but a slight disadvantage for experienced crews who were obliged to use ‘“new” as opposed to seasoned aircraft. If this reasoning is correct then a slight diminution in the loss rate may be expected in the future [once the new aircraft have

170 The Science of Bombing

been broken in].’ In the body of the report they made extensive use of the comments provided by Smith. It was also in February that Dickins reported to Harris and other seniors with results of the boffins’ comparison of the Halifax III and the Lancaster III. While the data samples were small, it was possible to show a preliminary difference of 40 per cent lower losses in favour of the Lancaster.37 ‘Very interesting,’ was Harris’s reply, but he then asked to see the following month’s figures when they became available,38 and these Dickins sent to him on 22 March. ‘The Halifax III is sustaining 30% higher loss rate than the Lancaster and having regard to the smaller bomb load of the Halifax the relative usefulness of these two aircraft may be given as 1: 2.6.’39 Harris now used this data to send a ‘Top Secret and Personal’ letter to Air Chief Marshal Sir Wilfred Freeman at the Ministry of Aircraft Production; he sent copies to the Secretary of State and the CAS as well.40 Harris opened the missive: ‘We now have sufficient experience of the Halifax III to make a firm comparison between it and the Lancaster.’ He continued, restating the ORS’s comments word for word before going on to discuss less quantifiable but still critical comparisons. Routinely the Lancasters were forced to provide a shield around the Halifaxes, taking top, front and rear waves of the raid, all with less protection than the Halifaxes in the centre. The Lancasters were also obliged to conform to the Halifaxes’ routing requirements, which were driven to large extent by their range limitations. Finally, said Harris, it could not be forgotten that, with a higher loss rate, the Halifax aircrew supply system was more costly than that of the Lancaster. Harris reminded Freeman and the others of his earlier concerns: ‘On the whole, therefore, it is apparent that my prognostication that the Halifax III would be in the same position by next Autumn as the Stirling and the Halifax II and V are today, shows every indication of coming true.’ There was no doubt in Harris’s mind, or those of the scientists, that the Halifax was not the aircraft that the Lancaster was. While it was up to Harris to fight the political battle with his seniors, he had been armed by the ORS with clear evidence of his doubts about the Handley Page product. Aircraft were just one of Harris’s concerns. His crews, the men that he sent into battle night after night, were another. Loss rates during the Battle of Berlin were now such that Bennett for one felt that crews were backing away from engaging the target, an observation which Harris did not reject. Whether the Pathfinder leader was right or wrong, there were several factors in play that could not have helped: winter weather, for example, could have led to an inability on the part of the crews to

Science versus Attrition: November 1943–March 1944 171

see and bomb markers; higher loss rates (during the Battle totalling more than the overall front line strength of the command) would have led to a diminution of experience and with it bombing effectiveness; finally, there were no new equipment, techniques, or tactics which might have provided a temporary advantage. The only thing that had changed was the size and composition of the force with its preponderance of Lancasters.41 In December 1943 the boffins tabled a report on crews that aborted their sorties within 4 Group.42 Data showed that while the ‘early return’ rate was generally in the 7–8 per cent range, spikes of 15–20 per cent had occurred. The scientists went further, breaking down causes into six categories: aircraft defect or incorrect maintenance; technical problems which could not be duplicated once back at base and faulty navigation; sickness; icing; enemy action; and ‘no reason given.’ Some categories were fully justifiable, while others, especially sickness and technical problems which could not be duplicated once back on the ground, merited close attention, said the scientists. It had also been possible to look at which crews were coming back early and pick out those with multiple aborts; work was being done to see if the problem lay with the aircraft or the crew. With regard to tour completion the report said that credit for sorties should be given to crews who had legitimate mechanical problems, but it was noted that these only amounted to about one-third of the early returns. The scientists were more concerned with the avoidable or questionable aborts and recommended ‘very close investigation on the part of Station and Squadron Commanders.’ While they might have restricted their analysis to the facts, the scientists had shown that they well understood the challenges facing leaders at the sharp end. At the beginning of 1944, at the request of the Ministry, the scientists had prepared a draft report dealing with the related use of body armour. Because the data analysed dealt with casualties, Dickins forwarded the document to Saundby, seeking, as had been the case with similar information a year earlier, a limited distribution.43 The study indicated that there had been 469 casualties among nearly 50,000 sorties for the year ending 1 December 1943; this represented a rate of one casualty for every 105 sorties. The report also looked at the usefulness of American flak jackets and helmets and, based on the number of casualties, concluded that their use would be of no appreciable value. Attempting to include data from missing aircraft, the report used input from crews of aircraft that had failed to return (presumably these were escapees; however, pages of the report are missing and this cannot be confirmed) and

172 The Science of Bombing

‘it has been shown that the figures for casualties per aircraft destroyed among missing aircraft is very similar to that for returning aircraft. Thus the reduction [through the use of flak helmets and vests] in casualties to personnel in missing aircraft will be equally small.’ Saundby made no recorded comment about the report although the last sentence of the above extract appears to have been added in his hand.44 The human element played out in other ways too. In November the ORS issued Report S113, which examined the growing circumstances of bombers firing on suspected rather than confirmed enemy aircraft. It was noted that 5 Group, which had interpreted previous rules of engagement most aggressively, was experiencing a higher rate of attacks by enemy aircraft. Moreover, there had also been some firing against friendly aircraft which could be attributed to a Group policy to engage unknowns. A/V/M Cochrane, AOC of 5 Group, wrote to the Command on 5 February 1944 to reaffirm his Group’s policy and to indicate that he intended, after both a careful review of the report and a follow-on discussion with Dickins, to continue the Group’s practice of engaging unknowns. At High Wycombe, the Ops staff commented that while perhaps 1 Group should be encouraged to be a bit more aggressive in engaging ‘identified’ enemy, by comparison 5 Group ‘seem to shoot with abandon at everything that comes their way and should be impressed that definite identification is most important if we are to avoid self inflicted losses.’ ‘To teach crews that their primary defence is in the use of their guns is wrong.’ Dickins reviewed these comments and said that there was ‘no doubt’ that 5 Group’s engagement policy was behind their higher attack rate. Armed with these conclusions the SASO wrote to Cochrane on 25 February reminding 5 Group that ‘Gunners are to open fire upon every identified enemy aircraft.’ To make sure there was no confusion the letter reiterated that ‘it is therefore most important to avoid the opening burst of fire until the enemy aircraft is identified as such.’ (emphasis in original).45 Here was a case, as usual, of 5 Group disregarding the boffins’ recommendations, but as often happened it was not the scientists who would have to adjust their views, but rather a fairly senior member of that Group. This was not to say that everything coming from Cochrane’s headquarters was problematic. The Group’s ORS officer, Mr Curry, had worked up some interesting numbers which he sent to High Wycombe just before Christmas. It seemed that there was a marked rise in losses and accidents for crews just passing their eleventh sortie. In raw numbers this was reflected in 46 losses among 1229 sorties flown between 22 October and 4

Science versus Attrition: November 1943–March 1944 173

December. While there was no minimizing the loss of life, the impact on efficiency was equally dramatic: ‘if crews with more than 11 sorties had been able to achieve the same low casualty rate as those with less than this number, 12 aircraft would have been saved over this comparatively short period. ... Put another way it would have been possible to complete 1700 sorties instead of 1229 sorties, for the same total loss, an increase of 470 sorties.’ The Group had concluded that ‘the crews are going back in their training’ and steps were now in place to tighten up the operation. These included extra training and putting the word out directly to those crews in this phase of their tours. In passing Curry’s letter to the Air Staff, Dickins expressed some surprise at this trend and said that his staff would be doing their own assessment of the Group’s losses as well as a comparison with the experiences of 1 Group, but that they had no reason to doubt Curry’s work. In an exchange of notes on 27 December, Oxland and Saundby discussed the conclusions and wondered if there might be some other causes. The issue now circulated rapidly among several of the senior staff and it was noted that 4 Group was finding the same phenomenon among its crews.46 It was quickly decided that more data was necessary and Dickins was asked to immediately prepare a letter for Air Staff signature asking all Groups for a monthly report along the lines of the statistics which 5 Group had provided. The scientists completed their review of the loss rate figures for December by 7 February; the raw totals were small and so they did not feel that it was possible to draw any conclusions other than that junior crews still appeared to have higher losses. The results for January were ready just two weeks later and these showed a slightly different story. Halifax losses were greatest during the first five trips. For the Lancaster there was a similar trend for the first five trips, but then another peak for pilots with eighteen to twenty-three sorties. There was some reluctance by A/V/M H.S.P. Walmsley, the Command’s new SASO for night strategic raids and former AOC of 91 Group, to take action, but once the February figures were assessed the scientists were convinced of their data. Dickins therefore wrote to Saundby, Walmsley, and the other senior staff: ‘The results show that the missing rate is highest for inexperienced crews. It decreases as experience is gained and shows a slight tendency to rise again towards the end of the tour.’47 The final act of this review took place in June 1944 with the publication of ORS S 187 ‘The Effect of Experience on Operational Efficiency.’48 Having reviewed some seventeen thousand sorties, including nine hundred losses, ‘the investigation confirms the higher missing rate, crash

174 The Science of Bombing

rate, early return rate, and combat rate among inexperienced crews. The tendency for the loss rate to rise above average just after the middle of the first operational tour is also confirmed, but is not found to be so serious as had been at one time suspected.’ The ebb and flow of losses was the result of an initial lack of experience, which was replaced by a growth of competence. An upward swing in losses occurred at the middle of the tour due to ‘over-confidence and consequent lack of care and a final fall [in losses due] to the care induced by realising the probability of safely completing the tour.’ Training was recommended as the only real means to mitigate the losses. ‘Efforts should be made to increase the veri-similitude of operational training, and in particular that more long practice flights with fighter affiliation should be regularly included.’ Perhaps as a result of the invasion of France the report did not get much review until close to the end of 1944 when Director of Bomber Operations staff at the Air Ministry began to push for the extra training. By this time the reaction of the Air staff at High Wycombe and of the boffins ironically was that there had always been high losses among new flyers and that experience was the best teacher.49 Despite this litany of problems there were a number of means to negate the fighters’ advantage. As seen, by the fall Window was not nearly as effective as it had been, but jamming (using Tinsel and Cigar transmitters) and spoofing of GCI radio transmissions (with RAF personnel transmitting false data on German frequencies under the code name Corona) were having some effect, and concentration was stepped up yet again. As a result of these measures the loss rate in October was actually just 3.9 per cent, although there were higher losses such as the 4.4 per cent at Kassel, three-quarters of these likely falling to fighters. In November, the first month of the Battle of Berlin, the rate sat at 4.0 per cent with sorties against Berlin experiencing only 3.6 per cent losses.50 For their review of the otherwise successful Kassel op, as for the analysis of all raids, the boffins had had access to a summary of the Luftwaffe’s defensive fighter activities within a few days of the attack. Report ZIP/ BMP No 50051 described the various elements of the night defence for 22/23 October, and, as was the normal practice, the document in fact covered all defensive activities for the twenty-four-hour period from dawn 22 October to dawn 23 October. In addition to a general narrative the report included a number of useful components: a sketch map of free-lance fighter activity; the time line of the running commentary relative to the raid; a map of the landing fields used by the fighters after the raid; a diagram showing the assigned altitudes at fighter beacons; and

Science versus Attrition: November 1943–March 1944 175

a table showing the weather data used by the fighters. The sources of these ZIP/BMP reports were deemed of utmost secrecy, but were clearly valuable to the ORS in putting together an accurate picture of what had gone on, and would normally have assisted the scientists in confirming or even identifying the time, place, and causes of various losses. In December the ORS issued a new analysis of German defensive tactics to update their earlier report on the impact of Window. This was an extensive document which reviewed how German tactics had evolved and were now making considerable use of running commentary and free-lance fighters. It explained how the Luftwaffe had introduced new control methods for GCI including Benito (using difficult-to-jam Morse code transmissions) and modified flak searchlight and fighter cooperation. It also commented on how Bomber Command tactics and equipment were continuing to degrade GCI and were also now able to confuse the free-lance fighters through the use of diversionary raids and feints. In addition, jamming and spoofing of enemy control frequencies had also proven effective in degrading the ability to direct the free-lance fighters towards the target area. 52 At the end of the month Dickins advised Saundby that AOC 4 Group had asked for additional copies of the report to pass to his squadrons. Saundby was not happy with this; he was concerned that 4 Group flyers would be negatively affected when they noted that losses among Halifaxes were 5.7 per cent, a full 2.5 per cent higher than for Lancaster units. Oxland defused the situation, explaining to Saundby that AOC 4 Group had simply reacted to the offer for additional copies that had been made with the original distribution. The AOC, said Oxland, did not ‘feel strongly’ about the matter but would prepare an extract for his crews rather than share the entire report with them.53 This manoeuvring, similar to the circumstances surrounding personnel casualty reports, is interesting to observe. From it one can clearly see that there was a concern among the commanders for the morale of the flyers. Conversely, despite the earlier events, this sensitivity was not something that the scientists might necessarily have noticed, since their work ethic generally was to paint as complete and unambiguous a picture of circumstances as they could. Contact with the Groups was a constant, and in January and February the Air Staff and the ORS reviewed a proposal from 5 Group on routing tactics. Two stratagems had been proposed by which to deceive the Luftwaffe about the intended target and thus negate the threat of fighters sent to the target area. The first suggestion was to have the raiders

176 The Science of Bombing

conduct a parallel ‘sidestep’ movement which would slide the raid onto the actual track leading to the intended target; the second would see the force over-fly the target and then reverse course once the whole stream was past it. The Ops staff felt that both proposals offered some value, but believed the second was problematic because of the risk of collision. Dickins commented that, while the course reversal was very risky, the sidestep manoeuvre did offer some merit in eluding free-lance fighters. He did caution that there could be higher numbers of stragglers than normally the case. Some days later the CNavO commented, confirming a dislike for the course reversal option, but also finding much to criticize in the sidestep proposal. He felt that there would be a requirement for a block of airspace about 100 miles long in which to conduct the sidestep and that the manoeuvre would have to be effected at a set time. This in turn would place its outcome at the mercy of found rather than Met winds, which would most likely lead to dispersion or blow the force over unplanned anti-aircraft defences. Finally he felt that turning the force 30° off its straight track would adversely affect concentration. The matter, along with these mixed reviews, was set before the DCinC on 26 January and he ruled that sidestepping on the way to the target was to be avoided. It could, if conditions were right, be used on return to the UK.54 The matter of the course reversal was still active, however, and at the end of the month Dickins forwarded to Saundby a paper examining the issue. The document pointed out that in addition to the risks of collision during the turn and the need to sequence the raid in reverse so that the PFF would still arrive first over the target, the proposed scheme had been overtaken by a modification of Luftwaffe tactics. Fighters were now attacking the stream while enroute rather than waiting over the target, so elaborate schemes to confuse the controllers as to the actual target were essentially pointless. Additional disadvantages were listed: the force would spend an additional forty-five minutes over enemy territory while conducting the manoeuvre; as well, the extra distance meant more fuel and a reduced bomb load. Moreover, the manoeuvre created a possibility for dispersion, and in attacking from the east, and normally into the prevailing wind with a resulting decrease in ground speed, flak batteries would have more time to engage. The overall conclusion, given these problems coupled with the change in German tactics, was that the reverse course proposal was impractical. Armed with these considerations, Saundby indicated to Harris that the 5 Group suggestion was not worth pursuing. Harris, after a review of the issue as laid out by the ORS, agreed, stating, ‘I have often considered this manoeuvre but it has too

Science versus Attrition: November 1943–March 1944 177

many snags.’ At the end of February the Air Staff replied to 5 Group indicating that while the sidestep manoeuvre appeared to have merit, the many problems associated with the course reversal suggestion made it unworkable.55 While such proposals were being considered, the loss rate was rising. The first raid of December against Berlin lost 8.7 per cent of the attackers; the monthly toll for the capital was 4.8 per cent. January was worse: 6.1 per cent losses against Berlin; 7.2 per cent in three other raids. The figures for February were no better: the first two raids of the month against Berlin and Leipzig saw losses of 4.8 and 9.5 per cent respectively. In the latter case most were the result of fighters which seemed now to be concentrated across northern Germany. These results led to an almost immediate change in tactics and strategy. Further attacks were aimed against southern Germany with approach routes from the south. Route markers had proven useful in steering the main force towards the target, but had also given the fighters an indication of the presence of bombers. Their use was now largely discontinued and attacks were broken into two waves that came in along separate approaches to confuse the defences. In the next three raids against Stuttgart, Schweinfurt, and Augsburg the loss rates fell to 1.5, 4.6, and 3.5 per cent. The boffins felt, however, that this reprieve was only temporary and that the Germans would adjust to the southern approach.56 In fact the scientists had been intimately involved in tracking these losses and in recommending the changes to tactics which had provided some succour. In early February the ORS had completed a tactical update, which made use of much of this data.57 In a highly detailed review of German and Bomber Command tactics (discussion of which spans the next two pages) they were able to describe how the Luftwaffe had initially reacted and was now finding effective counters for many of the procedures that the bombers were employing. The alarming result was that losses had gone from 3.7 to 4.9 per cent in November and by January were at 5.9 per cent. ‘It is therefore desirable,’ they said, ‘to examine the enemy’s improvements and to consider the desirability of modification to Bomber Command’s tactics.’ The scientists described the German tactics generally, but had also made ‘a night to night investigation’ from which they could draw several conclusions. First, free-lance fighters had more success when the raid passed close by German fighter beacons. Routes, therefore, should be as far from beacons as possible or should not parallel the line between beacons. They pointed to such an avoidance tactic on 29/30 December when an attack against Berlin had

178 The Science of Bombing

suffered just 2.9 per cent losses. It was also observed that the enemy was making use of PFF-laid route markers not just to spot and shoot down the marker aircraft, but also to get into the stream. As well, route planning needed review to avoid instances where the route in the vicinity of the target ended up in the shape of a loop which enclosed a German fighter beacon. Avoiding planning a loop meant in practice that the route home should be well clear of the outbound route so that enemy fighters could not easily loiter in one spot and hit the stream as it passed by in both directions. Finally, the scientists stated that the use of Mosquitoes for decoy raids no longer seemed to be working. It was surmised that the German observer organization was able to tell the difference between a few Mosquitoes and the main raid. From here the report went on to recommend new route-planning strategies. First, routes to the same or neighbouring targets on successive nights should be as different as possible. Second, routes should not pass over fighter beacons, nor should they pass along the path between two beacons. As well, routes should not form loops around the target area, particularly where fighter beacons were enclosed in the loop. Similarly, route markers should not be used in the vicinity of any such loops. Additionally, shortening of the stream should be attempted to minimize chances of fighters being inserted into the stream. Presuming that concentration had reached its practical maximum, an alternative would be to split the force onto different routes which would come together to form a single stream over the target. The routes should, however, be relatively close together so that the enemy did not have the option of bringing more defenders into the equation than if only one route had been employed. The scientists recommended that splitting the force on leaving the target might also prove effective if in the future the enemy started to put more effort into engaging the homebound bombers. They also said that since accurate winds were now being broadcast to the main force it was less important to mark the route and that route markers should therefore be used only if essential and in no case within 30 miles of the target. The ORS felt that feints and diversions should still be considered, although these feints would need to be larger in scope, and that the variety of feints should be increased in order to create the desired confusion and indecision on any given night. Finally the boffins noted that additional PFF resources would likely be necessary, particularly if more than one force was to be sent out on a single night. In concluding they stated: ‘it is considered that though present tactics have achieved much success in the past, the time has now come when variety

Science versus Attrition: November 1943–March 1944 179

of the type described above is desirable and likely to lead to a reduction in bomber losses.’ At the same time they admitted that the application of such changes would be difficult and involve more resources. On 10 February 1944 Dickins sent the paper to Saundby. In his accompanying minute he ‘suggest[ed] that the time has come when it is desirable to consider the possibility of modification to our general tactics.’ He added that Saundby might want to consider convening a small meeting for the appropriate Headquarters staff.58 Not coincidentally, it would seem, the Ops staff had the previous day made a similar proposal to A/Cdre H.A. Constantine, the A/Cdre Ops: ‘[Because enemy tactics had changed] ... there would be every advantage to be gained from careful investigation and detailed planning to counter them. No organized arrangements exist for the discussion and analysis of operations after their completion or to examine the tactics that are employed, and to see whether they were successful or not and thereby draw conclusions for future reference.’ Among the material to be reviewed was the ORS report just described and the senior scientists were included in the list of proposed members for the gathering. Constantine agreed59 and a meeting was organized for 20 February. Among those attending were Ops, Intelligence, Navigation, and Radar participants; the ORS representatives were Dickins and Smeed. Constantine opened the session, reminding those present that the Germans had generally shifted from protecting the target to inserting as many fighters into the bomber stream as they could.60 Smeed and Dickins participated actively in the discussion that followed. Dickins cautioned against splitting the raid into multiple streams which would require altitude coordination and would lead to an increased number of stragglers. He also said that flak losses were creeping up, attributing this to a failure by crews to discharge ‘Window’ according to standard practices. Discussion then turned to the notion of confusing the enemy by attacking two targets on the same night. Dickins felt that this approach would certainly confuse the enemy and reduce losses, but he also cautioned that it was felt by the Ministry of Home Security that a raid was likely to be more successful if it was as heavy as possible.61 He also indicated that problems with creepback would threaten the effectiveness of both attacks, given that about half of the already halved raids would be likely to bomb short of target. In the end it was agreed that despite some costs it would be worth trying a number of schemes, among them shortening and concentrating the stream; taking on two targets, which should contribute to increased concentration and reduce loss rate; using

180 The Science of Bombing

two routes to the same target; ensuring close coordination between raids and within multi-wave raids to maximize concentration and distraction factors; and, finally, establishing a permanent body with members from Ops, Nav, Intelligence, and ORS to review these issues.62 The scientists’ report having been reviewed and accepted, Saundby was thus ready to send it on to Harris on 24 February. His comment to the CinC was perhaps an understatement: ‘I think you will find ORS Report B 197 worth reading, if you have the time.’63 Harris did have the time and replied the next day. ‘I agree. Most of this has already been in our minds and some of it has been tried. We sh’d from now on work as far as practicable on these lines. But we cannot increase the size of the P.F.F. yet.’64 Saundby directed that the report be distributed to the Groups. Feedback was very positive. Cochrane at 5 Group: ‘I very much agree with all the points’; Addison of 100 Group: ‘This report sums up in an excellent manner and I think very usefully, the salient features of the enemy’s tactics’; Harrison, 3 Group: ‘most interesting’; Carr, 1 Group: ‘The above quoted Report covers a wide range of enemy defence measures and possible counter measures that might be adopted by Bomber Command to overcome enemy opposition’; Bennett at PFF: ‘I consider that this report is a very valuable one and my only complaint is that it came out too late. Hun tactics change and it is essential that the whole of the planning staff and the Command in general should be up to the minute in their knowledge of his tricks.’65 Based on the feedback from the Groups, the SASO was able to publish a short letter on 22 March which confirmed the new initiatives to be developed. It is evident that the report has aroused considerable interest and that the Groups have been thinking along the same lines. In general all Groups agree to: – (i) Large scale diversions. (ii) Divided routing on one target. (iii) Attacks on two targets on the same night. (iv) Split attacks in time. (v) Continual variation of tactics. (vi) Route markers should be dispensed with as much as possible. (vii) Routing over beacons is undesirable but cannot always be avoided, and should not be allowed to interfere unduly with route planning.66

Walmsley’s letter also reinforced the point that Harris had made about

Science versus Attrition: November 1943–March 1944 181

putting general thinking into concrete form: ‘it is evident that the report has aroused considerable interest and that groups have been thinking along the same lines.’ What is striking is the fact that it was the scientists who not only had analysed what had been happening, but also had subsequently conceived practical alternatives that were universally accepted and praised by the senior leadership of the Command. While this work was going on, March had also seen shifts in targets and losses. In the first instance the Command was now spending some effort in attacks against targets in France in preparation for Operation Overlord; loss rates in these raids were less than 0.5 per cent, although there were indications that the Germans would soon be deploying some fighters into France to counter these activities. Attacks against Germany approaching from the south were sometimes poorly challenged with losses in the 1–2 per cent range as a result, but there were others with losses closer to the normal 4 per cent range. A final raid against Berlin, however, resulted in 9.1 per cent losses. And then, on 30 March, the raid against Nuremburg ‘showed the extent to which Bomber Command’s position was precarious.’ The attack came in across the north with a straight leg of some 250 miles which passed close to two fighter beacons. The weather turned out to be clear and there was some moonlight; worse, the aircraft produced contrails not usually seen at their operating altitudes. No diversions or spoofs had been planned. Of 795 sorties allocated to the target fully 95 crews, 12 per cent, failed to return. 67 Now, as spring arrived, Bomber Command was obliged to change its plans and its tactics for two reasons. Shorter nights would once again rule out deep raids, particularly given the increasingly effective Luftwaffe fighter defences. At the same time, the approaching invasion of North West Europe signalled a shift from strategic bombing to tactical targets in France. In-flight fires, flak, fighter and crew experience levels had all contributed to a loss rate that put a significant strain on the Command, and the official historians say that Harris worked hard to ensure that the flow of aircraft, both new and repaired, was maintained. They say too that Harris had ‘an instinctive insight into the temper of his men,’ suggesting that he knew what he could and could not ask of them. Because of this, ‘the casualty rate was only contained at the expense of the major strategic object. The alarmingly high losses sustained in February and March 1944, when the latter was still pursued, were undoubtedly the cause which led Sir Arthur Harris to doubt whether the offensive could be indefinitely sustained by the “existing methods and types of heavy bombers.”’68

182 The Science of Bombing

Nothing that has been discussed in this chapter would tend to contradict this interpretation, but at the same time it should be apparent that Harris lacked the time, data, and means to analyse adequately the myriad tactics and procedures which had sometimes worked very well, but more often, and certainly during the Battle of Berlin, had not worked well at all. For these examinations he and the staff at High Wycombe had relied heavily on the ORS for not just the analysis of raw data, but also its ability to sift through the statistics and to distil the essence of a problem. The work done by the scientists had proven important on many fronts. Their analysis of the effectiveness of Window had allowed many commanders and planners to appreciate the value and the limitations of the system. They had been called to look into the effectiveness of navigation and of bomb aiming, going so far as to coordinate a review of the physiological and psychological attributes of the bomb aimers themselves. Questions of maximum bomb loads and aircraft weights had required the scientists’ attention; it was not difficult to see that making sure the aircraft had adequate fuel, the largest bomb load possible, and the manoeuvrability needed to deal with enemy fighters were all to some degree confounding variables which required a systematic analysis if the optimum combinations were to be achieved. They had investigated the curatives for Halifax deficiencies and pressed for improvements from the manufacturer and Air Ministry bureaucrats. When losses continued to mount, they found relationships between experience and losses, investigated the best use of new countermeasures, and eventually proposed a systematic review of defensive tactics and raid procedures. Often the scientists had themselves recognized the central element of a problem and on occasion had assumed some degree of executive authority in getting things done. That said, there was little doubt that they were in a supporting role, but it was a supporting role without which the Command and Harris could not have coped.

Basil Dickins, Head of the Bomber Command Operational Research Section from 1941 until the end of the war, at the time of his appointment as Director General Atomic Weapons in 1965. Dickins was a defence civil servant throughout his career. (Times of London)

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

AOC-in-C of Bomber Command Air Chief Marshal Sir Arthur Harris (centre) and his deputy Air Marshal Sir Robert Saundby (right) study a map of Europe. While the staff could make recommendations, the final discussions would be between these two with Harris making the call about the night’s targets. (RAF Museum.)

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

‘A Canadian bomber crew of an R.C.A.F. bomber squadron overseas reports to the intelligence officer after returning from a big raid,’ July 1942. The crew might have been returning from the Ruhr and would have provided their account of their route, the target, and any enemy activity. Hundreds of such reports would be collated by the ORS after each attack. (Canadian Forces Joint Imagery Centre PL 7978)

‘An air observer in training on a Fairey Battle bombsight,’ March 1944. The apparent youth of both flyers – student and instructor – is striking. Equally startling is the apparent ‘home made’ nature of the training device. Given these and other circumstances, that bomb aimers would have difficulties in hitting their targets might not seem so surprising in hindsight. (Canadian Forces Joint Imagery Centre PL 3575)

The complexity and confusion of what a crew would see and what the bomb camera recorded are shown on this slide. The photo, taken over Essen on the night of 5/6 March 1943, has been analysed by a photo interpreter to illustrate the variety of images that the long exposure of the camera would record. (Canadian War Museum, George Metcalf Archival Collection, CWM 19900192–015)

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

In this post-war photo, RCAF trainees receive instruction on the H2S. The official caption claims, ‘During the recent war, RCAF Lancaster bombers using this equipment were able to bomb with extreme accuracy through dense cloud.’ The truth was not quite as clear cut and the ORS followed the chequered performance of the system and its operators for over two years. (Canadian Forces Joint Imagery Centre PL 39325)

The potential of H2S is shown in this slide and map collage. In a radar picture of the River Humber it is possible to see the accuracy with which H2S could depict ground detail. It was still necessary for the navigator to keep track of the position of the aircraft at all times using both H2S returns and other navigation aids. (Canadian War Museum, George Metcalf Archival Collection, CWM 19900192–052)

‘V-Vic,’ a Halifax of 434 RCAF Squadron being inspected for damage after a raid against Stuttgart, December 1943. The scientists would note that 434 Squadron had terrible misfortune during this period of the war. (Canadian Forces Joint Imagery Centre PL 22451)

Halifax Mk II being ‘bombed up’ with a 4000-lb high capacity bomb and small bomb containers loaded with 4-lb incendiaries. This variant of the Halifax suffered from small vertical stabilizers and would have been prone to engine failures due to its liquid-cooled engines. (Canadian Forces Joint Imagery Centre PL 19507)

Bombing into a solid cloud deck, or undercast, a 4000-lb HC bomb and other general purpose bombs fall towards the target. The attack against Dortmund on 12 March 1945 consisted of 1108 aircraft of which only two were lost. (Canadian Forces Joint Imagery Centre PL 144267)

An attack on Calais on 20 September 1944. Target Indicators burn brightly through bomb debris and dust. At night the same debris, dust, and smoke would have made seeing anything but the TIs virtually impossible. (Canadian Forces Joint Imagery Centre PL 144271)

A stream of Window falls as the bomb photo is taken over Munster on 12 September 1944. While the effectiveness of the countermeasure was not as dramatic as it had been a year before, its use would add to the chaos facing the German air defence radars. (Canadian Forces Joint Imagery Centre PL 144263)

Halifax with wing fuel tanks on fire. The scientists made long and detailed studies of losses and found that fires in fuel tanks and engines were often the cause. They worked hard to convince staffs of the need for fire suppression systems. (Canadian Forces Joint Imagery Centre PL 144284)

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

A rare picture of a fighter interception of a heavy. The Halifax, over Hamburg on the night of 8/9 April 1945, appears to be manoeuvring (possibly evasive action), as its wings are not in the same perspective as those of the bomber towards the top of the larger frame. (Canadian Forces Joint Imagery Centre PL 144293)

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

This heavy has broken up in flight over Wangerooge on 12 April 1945. This late in the war there was little flak activity and so the cause would appear to be bombs falling from above. (Canadian Forces Joint Imagery Centre PL 144290)

During this attack on Wangerooge on 25 April 1945 the possibility of damage or loss from falling bombs was played out again as one Lancaster passed under another. Both of these aircraft would have similarly been under the one from which the photo was taken. Despite these circumstances, the rate of fratricide was quite low, but it did receive attention from the scientists as losses to enemy action dropped sharply in the final months of the war. (Canadian Forces Joint Imagery Centre PL 144281)

11 Debates Both Political and Technical: April 1944–May 1945

As the days lengthened into spring, Harris and his staff were faced with two major challenges. In the first case, there was a pressing requirement to find a solution to the German night fighter threat which had been building over the previous winter. As seen in the growing losses described in the previous chapter, if a means to neutralize the Luftwaffe could not be found, then the viability of the night bomber offensive was at risk. At the same time, strategic level imperatives were about to demand that the Command deliver a level of precision which had eluded crews since the opening raids of the war. In hitting transportation nodes behind the invasion beaches, many of them surrounded by French civilian housing, the area bombing that had been deemed acceptable and effective in raids against Germany and Italy could not be used. A/C/M Sir Arthur Tedder, former CinC of allied air forces in the Mediterranean and now Deputy Commander of the invasion force under Eisenhower, was adamant about the requirement to destroy the transportation network that would otherwise be used to bring reinforcements and materiel to the German defenders. As for the continued offensive against Germany, Portal, taking the line of much of the British strategic leadership, was in favour of an oil campaign, which would require similar precision, as a means of knocking the Germans out of the war. Harris, for his part, continued to see such precise targets as ‘panaceas’ and favoured a continuation of a ‘general area offensive.’1 This chapter examines these issues, showing the part played by the ORS in making good many of the tactical and technical challenges that continued to trouble the Commander and his crews. It also examines the role played by the scientists in supporting Harris’s concerns over bombing policy. In providing their advice they were able to draw upon an increasing expertise in identifying and solv-

184 The Science of Bombing

ing problems, employing a steadily growing data base that made their answers all the more comprehensive and reliable. In some cases events would give the Command alternatives that the scientists could not have foreseen, but in others the boffins’ hard work would reveal the best solutions to some significant dilemmas. As always, their conclusions would come under scrutiny, but as the strategic direction of the Command became more politicized it would not be uncommon for them to provide Harris with data and recommendations which he could use to argue for and against particular courses of action. In spite of disagreements at the strategic level, the conduct of bombing against various targets during the last year of the war demonstrated increasing effectiveness. Not only were there more bombers and bombs than in the past but the accuracy of the RAF’s bombing also increased until it was often more exact than the ‘precision’ day bombing being conducted by the US Army Air Force and more precise than even Harris had thought possible. The factors permitting this outcome included a reduction of the depth of penetration of raids into hostile skies as the Allies advanced across France and the Low Countries after D-Day; the steady replacement of Halifaxes with the more capable Lancasters; the overall growth in numbers of bombers; the improvement in munitions carried; and the development of new techniques and tactics.2 Taken together these changes meant that Bomber Command had become ‘more powerful, more accurate and more versatile, with the result that two important and hitherto established conditions of the offensive were changed.’ No longer were mass night raids the norm, nor was it necessary to saturate the target area when precision made it possible for a small force to achieve the desired results. As a result it had become possible to successfully attack more than one target on a given night.3 And yet the considerable benefits resulting from these dramatic and unexpected improvements in accuracy in fact served only to allow the Command to meet ever-increasing challenges. The target list was long, in part because of the divergent demands of the divided strategic leadership, such that there was little opportunity for returning to a target not destroyed the first time. Accuracy provided the only hope for meeting demands that increased more quickly than even the substantial growth of the Command’s resources. Accuracy proved to be all the more paramount in allowing the Command to grapple with the growing realization that the concentration of attacking bomber forces had begun to lose its effectiveness. Concentration was successful in concealing individual bombers from the GCI and AI radars, but could not, even with the help

Debates: April 1944–May 1945 185

of Window, prevent enemy controllers from directing the fighters towards the bomber stream enroute to, over, and on return from the target. Countermeasures including spoof raids had worked but took away from the weight of effort on the main targets. Taken together, the changing circumstances of 1944 drove the Command to find ways to engage multiple targets so that attacks would be both efficient in dividing the defenders and thus reducing losses, and also precise in destroying a target, particularly where the option of coming back a second or third time, as had been done at Hamburg for example, had become problematic.4 Planning for Bomber Command’s part in Operation Overlord had actually begun in the winter. On 31 December 1943 Harris sent Saundby a short memo calling for the preparation of an appreciation of Bomber Command’s potential to support the invasion of the continent. In fact he wished to limit the Command’s participation, and therefore directed that the appreciation should underscore that: pin point bombing by the heavy bombers is as a rule ineffective. Some of the aids are accurate enough for the purpose but if anything more than a very small force is used as backers up a big drag back or a throw off is bound to occur as soon as the target is obliterated by bomb smoke or incendiary glare. It is our experience that even small towns are a questionable objective for the Bomber Force and that if it is to be used with real economy the big towns are the best objectives because no matter what the drag back or throw off some damage is done somewhere. Railways are as a general rule an entirely unprofitable target and do not lend themselves to being put out of action except for a few hours. ... with the type of bombing we do at night the bombing of Occupied French towns outside the battle zone would be deplorable and is in fact at present forbidden by the Government.

Significantly Dickins assisted the air staff in producing the draft that made Harris’s arguments.5 In February Harris provided his thoughts on the Allied Expeditionary Air Force concept for the use of bomber forces. The scheme, although lacking precise target lists at this early date, seemed to call for Bomber Command to be able to drop some 29,000 tons of bombs monthly from February to April and then to increase this figure to 34,000 tons for May, June, and July. By Harris’s reckoning the best Bomber Command would be able to drop on average would be more like 22,500 tons based on 5000 sorties per month.

186 The Science of Bombing As regards the accuracy which our bombing can be expected to achieve in relation to ‘Overlord’ objectives, I wish to emphasise the following points, which are dealt with in greater detail in Appendix ‘C.’ (i) The accuracy of Oboe marking cannot be realistically assessed by considering only operations carried out under carefully selected conditions of moon and weather and without any substantial enemy opposition. The figures on which the probability of marshalling yards, let alone gun placements, being put out of action by night bombing are based, very seriously overestimate the degree of accuracy which is to be expected against targets of this type. (ii) I wish particularly to draw attention to the fact that the standard of accuracy in the attack of gun positions on which the entire success of the operation is admitted to depend [emphasis in original] has never been achieved under the best weather conditions by the use of Oboe technique. If weather conditions are such as to prevent ground markers from being clearly visible, the attack has no prospect of success whatever. (iii) Within Oboe range we have never in practice attained the degree of concentration round a specific in point on which the estimates in this paper depend. (iv) Many of the railway targets which the plan requires to be destroyed are outside Oboe range and some of these are in small towns which it would be extremely difficult to find and bomb successfully by any other method. It is therefore useless to hope that they could be destroyed by a single attack.6

Appendix C to Harris’s study provided calculations for bombing requirements against gun emplacements and marshalling yards. Three scenarios were depicted: Oboe, Oboe with backers-up, and H2S. For marshalling yards an average density of 2000 pounds – four 500 lb bombs – per acre was required to achieve the desired destruction: ‘On a 50 acre yard it is estimated that 11% of the bombs dropped should fall on the yard. Therefore, to obtain 1 short ton per acre 50 3 100 4 11 = 450 short tons must be dropped; and since 30% of the sorties would be abortive [actual aborts plus gross errors – this is described elsewhere in appendix C], 500 short tonnes or 110 3 18-bomb sorties must be dispatched.’ Regardless of how one looked at it, this was not the same as saying that, assuming perfect results, a 50-acre yard could be effectively attacked by 11 aircraft, which was the simple arithmetic requirement. By comparison, for a 500acre yard, where some 40 per cent of the bombs would hit the target, appendix C called for 310 sorties using Oboe. For H2S the figures were much less appealing:

Debates: April 1944–May 1945 187 In H2S raids the average density with half-a-mile of the aiming point is 0.008% per acre of bombs dropped and within 1½ miles 0.006%. Since the abortive rate is [?]5%, 14,800 short tons must be dispatched for a 50 acre yard and 19,750 short tons in the case of the 500 acre yard [emphasis in original]. The above figures show that attacks on marshalling yards requiring a density of 4 hits per acre are an extremely difficult proposition. The density required is equivalent to 640 tons per square mile, a figure 2–3 times that ever achieved by Bomber Command in its most successful attacks.7

To these were added diagrams showing average data from eight Oboe and nine H2S attacks of the previous summer. There is little doubt that the ORS had been the source of this data; and, as a result, of the four numbered duplicates of the document distributed within the headquarters, only Dickins and Oxland received copies.8 The scientists had portrayed the reality of past operations honestly and accurately. Clearly Harris, who some twenty years earlier had been a proponent and practitioner of precision bombing, would have liked nothing better than to guarantee that his crews could effectively and efficiently take on these targets, but to make that claim would have been an irresponsible exaggeration of the proven capabilities of the Command.9 Dickins would later write that a ‘knowledge of weapon effectiveness and of bombing efficiency and accuracy within Bomber Command has many applications in furthering the effectiveness of operations but one of its most important applications lies in assessing the weapon and force requirements for a given task.’ It was the ORS that was called upon to do this work, not just for the Command, but on occasion for the Air Ministry. These wider assessments had been done on an infrequent basis prior to the end of 1943, the Dieppe raid of 1942 being one such case. Generally, however, the available strength and effectiveness of the Command coupled with the nature of area targets meant that little attention needed to be given to the question, since it was unlikely that any target would be ‘overhit.’ Now in January 1944 the ORS was asked to look at figures coming out of Supreme Headquarters Allied Forces Europe (SHAEF) for the predicted effort required. The latter’s estimates ‘were considered to be grossly optimistic and independent estimates were offered to the Air Staff and were finally adopted by the Air Ministry.’ These new figures were derived by comparing the size and nature of the target with current weapons effects and bombing accuracy figures (along the lines of the

188 The Science of Bombing

analysis just discussed). This work allowed the calculation of tonnage requirements and subsequently of the force size for any particular target. Once operations began it was possible to compare actual with predicted results and show that the planning figures had been ‘reasonably accurate.’10 Harris confirmed this interpretation. It was essential to ensure not only that these small targets could be hit effectively, but also that the force would be used efficiently, so that the Command could engage all of its targets in the short time remaining before the invasion. Previously we had had no precise idea of just how large a force was needed to destroy a given target, which was natural enough, because we have never had anything like as large a force as we wanted for the work that was assigned to us. ... The necessary calculations were made by the Operational Research Section of my Command, and, apart from the unfortunate but inevitable fact that the mathematicians had not much data to use because there had been so few precision attacks in the past, the work was done in a thoroughly scientific fashion. From this point dates a most important revolution in bombing tactics, the significance of which might well be overlooked. For the last two years of the war it was possible to get from the Operational Research Section a very fair estimate of the weight of attack that would be required for the destruction of any given target. Naturally the estimates became more and more accurate as more operational data were accumulated, and as a result we were able to use our force with increasing economy of effort.11

The initial results of the pre-invasion campaign point clearly to the difference between what was needed and what could, at that point, be offered. The first phase of the Overlord campaign saw 14 ops beginning on 6 March. In all, 2513 sorties were mounted during these attacks, most involving 4 and 6 Groups bombing on blind Oboe groundmarking. In one case a modified Oboe technique (‘controlled Oboe’) was used where the master bomber laid his own marker, based on the blind Oboe marking, in an effort to improve the already notable accuracy of Oboe. Notable, but not without problems. It had been hoped for a bombing error of 640 yards but the actual error had been 680 yards. More importantly, while it had been predicted that ineffective bombing would amount to only 30 per cent of the sorties, in fact the ineffective rate was closer to 45 per cent. The initial figure had taken into account aborts, failure to detonate, and gross errors, each of 10 per cent. The first two estimates proved true, but the gross error rate was closer to 25 per cent and this

Debates: April 1944–May 1945 189

figure was attributed to marking errors rather than sloppy bombing.12 The whole question of Bomber Command’s ability to attack precise small targets had been the topic of heated debate in the weeks leading up to the invasion. Solly Zuckerman, a scientist who had worked on bombing effects before being assigned to the Mediterranean air forces and then to SHAEF for the operational planning for Overlord, was particularly involved in the air bombardment plan. It was he who had developed the Transportation Plan – a concept to attack the rail network behind the German coastal defences. As described in chapter 3, the first meeting between Harris and Zuckerman, with Dickins also in attendance, had been somewhat acrimonious and had led to Bomber Command reducing its estimate for the sorties needed to mount effective attacks.13 Some weeks later, Zuckerman again took exception to data being used by Harris to contest the ability of the Command to accomplish the plan. Zuckerman told Sir Arthur Tedder, his Overlord superior, that ‘whoever it was who had done Harris’s sums did not understand what he was doing.’ Zuckerman refuted what Harris and others had been saying about the origins of various calculations and conclusions. It was not the Bomber Command scientists or any other agency connected with strategic bombing that had devised a calculation for ‘translating areas of physical destruction into a functional assessment of what was happening to German war production. It had started ... with a study that Bernal and I had made of Hull and Birmingham.’ Moreover, Zuckerman later wrote, Harris also misspoke in saying that the Bomber Command scientists had devised the calculation for estimating the number of bombers needed to achieve a given bomb density – the true source was again Zuckerman’s work.14 A number of interpretations of this incident are possible. First, Zuckerman was perhaps happy to attack Dickins’s calculations and claims, as in doing so he could make himself and his plan look better. Related to this, there may have been an ego factor at play between the scientists, or perhaps with the author himself. Indeed, it must be recognized that Zuckerman described this disagreement in the course of writing his autobiography more than thirty years after the events. Why only then, when he had written about Operations Research in the 1950s? And, rhetorically one might ask, why would he not push his own agenda and reputation? It does seem as if there may have been some issues of ego at work. A third option is that Zuckerman was correct and Harris’s boffins had indeed used flawed thinking and numbers, but these possibilities require analysis. In this study it has already been seen that the scientists at

190 The Science of Bombing

High Wycombe were well connected with the greater military scientific community through the ORC, whose very purpose it was to exchange scientific data and findings. Would it not be reasonable that the Bomber Command ORS might well have used some of Zuckerman’s work, whether known to them or not, as the basis for their own calculations? Finally, it must be remembered that the ORS had already done a major study in the relative efficiency of munitions and thus one can conclude that they did know what they were talking about when it came to examining weapons’ effects. As the debate over the Transportation Plan continued, there were attempts to refute it, according to Zuckerman, based on an ill-founded prediction of civilian casualties. Zuckerman reported that these questionable figures had been provided by Dickins’s staff with help from Air Staff researchers, but that the work was based on Zuckerman’s own investigation of the German attacks on England. Zuckerman complained that certain assumptions had been imposed which led to a ‘gross exaggeration’ of probable casualties. He wrote that both he and Tedder suspected this concern over casualties being raised by Harris had little to do with the ethics, but rather with the target set which precluded the Command from carrying on its campaign against German cities.15 Zuckerman also indicated that he was ‘worried about the so-called “operational factor” which Bomber Command introduced into their sums when making their own calculations of the bomb-lift necessary for a particular task.’ Zuckerman said that his calculations took into account all the variables. Yet for some reason which was never defined, the Bomber Command chiefs kept multiplying the estimates they were given by their ‘experts’ of the amount of effort required to deal with a particular target, by a mystical number which they called the ‘operational factor,’ in order to make allowances for undefinable [sic] variables which they assumed had not been considered. It was useless pointing out to them that the standardised values of casualties or physical damage ... were based on actual experience. It was also impossible to get across the fact that any analysis of Bomber Command’s sums showed that the so-called ‘operational factor’ varied in a completely unpredictable way. Sometimes the calculated force that was required for a given task was multiplied by two, occasionally it was halved, and sometimes it was left as it was. It all seemed to depend on how Bomber Harris and his staff felt on the day. And it certainly showed that the men who planned and laid on the actual operations either did not understand what their Operational Research people were doing, or that the latter were not doing enough.

Debates: April 1944–May 1945 191

Zuckerman concluded in frustration, ‘The Air Force cry was always: tell us what you want done, but don’t tell us how to do it.’ 16 Anyone who has planned and executed military operations will find the whole notion of an operational factor a plausible concept, one easily equated to the Clausewitzian constructs of ‘friction’ and ‘fog of war.’ It might have been one thing for calculations to suggest an optimum bombing force or weight of bombs, but that was not a guarantee that the force would always find operating conditions to be optimal or even normal. Indeed the Services have always strived to include a reserve element in combat forces. For example, it is said that a ground attack should succeed with a 3:1 ratio of attackers to defenders, but the attacking commander will always prefer a larger ratio. And the defender always attempts to maintain a reserve element that can be thrown against an unanticipated weak spot in the line. Air force planners and commanders will always want to have 5 to 10 per cent spare aircraft for an op just in case of aborts in the early stages, even before takeoff in some cases. This said, if Zuckerman’s figures were always based on a worst case scenario, that is with a reserve factor included in the total, then there is reason to support his criticism; otherwise Harris’s ‘operational factor’ was not in any sense bluster. That it varied may seem odd, but then we have seen that some targets were known to be more difficult than others and would reasonably have required a greater operational factor if success was to be ‘guaranteed.’ One can also wonder why Zuckerman did not realize that, while scientific experiments can be normalized, if they don’t work no one suffers. Conversely, in these operations there was no experimentation, no gathering of data to prove some theoretical model. Each time men took off they were going in harm’s way, and commanders, particularly men like Harris who did think of their subordinates’ well-being, would want to ensure to the best of their ability both that losses were minimized and that targets were destroyed the first time round. And, once the invasion was launched, for the forces on the ground it was equally important that targets be destroyed on the first go; a second attack might well come too late for those fighting from minute to minute. Perhaps the final word on the operational factor should be left to Harris, the target of Zuckerman’s initial criticism. When we had our estimates of the tonnage required for each target we had to take into account the uncertainties of weather and the fact that we should be using new and experimental tactics; this meant that we had to be more lavish with our bombs than the strictly mathematical estimate re-

192 The Science of Bombing quired. Moreover we were working as never before against time. The date of D-Day had been many times changed in staff conferences, but now it was pretty well fixed for early June, which gave us just about three months, not only for the destruction of the railways, but for all the rest of the pre-invasion bombing that would have to be done. As always, much depended on the weather, and with this in mind I decided whenever we got reasonable weather to carry out one heavy attack, even when it might appear to be an unnecessarily heavy attack, rather than a number of light ones.17

It is significant here to note Harris’s philosophical approach to the use of scientific data. In no sense do we see distrust of the data, but at the same time he did apply a bit of ‘commander’s genius’ or commander’s insight to the problem, specifically ensuring that he dispatched at least the minimum force required as determined by the scientific calculations. Moreover, in order to ensure success so that the ground forces would be fully supported, he was prepared to use sorties in excess of what has been thought to be scientifically necessary. This approach was clearly at odds with Zuckerman’s comments, and could allow one to conclude that Harris was not trusting of scientific process. Rather, in fact, Zuckerman’s comments would appear to show that he had not yet understood that there was a difference between building theoretical models in the laboratory and the application of such models in combat situations. For Zuckerman to later criticize the Command’s technique was, perhaps, ill considered. A detailed examination of the ORS report describing the first fifteen Oboe attacks on the rail yards helps to explain what Zuckerman was missing. First though, it is necessary to consider a short explanatory note that Dickins had sent along with the report to Saundby on 9 June.18 Dickins stated that while the report indicated an average bombing error across the Command for the fifteen raids of 480 yards, data from more recent ops showed that 4 and 6 Groups had reduced their average error to 250 and 260 yards respectively. Saundby reported the figures to Harris some days later, stating: ‘There is no doubt that recent operations have done a lot to improve our bombing accuracy.’19 Despite this optimism, the detailed results found in the report were more pragmatic and showed a 680yard overall error. As well, the review stated that, while these new values had had a positive effect on the efficiency of the attacks, the effectiveness – the actual destruction of the targets – was not what had been hoped for. The result of applying these new parameters is that the overall requirement is increased by a factor of 1.43. The actual weight of attack dispatched on

Debates: April 1944–May 1945 193 the 11 targets considered in this report was 1.72 times the original estimated requirement, or 1.20 times the revised requirement, and even so, 4 of the 11 targets required further attention by tactical daylight forces in respect of priority buildings insufficiently damaged. The reason for this excess over the theoretical requirement [emphasis added] is that sometimes the priority buildings will be underhit (and the target will require further attack), and sometimes they will be overhit with a consequent wastage of bombs.20

These figures showed clearly that the bombers were incapable of guaranteeing specific results for specific raid sizes.21 And Dickins’s words showed with equal clarity that the scientists were laying the facts before the senior leadership. Harris was to revisit Zuckerman and his figures later in the year. In December, while debating the Oil Plan with Portal (a plan which would see the emphasis of renewed attacks on Germany focused on oil production facilities), Harris reminded the CAS that his OR Section had correctly predicted the requirement for nine thousand sorties per month to knock out the French marshalling yards earlier in the year: ‘On 12 December, he wrote to Portal saying that his ORS had just completed a study “into the feasibility of the oil plan,” reminding him that past estimates by Bomber Command ORS had always been reliable, giving as an example their statement that “it would take three times the effort estimated by the ‘expert,’ Mr Solly Zuckerman, to knock out the French marshalling yards and that was precisely what happened.”’22 Here was a clear example of Harris using ORS data to make his point, albeit a somewhat politically charged point. Had he not had confidence in the process and the results of scientists’ investigations one feels sure that he would not have been prepared to offer up such an argument. While these debates had been going on the flyers had been using the existing equipment and tactics to improve their accuracy. For example, 5 Group’s 617 Squadron had developed a low-level visual groundmarking technique where a Mosquito, often piloted by the CO, did the marking, permitting the squadron to achieve an 83 per cent effective rate and a bombing error of 380 yards (and eventually 285 yards). This compared most favourably with the Oboe results from the previous month. An offset marking technique was also developed by 5 Group wherein the master bomber and a number of pathfinders would arrive at the target early, find the wind at bombing altitude, and then establish an aiming point several hundred yards from the target (where it would not be obscured by smoke and fire once the bombing started). From this point a ‘false

194 The Science of Bombing

vector’ to the target was calculated and this, along with the actual winds, was used to determine a ‘false wind’ which would be broadcast to the main force. Once this wind was set in the bomb sight, the bombs aimed at the aiming point would impact the actual target. The boffins attributed the technique’s effectiveness to three factors: the use of a master bomber to eliminate sources of error; the use of a single aiming point; and the use of a common wind by all crews.23 The technique was in no small measure due to the ‘brilliance of the crews’ who did the marking, backing up and wind finding, as these were demanding jobs on which the success of the attack depended. The technique worked well but it was complicated and required clear weather. Technical problems could lead to the pathfinders being required to remain over the target area for long and nerve-racking periods. Three attacks on German targets at the end of April were ‘severe tests’ of the technique. The results ranged from excellent concentration against Munich on 24 April to typical errors of marking with significant bombing errors and losses in attacks against Brunswick and Schweinfurt on the 2nd and 26th.24 Other techniques were also being tried in the pre D-Day campaign, and all were meeting with some success. In addition to Controlled Oboe, Musical Newhaven was introduced; it used a visual check of blind marking and verbal guidance by a master bomber to enhance accuracy. Over the course of the two months these techniques achieved a range of success: ‘hits achieved as a proportion of hits expected amounted for blind Oboe marking attacks to fifty-seven per cent, for Musical Newhaven to sixty-four per cent, for controlled Oboe to seventy-nine per cent and for the 5 Group technique to ninety three per cent.’ Given these results, Harris’s concerns that precision attacks were beyond the Command’s abilities had been proven ‘largely without foundation.’ This not only meant that the Overlord bombing plan would work, but also suggested to seniors at the Ministry that precision attacks against the German oil industry should be possible. It remained to be seen if these could be mounted effectively and with tolerable losses. Yet when four night raids against refineries were attempted in June the results were ‘little short of disastrous’: of 832 sorties dispatched fully 93 failed to return. More Oboe-equipped Mosquitoes were available for marking, but otherwise the tactics were not altogether different from those used during 1943 and suffered the same sorts of problems in marking errors and creepback. Moreover the weather was bad for three of the four attacks. Only the attack on Gelsenkirchen on 12 June caused significant damage to the target refinery. With these sorts of results, the official

Debates: April 1944–May 1945 195

historians say, Portal had a strong case for recommending the switch to day bombing.25 The whole question of night bombing technique had been under close review in the spring of 1944, given the dismal results from the winter. On 9 April Dickins and Smeed attended the second meeting of what was now being called the Bomber Command Tactical Planning Committee.26 A/ Cdre Constantine opened the meeting reminding the participants that recent successful defensive actions by the Germans had convinced the CinC of the need for new tactics. Harris had already agreed to split the force in two and would now send 5 Group out independently using their own marking techniques, based on the work by 617 Squadron, to conduct attacks in cooperation with the main raid. When general discussion began, Dickins spoke about the issue of concentration, indicating that the tactic might not be as compromised as was now thought the case. ‘The success of the enemy finding the stream is proportionate to its length, but once in the stream the successful destruction of aircraft is proportionate to the concentration. It should be possible therefore, to calculate the length of the stream and concentration which would result in the least success for the enemy.’ As the meeting progressed, the Ops staff proposed sending all aircraft over Germany individually, the argument being that better navigation aids and self-protection now gave individual raiders sufficient chances of success, but Smeed cautioned that individual routing would likely play into the hands of enemy GCI controllers and lead to significant losses. The logic of his comments and of Dickins’s obviously carried the day, for the proposal was subsequently discounted. When the discussion turned to concentrating the force by using different heights, Smeed was similarly prepared to comment, indicating that he had already been reviewing the matter and had found that there could be problems, particularly when a small force used too great a vertical spread where wind variations at different altitudes could play havoc. Nevertheless, the idea was considered of some potential and so Smeed was asked to continue his work and to prepare a paper in collaboration with G/C Plans. The meeting ended with a number of ideas being left for review and implementation, all with the intention of shaking up current tactics and catching the Germans off balance. Key among them were the use of different and widely separated routes, the possibility of attacking three targets in a given night, and, against small targets, the use of blind bombing without skymarking by an H2S-equipped force.27 The whole question of the viability of H2S was under intense examination and would be for several more months. While the boffins and the

196 The Science of Bombing

Air staff at High Wycombe were wrestling with the very real problems of getting the equipment to do things for which it had not initially been designed, others were convinced that the Command was just not getting on with using the system to its capabilities. There was a bitter argument over the best employment of H2S which led Harris to complain that the Telecommunications Research Establishment (TRE) should stop playing in operational matters, specifically the use of H2S for blind bombing. This seems to have precipitated a ‘get Harris campaign,’ which led to a meeting chaired by the D/CAS, A/M Bottomley, on 22 April 1944 that was attended by Saundby and Dickins. Saundby, according to a later account by a participant, apparently admitted that only Harris was against H2S blind bombing and that he himself would like to see all aircraft fitted with H2S. This would permit main force blind bombing and would simplify mission planning and execution.28 The actual minutes of the meeting and the associated Bomber Command file do not indicate these comments but rather show that the Air Ministry and TRE were pushing H2S without appearing to recognize the problems which the Bomber Command scientists had been reporting. Saundby had warned Harris of the meeting; the D/CinC questioned data sent out in advance from the Ministry and confirmed for Harris that, as far as the Command was concerned, H2S Mk II was of little value even for the PFF. Only the Mk III and Mk IIA allowed accurate target finding. At the meeting Saundby and Dickins were accompanied by Saward while across the table there were a range of actors from the Ministry plus A.P. Rowe, Mr (later Sir) Bernard Lovell, and one other scientist from the TRE. The thrust of their argument was that the Command didn’t really seem to know what it was doing with H2S. Were crews to use the system the way TRE suggested – best crews only, concentrated in few squadrons, etc. – things would be fine. Saundby was able to rebut or explain most of what went on, but one of the main decisions of the meeting was that a full blind bombing raid using just H2S was to be executed as soon as possible. The record of decision stated that ‘the conditions governing this experimental attack [were] to be laid down by Bomber Command.’ To this, Harris later made the marginal note: ‘That’s kind of them!’29 While there was no record of Dickins having actively participated during the meeting, he was fully prepared to develop the experimental plan and did so within days of Saundby giving him the task. The operation order was ready by 16 May and Dickins provided a list of suitable targets at the same time. Saundby further specified that the raid was to be conducted using only H2S as a navigation and target recognition aid; no

Debates: April 1944–May 1945 197

flares or TIs were to be carried. The experiment would take place on a cloudless night and all aircraft would use bomb cameras. Crews were to be briefed on the purpose of the experiment and specifically warned against using visual cues for bombing. Then nothing happened. By the end of July the experiment had still to be conducted, and although in the meanwhile the Command had been fully occupied with Overlord with only a few raids on German targets, the Ministry had continued to press for the results. On 6 August Saundby explained this badgering to Harris, and said somewhat editorially that any raid was intended to ‘employ the best possible plan for the circumstances,’ perhaps inferring that there was little support for a Ministry-ordered fool’s errand. In any case, the CinC directed the experiment be mounted against Berlin on the first completely overcast opportunity. The Air staff pointed out to Saundby that this would nullify the ability to collect data. Subsequently the experimental raid was mounted against Brunswick on 12/13 August and turned out to be a failure. Harris directed Saundby to communicate with the Ministry ‘to the effect that the attack – as anticipated – was virtually a complete and certainly an expensive failure.’ Dickins sent Saundby the data to explain the fiasco. Of those aircraft reporting attack, only 5.9 per cent bombed Brunswick and only 14 per cent of those reporting attack bombed within three miles of the aiming point. Using cloud plots it appeared that another 15 per cent had bombed the Herman Goring works, which presented a confusingly similar radar echo, some 10 miles away. While a failure, Dickins thought the experiment worth repeating, but with the modification that the aiming point be identified by skymarkers to preclude gross errors. He had discussed the matter with Saward, who was in agreement. Saundby communicated these facts and views to the Ministry. At the end of September, almost five months after the original decision to mount the trial, the response reminded the Command that the original direction had specified the use of marking and then directed them to repeat the experiment.30 While it might seem from Dickins’s attempt to make good on the experimental raid that he was confident in the radar navigation and bombing aiming apparatus, it was during this period that he wrote to Harris to express his concerns about H2S and its effectiveness in the hands of the pathfinders. In a long minute to the CinC he presented the circumstances regarding the effective use of H2S, once again urging better selection and training of personnel, as well as an increase in the PFF should the Command wish to engage more than one H2S target on any given night.

198 The Science of Bombing In accordance with your instructions I have been considering the results of recent attacks led by the P.F.F. I have also discussed with A.O.C. P.F.F. possible ways of increasing their efficiency. I think there is no doubt that the weather is the real difficulty. When conditions are such that visual identification of the aiming point is practicable in the light of flares dropped blindly by means of H2S, the results obtained are, in general, about the same order as OBOE marked attacks. Failures generally occur when a blind ‘Paramatta’ is carried out due to the fact that the blind bombing with the present H2S Mark III and the present average standard of P.F.F. crews is not sufficiently accurate. While, therefore, better equipment is undoubtedly required, I feel that we are not getting the best out of the existing gear due to the lack of sufficient number of really good crews in the P.F.F. and insufficient training. It is certain that only a proportion of the existing P.F.F. crews are of the desired calibre and steps are urgently needed to ensure the P.F.F. obtain a higher standard. Further, although P.F.F. have now instituted a much improved method of selecting blind [marker crews], there are insufficient facilities to train them up to the necessary standard quickly and many potential blind markers are lost on operations before they are fully trained. With a return to attacks on Germany and consequent increase of losses, I fear that the P.F.F. will not be able to maintain even their present standard of accuracy unless the training organization is speeded up. We have given consideration to the number of crews necessary to carry out a satisfactory H2S attack and it is apparent that with the present standard of crews and the necessity of providing for emergency sky-marking in addition to ground-marking on many occasions, the present strength of P.F.F. is not really adequate for the marking of two targets outside OBOE range. A brief memorandum expanding the above points is attached hereto for your consideration.31

Now followed another two months of attempts to set up the Ministrydirected second trial raid. All manner of issues caused delay. These included selecting the Groups to participate; providing adequate warning to the PFF; and in one instance planning for a daylight operation and subsequently switching to a night attack. Then on 3 December Saundby minuted A/Cdre Ops: ‘The more I think about this the less I like it. I shall discuss the matter with the CinC in the near future. Meanwhile do NOT send [the plan for the follow-on experiment to the Groups].’32 At the end of the month he communicated with the SASO indicating that even if successful the concept could only be applied to a very small list of

Debates: April 1944–May 1945 199

targets. ‘Even if we have one or two successful experiments we shall not have proved very much and there will be little, if any, further opportunity of using this method operationally.’ There was no harm, he felt, in letting ‘selected’ (emphasis in original) H2S crews bomb blindly to see what might happen, but proving the results would be difficult.33 The entire episode is intriguing. It seems fairly obvious that there was no great desire to undertake the experiment and certainly not the repeat attempt. The reasons for this are less clear. Had the seniors lost confidence and interest in trying to make H2S an effective bombing aid? Were they less than pleased to have outsiders telling them how to do their business? Were they extremely busy with Overlord and with the rapid progress being made on all fronts? Were they fully satisfied with the various techniques and technologies which did seem to work? It might have been a combination of some or all of these factors, but it is clear that, while the staffs did what they could to make the experiment work, the leaders were not all that pressed to get on with it and not all that concerned when things did not work out. Before leaving the question, it is also important to return to Dickins’s note to Harris. Here was a prime example of what the scientist could do for his commander. All the issues had been neatly pulled together in a tight, cogent summation. There was no need to speculate about sources of difficulty or remedies; Dickins had laid out the problem and the solution. That Harris did not seize the issue and press to fix the H2S conundrum was probably influenced by the strategic priorities and factors described above. At the same time, the CinC’s thinking and the caution shown by Saundby were no doubt affected by the ORS reviews of H2Srelated losses and performance which had been coming in steadily during the spring and summer and which pointed to the problematic nature of the device, and particularly the crews’ difficulties in making the most of it. Even as the debate over the system’s viability had been joined in April, the scientists were generating data and conclusions which pointed to problems. That month Dickins asked Smeed to look at some 5 Group figures on recent losses, asking that the numbers be analysed on the basis of the different equipment carried. Smeed replied on 11 May saying that the loss rate of H2S-equipped aircraft was 15.5 per cent while non-H2S losses amounted to just 2.2 per cent.34 There was no doubt that the difference was significant, but it was not possible to determine why. Perhaps, thought Smeed, the H2S aircraft stayed on the planned route while the others, without the navigation aid, wandered off track

200 The Science of Bombing

while enroute. If the fighters were then able to find the H2S stream this might lead to the difference. There might be other explanations, too, and Smeed said that the staff would keep an eye on developments. Just weeks before this Roberts had asked the CNavO to publish another ORS report – this one analysing the 24/25 March raid on Berlin. After discussions, the scientists and navigators both agreed that ‘the performance of all H2S aircraft should be investigated.’ ORS Report B206 showed that little had been learned by the navigator corps, for there were several problems with gathering, transmitting, and then using the common wind that had been intended.35 As the detailed investigations got underway, Dickins provided A/Cdre Ops with a brief synopsis of the report. It is concluded that losses were mainly caused by flak as the result of flying in very low concentration over a large number of defended areas due to a general southerly displacement and scatter of the force. This was brought about by the fact that many wind finding navigators disbelieved the high winds they obtained and either watered them down or sent back only the lowest values, thus leading to a systematic error in the broadcast winds.36

Matters had been made worse by the delay in receipt of the winds at Command and subsequent delay in broadcasting; all this was further compounded by the fact that a large proportion of the force did not use the broadcast winds. Dickins concluded: The lessons to be learned from the investigation appear to be: (a) That the delay in the receipt of winds at this Headquarters and passing them back to the main force must be reduced (steps have already been undertaken in this connection). (b) That it must be emphasised to wind finders that they must send back the actual winds found and not attempt to adjust them. (c) That all navigators must use the broadcast winds in preference to their own found winds or their own estimates.

The same lacklustre performance was subsequently reported in ORS S 139, which, unlike B 206, was widely circulated.37 The great value of the ORS was its ability to look at problems over an extended time frame so that trends became clearer. Thus, while individual raids continued to be dissected, the scientists were also conducting a relatively long-term analysis which they were able to publish at the end of

Debates: April 1944–May 1945 201

July. In it they reviewed the navigation evidence for the period between 21 January and 24 April. A typically exhaustive review of the statistics and crew reports confirmed many of the problems that had been seen with individual raids. In the end, however, they made just three recommendations: • In order to obtain better timing and a higher reliability of fixes navigators should take fixes every 5–10 mins. and should in all cases check them by [dead reckoning]. • Winds should be found over periods between 15 and 40 mins. Others should not be broadcast. • Greater care should be taken in calculating winds, checking the measurements of distance and time.38 The scientists continued their efforts throughout the year, seeking to find ways to resolve the limitations of H2S. One such proposal had been made in April and suggested the use of a ‘reference point’ method where a landmark readily identifiable by H2S would act as a check point for a short final dead reckoning leg to the target. This would do much to mitigate the apparent impossibility of finding precise points in the midst of large built-up areas such as Berlin. Dickins passed the proposal to the Radar staff in late April, asking for their thoughts. The somewhat dismissive response came back within days indicating that the concept was already well known and that it had been used by 4 Group in minelaying operations; if anything the scientists’ version was much more cumbersome. This determination was quickly overruled by more senior staff in the Radar branch and in discussions with G/C Saward it was agreed that the idea was in fact worthy of testing.39 5 Group was to be tasked to conduct the trials, while the Bombing Development Unit (BDU) would do additional work with the Mk IIIA sets when they arrived. At the same time another study of H2S was underway based on the 24/25 March raid on Berlin. This study dealt with navigation and confirmed the results of the earlier navigation investigation, namely that H2S was an excellent tool for navigators who took frequent fixes in addition to normal navigation. In the new report navigators’ accuracy had been grouped into three categories. The most accurate took fixes at least once every ten minutes and of these 78 per cent kept the aircraft within 10 miles of track; conversely those taking fixes not more than once in twenty minutes were within 10 miles of track only 25 per cent of the time. There was no recommendation in this study, as the corrective action was

202 The Science of Bombing

obvious. If some navigators could make frequent checks on their position and thus keep to the designated track, then surely the rest could too. In putting the study before Dickins, his staff also felt that there were some additional points to be made. First, it was important that a common method of using H2S be promulgated and got out to the squadrons; it was thought that a meeting with the Group Navigation and Radar staffs should be convened. At the same time it was considered important to revamp training and to get the necessary directive on this aspect of H2S out to the Groups. Dickins passed the document to CNavO for his endorsement, but the latter was not inclined to give his unlimited support to the study, as, he stated correctly, the conclusions were based on just one raid. Moreover, he felt that training and procedures were fairly well standardized. As a result the report was toned down before being issued.40 It was one of the boffins’ rare setbacks, but the fact that they did not take the matter to the SASO or the D/CinC is indicative of the normally effective working relationship among the staff. The scientists did not give up on looking for curatives to H2S’s limitations, and during the summer they reviewed the data from several months of H2S bombing. The resulting report noted that the introduction of the more capable Mark III set had not seen any appreciable improvement in performance of H2S overall; indeed, the bombfalls plotted within the mean radii of targets were now only 34 per cent compared to the 50 per cent seen in earlier studies, but this was mitigated by the fact that H2S raids were now being conducted against smaller cities and that the data for Berlin, though ‘scanty,’ appeared to indicate better results than in the past. That said, the report still recommended the adoption of the reference point method for blind bombing whenever suitable landmarks were available. The boffins were ready to put forward these fairly severe criticisms of the performance of the equipment, but the Radar and Nav staffs once again made a number of rationalizations that included the problems of getting the Mark III into service without proper training as well as the usual teething problems that saw many cases of technical failures. Dickins agreed with these counterarguments and had the authors revise the study before presenting it to Saundby in the middle of August.41 There was no indication that the navigators of the Command were to learn from their mistakes. An analysis of a raid on Munich on the night of 7/8 January 1945 pointed to the failure to use actual winds, which differed from those forecast.42 Timing was poor and the raid was compromised as a result. Failure to update and use actual winds had diminished the effectiveness of yet another attack.

Debates: April 1944–May 1945 203

Much of the struggle to obtain accurate winds and then get the navigators to use them was, as was mentioned earlier, a team effort. The boffins were the only ones with the expertise and personnel to identify and isolate the problems, but it was quite appropriately the CNavO who, somewhat belatedly during this period, sent the criticism to the field. The problem seemed, however, one without solution. It might have helped to bring Harris or Saundby and perhaps the group commanders in on the debate, but one can hypothesize that there was little to be done. For one thing, the only way to improve standards would have been to take crews and units off line to conduct training, which might well have improved individuals’ abilities to deal with the complexities of navigating accurately in the dark. But training could not simulate the stress that came with navigating not just in the dark, but over a hostile nation which was putting up every defence it could to bring down the bombers. There was no doubt that the degree of perfection of the navigators’ skills left much to be desired; whether their accuracy could have ever been brought up to that high degree of proficiency hoped for is another matter. For once, it appeared that the boffins were caught it what amounted to a lose-lose situation. Ironically, much of this work and the associated frustration were of marginal value, since, of all the systems to be optimized over the summer months, H2S was arguably at the bottom of the list. In truth Saundby and Harris were alive to the big issues. At a January 1945 navigation conference the D/CinC commented on time-keeping and track-keeping. He ‘referred to recent O.R.S. reports on this subject and spoke of the value of Window. When aircraft were as much as 20 miles off track they were out of window [sic] cover and an easy prey to the night fighter. He further stressed the need for all aircraft to take advantage of the period of marking by the P.F.F. and to avoid being late and having to resort to bombing fires or bombing on navigation aids.’43 The discussion on this issue continued and it was ultimately decided that, while both were important, it was more critical to arrive over the target on time rather than to focus on maintaining track. It was thought better to make track corrections with the intent of getting over the aiming point at the right time than to make abrupt and significant changes in direction which might result in collisions while attempting to regain track, a manoeuvre which could also mean a significant loss of time. Even in the closing weeks of the war, the leadership within the Command found it necessary to drill home the need for vigilance and basic competency in the air. In March 1945 the SASO sent a fairly blunt letter to the Groups based on an ORS analysis of navigation from earlier

204 The Science of Bombing

in the year.44 Comparing ops done in January, the scientists had noted that when a raid against Munich (described in more detail earlier in the discussion of H2S problems) experienced winds different that those forecast the navigators in large measure had done nothing to adjust their times to meet the target ETA, even when they knew about the wind change. 5 Group, for once accepting the Command’s thinking, at least to some degree, sent this blast down to the squadrons indicating some disappointment with the matter,45 but then subsequently replied to the Command saying that an internal review found that most of the Group’s crews had used the proper means to adjust for the winds despite severe fatigue from an intense raid program in the preceding days. At the same time, however, they asked if in future the Command might be able to indicate which crews and units were at fault so that not all would be tarred with the broad brush.46 There was no question that progress through 1944 and 1945 had been problematic in getting H2S to do what the Command needed of it and that there had continued to be even more basic challenges in achieving the level of basic navigation skills that would allow crews to use the equipment effectively and efficiently. At the same time, however, many of the other systems and tactics were producing a winning combination. The thought of precision bombing was no longer a faint memory from 1939. As spring turned to summer and fall, Bomber Command could, within limits, go pretty much where it wanted and when – and hit the target when it got there.

12 Round-the-Clock Bombing: April 1944–May 1945

During much of the period that the scientists and specialist staffs were chasing the H2S problem, the Command had for the first time in almost five years begun wide-scale day attacks. Indeed, the period from June to October 1944 saw the crews conducting both day and night operations. The former were directed largely against tactical targets in support of the invasion of the continent and against flying bomb launch sites, and contributed to the development of daylight precision bombing capabilities. As discussed earlier, Portal and the Ministry had been proponents of day bombing for some time, but Harris remained opposed to the concept, a perspective held since the few expensive daylight raids of 1942. He argued that little had changed by the spring of 1944; the Ministry did not agree. It was thought that escorted attacks could be effective over the occupied countries and western Germany and that there needed to be some capacity for daylight ops should an emergency arise during the invasion. The issue of losses at night was, of course, added to the list of reasons to shift to daylight raids. Yet Harris remained unconvinced, citing a range of reasons why daylight ops would be costly and no more accurate than the night ops already being undertaken. There might be a case for day tactical operations, he said, but only if the Americans could not take on the tasks, adding that, in any case, good weather was a rare commodity which made much of the debate moot.1 The Canadian official history says that Harris had been aware for months that switching to day operations might prove the most efficient use of his resources, given that loss rates by day were low and his crews were capable of successfully engaging precision targets. Yet he chose to remain committed to the stalemated night campaign. He believed that there were insufficient escorts for sustained day operations and that the

206 The Science of Bombing

bombers had inadequate defences to counter a well-executed attack by fighters. Moreover, the Americans had the tactics and training to concentrate their heavier firepower, while the Command was unable to devote the time to become proficient at anything better than flying in a ‘gaggle.’ Finally, unlike the high-flying Americans, British aircraft operated below 18,000 feet where enemy flak was effective. 2 What caused Harris to change his mind was the apparent collapse of the German day fighter force when it failed to meet the invaders on 6 June. As a result, Bomber Command started a series of daylight raids, escorted by fighters, at mid-month. The loss rate for these sorties in June and July was just 0.4 per cent. At the same time, however, there had been no opportunity to develop or train for any sort of formation flying and the bomber stream in daylight was awkward in the extreme. It was found that night marking techniques were the only means of identifying the target; these tactics of course meant that each bomb aimer remained responsible for identifying and bombing the AP.3 Harris’s gradual acceptance of day attacks can be seen in the work of the scientists.4 The Groups had been conducting extensive night raids over France before and after D-Day, but losses began to creep up as they operated in what were not true night conditions. Smeed’s staff produced a short examination of the issue and he passed it to Dickins on 17 June.5 It pointed out that any aircraft approaching a specific point, in this case the target area, ‘from the sun’ were much more visible than those that were ‘down sun,’ where there was not much difference in visibility between twilight and normal dark. It so happened that this put the fighters in the area of low visibility while the bombers, coming from England and with the summer twilight directly behind them, were in the worst possible position. Making matters worse, the higher the bomber’s altitude the greater the level of illumination. It was shown that at the end of June a Lancaster flying at an altitude of 20,000 feet at 53q N (north of London) would be visible for distances greater than 10,000 feet; this value dropped to 6500 feet at an altitude of 10,000 feet and to 2500 feet for the same aircraft at 50q N (Normandy and the Ruhr) at 10,000 feet. It followed that aircraft should be operated, for the balance of the summer, at as low an altitude as possible and that their routes should be planned to enter and leave enemy territory from as close to a southerly direction as possible. Dickins sent these remarks on to A/Cdre Constantine, suggesting that the Tactical Planning Committee would want to review the boffins’ data. An information copy of the proposals was sent to Saundby, who immediately recommended the scheme to Harris; the CinC

Round-the-Clock: April 1944–May 1945 207

agreed that the ideas were worth a try.6 The attached synopsis proposed that attacks should take place, during this period, using multiple twominute-long waves of fifty aircraft (i.e. a density of twenty-five aircraft per minute) which would proceed to and leave the target by separate routes, but which would attack in a coordinated concentrated attack. The concentration within waves would be such that a free-lance fighter would have difficulty finding multiple targets, although it was admitted that GCI controllers might have more success in directing a number of fighters to one of the waves. It was also proposed that Window be used only within 30 miles of the target, as the countermeasure tended to produce a clear indication of the general position of the attacking force. The note went on to suggest the use of a full suite of radio countermeasures to neutralize GCI. As well, it was restated that visibility increased with height and latitude and thus routes should transit as far to the south as fuel would permit, with the aircraft flying as low as possible consistent with the flak threat.7 With general agreement to the proposal among the senior leadership of the Command, the matter was placed on the 26 June agenda of the Tactical Planning Committee along with several issues related to day attacks.8 Constantine explained that the CinC was concerned that there was insufficient darkness for mounting raids deep into Germany. As a result it had been decided to conduct attacks against Germany in daylight, making use of fighter escorts. To do this, however, meant a modification of tactical procedures, and the current session had been convened to review the factors and take the necessary actions. Most obviously the threat of flak would call for the maximum use of altitude. While flak was to be avoided as much as possible, it would still be necessary to use Window within 30 miles of Berlin and the Ruhr and within 20 miles of other targets. Visual marking would continue, although strangely the questions of blind marking and bombing were deferred. The earlier debates with Zuckerman had been over accuracy and raid size and here the scientists had been thinking about these issues. Dickins spoke on the question of ‘over hitting’ suggesting that stick length (the distance between the impact point of the first and last bomb from a single aircraft) needed to be shortened so that bombs were not wasted beyond the target. This matter, too, was deferred, although it was agreed that the question should be reviewed once sufficient data became available. Putting off discussion on so many items would in hindsight appear somewhat odd, but, just two weeks after D-Day, undoubtedly the Air Staff were working at peak capacity and they had made the sage decision to

208 The Science of Bombing

allow the boffins to collect more data and report back. In fact, the experience gained over the summer was refined and subsequently permitted the adoption of these tactical procedures for the main campaign once it was re-engaged.9 Two of the night precision techniques, Oboe groundmarking and controlled Oboe, were directly transferable to day operations. Additional techniques were added, the simplest of which were visual bombing and Gee-Dead Reckoning, the latter involving bomb release over a Gee fix or after a timed run from a Gee fix. Finally, ‘the much more ambitious and complicated Oboe or G-H formation technique[s]’ were introduced. In these types of attacks the leader would use one or the other device to bomb blindly on the remote signal from the control station and the balance of the raid would bomb on the leader, a system not unlike the USAAF ‘toggling on the group leader.’ Two factors determined the success of these new techniques: the accuracy of the leader’s bombing and the ability of the raid to keep the leader in sight. The advantage of these attacks was that they could be conducted regardless of the weather on the ground, but they were limited by range from the transmitting station, unlike 5 Group’s visual marking techniques. The latter, on the other hand, were limited by weather. Each method, therefore, had its strong points and its Achilles’ heel.10 From October to the end of 1944 numerous raids made use of these updated tactics and bombing was highly accurate. The Command employed a variety of marking techniques, selecting that most appropriate for the weather and the target. There were instances when both daylight and night ops of over one thousand sorties each would be dispatched in the same twenty-four hours, while smaller attacks too would be sent out.11 The final phase of the war, from October 1944 to May 1945, was the most spectacular in terms of results achieved, ‘but, from the point of view of the operations themselves, it was, perhaps, the least interesting. The success of these final operations was the logical and almost inevitable crowning of an edifice whose foundations had by the painful processes of trial and error and of courage and endurance been laid in the past and to a not inconsiderable extent in the recent past.’12 In less eloquent terms it could be said that for the decision makers, including the boffins, this was the dénouement of the piece; it was almost anticlimactic to watch the bombers return op after op with general success and few casualties. But there was still one strategic issue which would involve the scientists. During fall and winter 27 attacks involving some 5194 sorties were made against oil refineries. 13 Yet the transition to this oil plan was a most

Round-the-Clock: April 1944–May 1945 209

acrimonious process where Harris and Portal, colleagues for decades, were at loggerheads. Harris’s reluctance to convert to the oil targets seems strange, given the excellent results his crews had been achieving, but the evidence gathered for him by Dickins and the boffins did suggest that these targets would not be easily destroyed. As the fall advanced and it became likely that Harris would be required to take on oil targets, he had asked Dickins to prepare an estimate of the sorties required to do the job. The brief analysis was ready by the end of November and showed that once location, degree of precision, bombing aids, and weather were taken into account, and not forgetting the Germans’ ability to repair damage, it would take some 9000 sorties a month, split between 13 day raids of 200 aircraft and 18 night attacks of 350 aircraft each to accomplish what had been asked for.14 While formation attacks using Oboe or GH were possible for shallow penetrations, deep raids would not have these precision aids and it was expected that accuracy would thus fall off from the maximum possible 80 per cent effectiveness for day precision attacks to no better than 26 per cent for night raids, assuming good weather and little moonlight. The final calculus, however, hung on the weather, and here it was likely to let them down. The predicted goodweather, low-illumination nights in the winter were only three or four per month, and it would not be possible, therefore, even if attacking two or three sites in one night, for the Groups to hit all fifteen deep targets during a thirty-day period. Perhaps taken aback at these figures Harris asked if the Command could expect help from the Americans or Mediterranean-based forces. Allaying some concern, no doubt, Saundby was able to reply that the estimates were based on the assumption that there would be no support from the Americans, an eventuality that would have seemed remote.15 At the end of the year the scientists were able to comment on the results of six raids into Germany.16 Thirty-seven per cent of the sorties had been ineffective either due to early returns, gross errors, or other problems. Of those that bombed effectively the systematic error was 1080 yards and the overall average error was 1460 yards – almost a statute mile. These statistics weighed heavily on the Command’s ability to guarantee success on deep and small targets. If ever Harris had a case to make for not wanting to be held accountable for a target set that remained at or beyond the limit of his crews’ capacities, these predictions and results were surely convincing reasons why the Command should not be given the task. While the decision was clearly one to be made at the highest levels of strategic command, the scientists had very expertly provided

210 The Science of Bombing

the best possible analysis for that discussion. That the task did come to the Command was simply a prerogative of the strategic leadership, but at least the decision could be made with knowledge of the issues. At the same time that the bombers were going after the oil refineries there was also a Communication Plan under way which saw the crews take on dozens of targets, including canal and rail systems. These attacks were all largely successful, the only significant difference being that, unlike transportation targets in France, where collateral damage was to be avoided, such was not the case in Germany where damage both to the transportation system and to the surrounding towns was a desired outcome.17 In reviewing these ops, and while still cautious, the scientists seemed more positive in their thinking and calculations. In January 1945 the boffins received a request from the Ministry to provide planning figures for attacking German marshalling yards.18 They prepared another comprehensive analysis similar to the work they had just done on oil production facilities and sent it forward for approval. A number of factors were included and, again, distance from Oboe ground stations was a definite handicap in engaging eastern targets. Based on a density of three bombs per acre and providing figures for 50 or 75 per cent level of success, they concluded that in the best circumstances, day Oboe attacks would require 114 sorties to achieve a 50 per cent success while 163 aircraft would be needed for 75 per cent. Comparable figures for H2S Newhaven attacks in the east were 540 and 958 sorties. Clearly these far-off targets were not going to be cheap investments, if ordered. Saundby reviewed and submitted the scientists’ document to Harris.19 In his minute to the CinC he indicated the study’s purpose and the fact that the conclusions reached were fairly preliminary. He said that the proposed covering letter would ask specifically that the Command be given the chance to review any plans that might be developed on the basis of this data. Harris gave his approval for this approach. The report was quickly dispatched and, in his covering letter, Saundby reminded the staff at the Ministry that the figures were largely theoretical and that the realities of the day would need to be considered should the operations be put into action.20 Just four days later, A/Cdre Bufton, now Director of Bomber Operations, replied, thanking Saundby for the excellent information and informing him that it would be used, with the stated caveat, in developing plans for attacks on German transportation. A long list of possible targets was included with Bufton’s reply, and he asked that the ORS develop ‘an authoritative estimate of the scale of effort which would be required

Round-the-Clock: April 1944–May 1945 211

against each target, to achieve 75% probability of achieving the effect specified.’ Bufton made note of the fact that the target list contained some bridges. While those targets were normally allocated to the US Eighth Air Force, he simply asked that the boffins identify the necessary resources should the bridges be given to the RAF.21 With approval from Saundby, Dickins took on this task as one of his highest priorities and within days the scientists had worked their way through the list, providing figures for the 75 per cent probability, given good weather.22 It was felt that the bridge targets were quite feasible and Dickins relayed this as well. His staff provided the details for taking on the bridges, stating that only two hundred sorties, on average, would be required for a 75 per cent chance of success. This was tempered with a recommendation to employ modified groundmarking so that the TIs were dropped on the banks and not inadvertently into the rivers where they would be extinguished. It was also noted that the numbers of sorties were a strict mathematical minimum that did not include pathfinders and that the figures might well need to be adjusted upward to suit tactical needs.23 Armed with this information, G/C Plans now developed the actual plans for the raids. He minuted Saundby underlining that, while the numbers of sorties for each target were relatively small, they did add up.24 There were fifty-seven bridges on the target list and the sorties needed to take them on totalled some thirteen thousand. As had occurred so many times over the previous years, the work of the scientists had been crucial in allowing the commanders both to make an informed decision and then to put that decision into effect. It might be possible to conclude that the Command had by now come to grips with its operating procedures, yet the scientists continued to find problems which required fixing. Several reports on bombing accuracy had been published over the summer and each built upon the knowledge and conclusions from the earlier reports. For example, studies were made of attacks against lightly held targets in occupied areas and German towns which would bear the brunt of deep night raids over the last winter of the war. In September ORS S 184 described bombing against lightly defended targets, such as the marshalling yards, which the Command had been engaging during the previous six months.25 It was observed that there were three types of bombing used for these targets: Oboe; medium-level visual bombing using the Mk XIV bombsight; and low-level bombing. Where no master bomber had been used crews engaged the MPI of the TIs; where a master bomber controlled the raid he would indicate which TI to engage. Overall fifty-seven raids had been

212 The Science of Bombing

analysed and it was concluded that 82 per cent of the bombing had been effective; it was also calculated that the mean overall systematic error was 300 yards and the mean random error stood at 385 yards. Surprisingly, controlled Oboe raids by day had higher error values than regular and controlled Oboe night raids. Looking at the Mark XIV sight as a possible source of error, the study reported that with good winds and a qualified crew the sight could give an accuracy of ‘about’ 130 yards at 10,000 feet (in other words only about 30 per cent of the random error). By comparison, the more advanced, but more complicated, Stabilized Automatic Bombsight (SABS), employed only by 617 Squadron for highlevel precision (attacks using a range of oversized weapons such as the 22,000-pound Grand Slam), had an operational error of 120 yards at 10,000 feet and 170 yards at 16,000 feet. Several ideas were advanced to explain why the Mk XIV could be so much less accurate in operations than in training: operational bombing was done using sticks of bombs rather than a single practice bomb; there was a random wind error; markers were dropped in salvos and could easily cover an area of 400 by 500 yards; there was a crush of aircraft over the target; and the master bomber adjusted the aiming point as the raid went on thus producing new aiming points and affecting the ability to judge distances. By comparison, the very few crews who used the SABS were affected by few if any of these friction points. At the end of the year the scientists were able to comment on the circumstances of bombing German targets as a counterpoint to their analysis of raids against lightly defended points. In their follow-on report, S 197, six attacks had been analysed.26 The majority of these raids had used groundmarking with illumination by flares and five of the six had been controlled by a master bomber. The other had used a combination of Oboe and H2S marking. It was felt that, while techniques were not standard across the sample, the results were valid and they were not all that good. As described above, 37 per cent of the sorties had been ineffective either through early returns, gross errors, or other problems. Of those who bombed effectively the systematic error was 1080 yards and the overall average error was 1460 yards. While it was noted that this was more than double the error for lightly defended targets, it was still much better than the 2550-yard average from the previous year. The report offered a number of reasons to explain the difference between the figures for the summer and the year end: increased defences over Germany; the greater size of the targets; the fact that targets were found and marked using H2S rather than Oboe; and the need for a larger number of TIs

Round-the-Clock: April 1944–May 1945 213

which would make the bomb aimers’ ability to define the aiming point more difficult. If ever there was a reason to have told Zuckerman to back off in his criticisms of the Command’s reluctance to engage precision targets this could easily have been it. At the end of September, while this work was going on, the SASO asked Groups to adopt, on a limited basis, a new method of assessing bombing accuracy.27 The ORS had developed a system which used the photo frame taken at the point of bomb release, rather than the later bomb strike frame, to help identify gross errors which resulted, said the scientists, in about 15 per cent wastage on precision targets. The study explained that this new system of assessing bombing accuracy for daylight ops was needed, since the existing system did not offer sufficient definition when assessing precision targets. The new approach, while more labour intensive than the existing system, could be used to analyse what was going wrong with those crews or sighting units that were constantly producing problems. Groups were asked to try the system on this limited basis and report back. 4 Group was the first to reply, having used a prototype of the system since early August, and said that they had not only introduced the system on a full-scale basis, but that it was much more useful than previous methods and caused little additional work.28 6 Group reported back several days later with similar views. They said: please be advised that the assessment of bombing from the second (i.e. release) frame is now used exclusively in the Bombing Sections of the Squadrons of this Group. It is considered that this method ... although by no means entirely free from error, does provide results which are more informative and reliable than those obtained from the third or fourth frame. It is therefore the intention to continue to use this method of assessing individual bombing accuracy in daylight operations undertaken by this group.29

Similar feedback were received from 1, 5, and 8 Groups, although the latter preferred to stick to the recommended use for review of errors only due to the additional workload placed on its photo interpreters.30 All of these techniques were adaptations in one way or another of existing night tactics, but the use of G-H for formation attacks was something new. During the summer, 3 Group had been equipping with G-H, and while the transition was slower than expected, the Group was ready for its first operation into Germany by the beginning of October. Curiously, while the new system was supposed to provide a precision capability, all the proposed targets for the Group were area bombing objectives,

214 The Science of Bombing

like town centres. Bonn was such an example and the city was attacked on 18 October. The weather en route was completely overcast and there was significant cloud over the target. However, the Group approached in a by-now organized formation and bombed using G-H indications only, which resulted in ‘heavy and concentrated damage.’31 Yet not all G-H attacks were as successful and by the end of the year the accuracy of G-H was under some suspicion. In reviewing five raids where sufficient photos had been taken for analysis, the evidence showed inaccuracies that were greater than earlier results, with average errors of over 1000 yards. It was concluded generally that G-H was not as reliable as Oboe or visual groundmarking in daylight raids, although it was felt that a force of 70–120 aircraft could achieve sufficient weight of attack to ‘devastate a built up area.’32 3 Group responded to the criticisms of G-H in a letter in early 1945.33 The Group pointed out that, unlike PFF Oboe crews, the squadron crews generally flew a shorter tour than the pathfinders and as a result their level of experience was lower. As well, it was noted that 3 Group switched back and forth between G-H day raids and regular main force duties at night so it was harder to maintain proficiency. Moreover, it had proven impossible to use the G-H ground training simulator. Finally, the Group reminded the boffins that while an Oboe operator sat safely in a control room in England the G-H operator was himself in the thick of it with many distractions potentially throwing off his accuracy. These criticisms were all well founded and served as a reminder of the risks which had not disappeared from the average crew’s day-to-day existence. While mounting several raids on a given night did tend to reduce losses overall, it did not mean that the night fighter force had become ineffective. The Luftwaffe’s day defences had been neutralized by April thanks to the introduction of long-range American fighters such as the Mustang, but the same did not hold at night. Only as the summer progressed, as their fuel stocks dwindled and as the advance of the Allied armies across France deprived the defenders of much of their early warning network, did a noticeable drop occur in the effectiveness of the defence. While in January 1944 Bomber Command losses had been 5.5 per cent, by June they were down to just over 2 per cent. This did not mean that the night defences were everywhere defeated. When they did catch the stream they could and did inflict serious losses such as the 3 May attack on Mailly-le-Camp by 362 aircraft with 11.3 per cent losses and a raid against Lille one week later that lost 13.5 per cent of the attackers. By comparison five other raids on the latter night came away with no losses.

Round-the-Clock: April 1944–May 1945 215

Results for June were similar, causing Portal to state to the prime minister that there remained an ongoing need for improvements in tactics and defences to counter the enemy’s effectiveness. Portal also suggested that an increase in daylight operations might prove beneficial. This shift was already underway and by August approximately 50 per cent of all sorties were day trips. In September, of seventeen thousand sorties some two-thirds were daytime flights. Despite the shift to day flying, loss rates remained in the 1 per cent range, and while damage from flak increased greatly during daylight attacks the actual loss rate was not affected. Perhaps more important, August and September saw the sudden drop of losses at night over Germany. The overall night loss rate for August was 1.2 per cent and for sorties against Germany it was just 3.7 per cent; the corresponding figures for September were 1.5 and 2.2 per cent. These shifts reflected the decline in German defences as a result of fuel and early warning deficiencies.34 A number of steps were taken to reduce night losses during the spring and summer months, although some tactics were more effective than others. Intruder (RAF night fighter) and jamming missions mounted by the newly formed 100 Bomber Support Group were helpful, but limited in scale. While there were insufficient night fighters to neutralize their German counterparts, jamming was frequently effective, with the use of airborne Cigar (ABC) to blot out voice transmissions of German GCI controllers and Mandrel to neutralize Freya early warning radars. Nevertheless, these advances would not likely have had any long-term success had the Germans not been faced with a long list of challenges associated with the June invasion and the resulting loss of much of their air defence network over the following months as the Allied armies moved east.35 Another layer of the Command’s defence was the ability to detect threatening fighters before they attacked, and here the scientists were well into the analysis of the relative merits of various equipments designed to provide the crews with on-board early warning. After the exasperating delays in getting Monica into service there were now concerns that its emissions might be providing the Germans with a signal on which to home in. The ORS had compared the performance of visual Monica (a version of Monica with a visual warning indicator) and Fishpond (a different warning device based on H2S transmissions) in March.36 Of the two, Monica had the better record and there was ‘as yet no good evidence that MONICA is being deliberately jammed or used for homing.’ This conclusion was to change by the end of the summer. On 19 July Dickins advised the D/CinC that aircraft equipped with Monica now

216 The Science of Bombing

had a higher loss rate than those without.37 There were some specific cases yet to be looked at and this was being done, but the results showed losses of Lancasters with Monica at 8.3 per cent compared to 5.4 per cent for aircraft without the device; the comparable figures for Halifaxes were 7.6 and 3.4 per cent. As a result, various staffs, including the ORS, were examining the practicality of turning on H2S, Monica, and AGLT (a radar-equipped automatic rear turret) only on approach to the enemy coast as there was concern that the enemy was able to plot and home on these transmissions. The CSO put forward a practical caution, however. Switching on H2S and AGLT while airborne was not possible, since neither could be warmed up in the low temperatures found at altitude and H2S could suffer from icing of the scanner. Technical investigations to work around these problems were in progress.38 From these indications it did seem fairly certain that the Germans were using Monica emissions for their own devices, and consequently in the coming days the boffins prepared a study, ‘Appreciation of the Effect of the Use Made by the Enemy of Transmissions from Bombers,’ in which they identified three ways that the Luftwaffe was exploiting both H2S and Monica transmissions: (i) To plot the bomber force (by his ground plotting organisation); (ii) To enable his fighters to ‘home’ into the bomber stream. (By the use of FLENSBURG utilising MONICA and NAXOS on H2S). (iii) Possibly to ‘home’ on individual H2S and MONICA fitted aircraft.39

The boffins marshalled considerable evidence to support their suspicions, including the telling and obvious fact that Monica-equipped aircraft were experiencing higher than expected loss rates. They closed the report with a series of recommendations intended to segregate the use of Monica to certain Groups and/or raids so that the suspected weaknesses and risks could be better defined. At the same time, they strongly urged that alternatives to Monica be examined; these included a modified form of pulse from Monica so that it could not be homed on as easily, as well as an increased use of Fishpond and the rapid acceleration of the introduction of AGLT. The primary recommendation, intended to stop the immediate problems, was more basic, however: ‘if losses of Monica fitted aircraft for July operations are not appreciably lower than non-equipped aircraft its use should be discontinued.’ The solution, while not elegant, would at least be a lifesaver. The story did not end there, for within the first days of September

Round-the-Clock: April 1944–May 1945 217

the Air Staff’s attention was drawn to another alarming ORS report.40 Trials had been conducted using a captured German Ju 88 night fighter equipped with Flensburg and SN2 radars. While the SN2 could be neutralized by Window, the Flensburg set was shown to be able to actually home in on Monica. In commenting on the ORS material the CSO concluded that the jamming of SN2 would lead to an increased use of Flensburg and hence an increased risk to Monica-equipped bombers. While he felt that stopping the use of Monica was a partial answer he cautioned that this would leave the enemy free to home in on H2S transmissions. Some workarounds were contemplated to deal with that eventuality, but for the moment he supported the boffins’ recommendation to temporarily cease the use of Monica. The ORS report quickly found its way to Harris’s desk, with a recommendation from Saundby to stop the use of Monica. Harris had no hesitation in approving the scientists’ and his senior staff’s advice. Monica was not to be used from that point on; the next day, 12 September, Saundby directed that it be removed from aircraft.41 On the same day the Director of Air Tactics at the Air Ministry wrote to the Headquarters about German fighter attacks from below.42 DAT wanted to distribute information on upward-firing guns and on a number of proposed countermeasures. Their sources of information on these weapons were limited to data provided by prisoners of war and suggested that upwards of 10 per cent of losses could now be attributed to upward-firing guns. Before distributing the DAT letter the Ops staff asked the scientists to review it. Some three weeks later the ORS completed its examination. Smeed did not believe the DAT statement that a ‘high proportion’ of losses were the result of this new technical innovation; nonetheless, he said that the scientists would keep close watch. Attached to Smeed’s minute was a draft report which, unlike the DAT letter, was based on hard data and immediate crew interrogations.43 The inspection of particular aircraft, the authors said, was felt to be the most reliable source of evidence, and the data from crews and repatriated members of missing aircraft served to confirm the hard data. Among other conclusions the scientists stated that upward-firing guns posed ‘a very serious danger to bombers’ for two main reasons. First, the crew was less likely to notice the approaching fighter and if they detected it they would be unable to engage it under most circumstances. Second, an attack from below exposed much more of the bomber than an attack from behind. Such circumstances had contributed, the scientists were confident, to 7 per cent of the confirmed losses and 10 per cent of the non-lethal attacks for 1944. Smeed was not particularly happy with the DAT suggestion that an

218 The Science of Bombing

early warning device would make the situation less troubling. This sort of thinking would be likely to lead to a demand from the squadrons for a reintroduction of Monica, which he felt was not then justifiable. But he did point to the fact that H2S was now showing some returns from enemy aircraft approaching from below and suggested that the radar might be employed for its early warning potential. This suggestion was found impractical by the Radar staff, but did lead to a recommendation for employing Fishpond for this purpose.44 Discussion now developed between the Ops and Signals staffs on how to get this information out to the Groups, and it was ultimately decided to issue a short factual statement based largely on the data provided by the scientists, this being done by the SASO on 25 October. A second note issued two days later describing how best to use Fishpond was signed by Saundby himself, who had been kept informed during most of the staff discussions.45 It was also thought appropriate to conduct trials at the Bombing Development Unit (BDU) to see how well the proposed defences might work. DAT had been kept informed of the staff’s thinking and it was they who tasked the BDU, through Bomber Command, to conduct trials of the efficiency of Fishpond and other tactical counters to the enemy. 46 Here was an instance where the scientists were able to confirm doubts held by the operators. The ORS had been able to rapidly conduct a detailed review of the issue and give the Ops and Signals staffs the necessary data to allow them to make a number of informed decisions. Had they not been tracking the efficacy of Monica through the preceding months, it is clear that the Service staff would still have had doubt and concerns, but neither the evidence nor the strong recommendation to take drastic but necessary action. In mid-November the boffins provided Saundby with additional data on attacks from below. They had now been studying the matter for some time and were even more convinced that damage reports, as opposed to crew reports, gave them a much more accurate depiction of the real circumstances. While crew reports had placed attacks from below at about 40 per cent of the attacks, the damage evidence moved that figure up to close to 60 per cent. At the same time the boffins added information on upward-firing guns, showing that the system was being used in about 7 per cent of attacks. Harris quickly directed that a letter be prepared for the Under-secretary of State in which he firmly refuted, for a second time, lower estimates which had been accepted by the Ministry almost a year earlier. He set a figure of 65 per cent for attacks from below, including the 7 per cent from upward-firing weapons. But he also added the

Round-the-Clock: April 1944–May 1945 219

comment that a large percentage of aircraft losses could be presumed to be caused by attacks from below. Given these figures, he asked that any decisions on aircraft design, and particularly armament and visibility, be revisited with urgency in light of this new data.47 But all of these developments were still several months into the future. The night bombing of Germany which had been bloody in June resumed in July with much lower casualties. Losses among 3419 sorties flown that month were held at 3.9 per cent, which compared favourably with those of the preceding year.48 In August the missing rate dropped slightly to 3.7 per cent, but this did not tell the whole story as not only were many of the raids deeper penetrations over Germany, but also the Allied ground advances meant that the defenders, having been pushed back, were all the more concentrated in the path of the raiders; the chances of losses had, in other words, increased while in fact actual losses had dropped. Indeed, despite the increased density of defenders, the September loss rate dropped further to 2.4 per cent; the failure of the German night fighter defence was beginning. This collapse was the result of three factors: loss of territory, loss of fuel stocks, and the development of enhanced radio countermeasures.49 The effective use of Mandrel and Window to jam early warning and GCI radars made it hard for the controllers, with little depth in their defences and hence only a few minutes to intercept the stream, to get the fighters vectored against a raid. Additionally, the Command was now regularly launching spoof attacks and making false transmissions simulating raids such that it became almost impossible to know if there was a raid behind the Mandrel screen or not. At the same time, as fuel stocks decreased the defenders became all the more cautious of wasting their scarce resources. The result was that some raids were not engaged or engaged so late as to make the attempt pointless. The corollary of this was that within Bomber Command losses dropped and crew experience rose, as did the confidence of the flyers and commanders. There was a strong synergy of success and the accuracy and efficiency of the raids contributed even more to this.50 It will be recalled that the issues surrounding daylight attacks against Germany had been considered by the Tactical Planning Committee in late June. By the end of August it was thought possible to attempt such attacks and five raids were mounted in the following weeks. Of some 780 sorties only 10 were lost and these to flak; heavy fighter escorts encountered no German defenders.51 In fact, the whole question of losses to flak was under review, and while the numbers were dropping the scientists were still keeping a wary eye on what could be a significant threat

220 The Science of Bombing

to daylight ops.52 The use of flak against daylight raids, first studied by the boffins in early 1943, was now reviewed again. While evidence was limited, it was still possible to compare formation and non-formation attacks on V rocket launch sites. The conclusion reached was that there was little difference in actual losses, although the formation attacks had an appreciably lower flak damage rate. In reporting these findings to Dickins, Smeed suggested that there had been doubts about the formation tactics. Formation flying ‘is no more liable to cause our aircraft to be hit than non formation flying of the concentration usually used.’53 Smeed referred the Air Staff to these findings later in the fall when A/ Cdre Ops asked to see figures on daylight losses and damage.54 A short paper reviewing the losses for the June–September period was also sent forward. Losses and damage were noticeably higher in 5 and 8 Groups, the former because of its daylight precision attacks requiring long runins to the target and the latter because of the need to remain in the target area for extended periods. It was estimated that three-quarters of daylight losses were due to flak while the most of the others resulted from being struck by falling bombs and from collisions; few losses were attributed to fighters. Overall loss rates were generally less than 1 per cent for the main force groups. There was one final round of review of daytime losses to flak. At the end of November Dickins received a paper by Sir George Thomson, the Scientific Advisor at the Air Ministry. In it Thomson proposed that the loss rate to flak could be halved if the stream could be shortened by broadening the width from about two miles to three. There would still be opportunity to narrow the front for bomb release by having the attackers converge when about three to five miles from the target. Dickins felt that the logic was solid and put the proposal forward to the Air Staff.55 But for once, the response from the Ops staff was not positive. Major counter-arguments included the fact that a fighter escort was more effective on a narrow front; moreover, there was already a fair bit of ‘jostling’ as the gaggle approached the target and this new tactic would only make accurate bombing that much more difficult. Constantine, adding his concerns, kept his remarks short, saying that ‘Thomson’s report is based so much on theory that it has little relation to practical operations.’ He added that the matter was already under review with 11 Group (of Fighter Command), which provided the escorts.56 The last word went to the SASO, with whom the D/CinC agreed. With reference to [Dickins’s and Thomson’s remarks], I consider that our

Round-the-Clock: April 1944–May 1945 221 losses in daylight operations due to flak have so far been extremely light and certainly compare favourably with the Americans. ... Groups are doing everything possible to reduce the length of the ‘gaggles’ in order to eliminate entirely the ‘tail.’ This will automatically tend to increase the width of a ‘gaggle’ which is what Sir George Thomson is after. So long as a ‘gaggle’ is our accepted formation (a ‘gaggle’ implying that each aircraft is flying and bombing as an independent unit as opposed to a set place in a fixed formation), I do not think that much more can be done to ensure that aircraft are placed in the best possible theoretical position in regard to freedom from flak. It is largely a question of method and training and I feel that many Americans would gladly give up their set formations for our ‘gaggles,’ but their operational and pre-operational training makes this impossible in the same way that [it] would be difficult for us to change to their or any other set formations.57

A/V/M Walmsley, the SASO for day ops, met with ‘Sir George’ and Dickins to explain the situation. He noted that Thomson seemed satisfied with the decision.58 As the losses to German defences began to fall, the impact of other sources of losses on the aircrew became more noticeable, and the boffins now spent considerable time dealing with these causes as well as the issue of tour credits and second tours, for the fact was that the senior leadership was not about to waste lives. As early as April 1944 Dickins had had the boffins look at the relative value of sorties over France and Germany. Smeed provided a detailed reply explaining that current thinking was that a sortie over occupied territory equated, in calculating the thirty-trip total, to one-third of a trip by comparison to the high risks involved of ops over Germany. A quick review of losses rates (5.3 per cent over Germany and 1.8 per cent over France for the preceding four months) confirmed this weighting, but Smeed made a convincing case for review of the policy should the upcoming invasion see a higher concentration of ops over France and with it a correspondingly higher loss rate. If losses over France rose to 3 per cent and sorties over the area were counted even as one-half of a Germany trip it would still reduce the tour survival rate to 17 per cent. In these circumstances he argued that consideration be given to counting French sorties at a two-thirds rate. 59 As the year progressed, loss rates remained a concern to Harris and he had asked Dickins to review actual losses against what the boffins had predicted.60 The CinC was particularly interested in the chances of com-

222 The Science of Bombing

pleting not just one, but two operational tours. The first review done had used a 4.5 per cent loss rate – the average sustained during 1943 – and this corresponded to a 25 per cent chance of completing the first tour of thirty trips and a 10 per cent chance of getting through the additional twenty sorties of a second tour. The scientists had now compared their predictions against the actual figures from 1 and 4 Groups; the corresponding first tour results were 30 and 23 per cent respectively. Taking the average of 27 per cent, the scientists felt their model was reasonably accurate. Harris, however, was more concerned about the second tour figure and this because of a larger argument with the Ministry. He was in a battle with those seeking ways to reduce the personnel overhead. London wanted flyers who finished their first tour to be posted to a second tour. In a letter to the Under-secretary on 9 August Harris protested, saying that the casualty rate in Bomber Command exceeded that of any other command.61 He indicated that the survival rate was just one in ten and felt that a second tour would reduce the chance of getting through the war to next to nil. In the middle of September Dickins became aware of Harris’s letter and the data he had used. He minuted the CinC immediately, spelling out that the one in ten figure that Harris used was for attacks on all targets, but that the survival rate over Germany was in fact only one in fourteen.62 This was a more accurate figure, thought Dickins, as it was based on solid long-term data and could well be indicative of the winter to come. The following day Harris used this new information to once again seek a change in policy from the Ministry, saying bluntly that ‘the arguments set out in my previous letter against a compulsory second tour in Bomber Command are, therefore, considerably enhanced by this new figure.’ On this basis he asked that the two-tour concept for bomber crews be scrapped.63 Another long-standing concern which now received additional attention from the scientists was the matter of rescuing those flyers forced to ditch their aircraft in the North Sea. Towards the end of August the scientists prepared a report on the effectiveness of the Air Sea Rescue service.64 It appeared that the number of ditchings was actually higher than had been previously thought, accounting for as much as 15–20 per cent of losses. It was concluded, too, that crews were not well prepared for the eventuality of having to ditch: often aircraft did not seem to be sending out SOS messages and in some instances crews did not use their survival equipment effectively. It was apparent that there was little Air Sea Rescue (CASR) could do to save crews if they themselves did not take the first proper steps in their own rescue. Dickins asked that the file

Round-the-Clock: April 1944–May 1945 223

be passed to the senior Air Sea Rescue staff for comment before sending it forward to Saundby. W/C CASR subsequently provided a detailed response indicating overall that the paper had captured the main themes and concerns that were already well known to those working in the rescue service. Some limitations in the paper’s logic were identified, however. It was felt that, just because distress calls were not often received, it could not be concluded that the calls were not sent: transmitter problems, low altitudes, frequency congestion, and wrong frequencies meant that crews could not always succeed in sending their distress messages. As well, many ditchings, particularly at night, were catastrophic and survival equipment could not be got at in the few seconds available before the aircraft sank. Finally, it was noted that some ditchings were either ‘impossible or impracticable’; the former were those instances where a crash landing would have been unsurvivable whether over water or land and the latter dealt with those cases which took place so close to the German coast as to make rescue unfeasible. Having made changes based on these observations, Dickins passed the report to the D/CinC on 20 October 1944. The scientist suggested ‘that the report be sent to Stations with a covering letter emphasising the importance of crews sending out distress messages as soon as they realise they are in trouble.’ Saundby noted his agreement.65 Once again the boffins had identified an important contributing factor to losses and focused efforts on viable solutions. Collisions, too, were a sort of self-inflicted loss – one which had been many times reviewed, but was again raised as a problem by the Ministry in the fall. The boffins had been aware of the collision risk for some time and in September an M series study was prepared for internal use only. Further work was done on it during the fall months, and a much-revised document covering night and day operations from April to December 1944 was issued in February 1945.66 This document acknowledged the earlier version and built upon its conclusions. In summarizing the collision experience of the Command, the scientists said that the rate of mishaps had been steady at about 0.13 per cent until December when it had increased to 0.25 per cent. In terms of losses alone it was estimated that in December and January, of some two hundred sorties that had gone missing, fully thirty-three were due to collisions. The increase was thought to be due to increased accuracy in track keeping; in other words, more aircraft were managing to stick to the correct track than had previously been the case. On short-range raids the collisions seemed to take place at a constant rate throughout the raid, while during long-distance raids collisions were more prevalent

224 The Science of Bombing

during the enroute phases of the mission. The rate of collisions, the authors said, could be reduced by increasing either the lateral or height spreads of the raid and they intended to discuss those notions in a later paper. There is no indication from the file who saw the report, but there had been considerable badgering about collisions from the Ministry in the preceding months and personnel in London were much interested in finding ways of reducing the rate, which they believed to be at about 0.5 per cent of all sorties.67 ORS staff had firmly explained to them that reducing the concentration would only lead to greater losses due to enemy action. Now they could add that the rate was at most just half of what had been believed.68 While there was no intention to dismiss the problem, the boffins had again brought accurate data to light so that a properly informed discussion could take place. Collisions were sometimes the result of human error and it was only too obvious to the senior leaders and to the ORS that flyers could make mistakes either through error or negligence. It was to underscore the latter problem that Walmsley sent out a stern missive in December.69 He reminded the Groups and ultimately the individual crews that operation orders were just that – orders. He cited two cases, one where the orders had been obeyed and only one loss had occurred, and a second where there had been a flagrant breach of the raid order and somewhat miraculously only three aircraft had been lost. In the latter case, he said: ‘Occasional nonadherence to these orders by some of our aircraft makes it obvious that a proportion of our aircrews either do not appreciate the great importance of these tactics or else they are guilty of criminal disregard for the needless danger to which they expose not only their own but all the aircraft participating in the operation.’ He went on to describe the two raids. Against Duisburg on 30 November/1 December electronic silence and altitude restrictions had been imposed for the first part of the flight to reduce the chances that the Germans would have any early warning. Evidence from the raid and a similar simulation the following night indicated that some sorties of the previous night’s raid had been operating in contravention of these restrictions. By comparison, the raid on Hagen on 2/3 December had gone off exactly as planned from a security point of view. Looking at the data used there is no doubt that much of it had come from the boffins’ night raid reports for these two ops. Sorting out the negligent crews was clearly a chainof-command responsibility, but this breach would not in all probability have been detected in the first place had not the scientists provided their typically comprehensive and immediate follow-up of operations.

Round-the-Clock: April 1944–May 1945 225

With the Command capable of effectively attacking any target, the official history says: ‘The question in 1945 was no longer what could be destroyed but what ought to be destroyed.’ 70 Nevertheless, this final phase of the bombing war had not started off so positively. April 1944 had seen Bomber Command stymied after a long winter of deep and frequently less than effective raids, raids which had suffered damaging losses. With both elements of the overall efficiency equation so far below their optimal level, the Command itself had needed new ways of operating in order to recover. As the staff and the scientists began to review and modify the tactics, however, larger forces had intervened, giving them new problems to address, but also more time to resolve long-standing issues. The boffins had played, as was by now the norm, a central role in helping the Command as it shifted first to tactical targets and eventually to day tactical and day strategic bombing. Dickins himself had been one of less than a handful of seniors at High Wycombe to be shown the first conceptual documents for Operation Overlord, and the scientist’s calculations formed a basis for Harris’s formal comments on Bomber Command’s contributions to the pre-invasion bombing. The accuracy of the data and recommendations was such that Harris would call upon the scientists later in the year when vexed by the prospect of taking on the oil production system as a war-winning target. While these political issues were being played out, the Section had continued to assist the specialist staffs in finding solutions to the limitations of H2S, joining forces with the Navigation and Radar branches to coax better performance out of the equipment and the aircrew. The boffins had also reported on the effectiveness of other bombing techniques and of the tactics for smaller but at the same time more precise raids. The scientists had been equally occupied with keeping losses down, for while the actual loss rate fell dramatically in the fall of 1944 and stayed in the 1 per cent range, the senior leaders remained committed to protecting every flyer they could, for such was the responsibility of command.

13 Conclusions

When Bomber Command was given its motto, ‘Strike Hard, Strike Sure,’ in 1936, the proponents of strategic bombing had, for close to two decades, been preaching that this form of warfare was certain to bring victory. They had been so convincing that in 1932 British Prime Minister Stanley Baldwin had stated: ‘The bomber will always get through.’1 Yet, for the Royal Air Force and the British government the events of 1939, 1940, and 1941 suggested anything but this certainty. By the middle of 1941 the record of Bomber Command was so alarmingly poor that Churchill had commissioned a special study to determine the nature and scope of the problem. The Butt Report made the extent of the troubles somewhat clearer and hastened, while not precipitating, the establishment of the Operational Research Section at Bomber Command Headquarters. From September of that year more than fifty scientists and their support staff, under the direction of Dr Basil Dickins, took on the task of making the conduct of the bombing campaign as efficient as could be. In its broadest sense, their mandate allowed and required them to investigate any circumstances which might contribute either to inaccurate bombing or to losses which could be reasonably avoided. The two concerns were the essential elements of an ‘equation’ for the efficient use of limited resources: maximum bombs on target per aircraft lost. Their work went much further than this. Over almost four years the OR scientists were involved in hundreds of studies and investigations which looked at all aspects of how the Command planned and conducted its raids. They considered the effectiveness of training, of navigation devices and practices, of target recognition methods, of aircraft performance, and of enemy air defence capabilities and the means to neutralize the Luftwaffe’s fighters and flak. But they went beyond simply reporting

Conclusions 227

on what was observed: they often made significant proposals for new tactics and techniques which would give the Command an edge in getting to the target with fewer losses. Within weeks of the Section’s activation in the late summer of 1941 Dickins and his initial cadre had been called upon to comment on those conditions which appeared to promote or hinder accuracy. Many of these had been sensed by the flyers and the staff officers, but no one had put them to scientific scrutiny. This research was a first and major step in sending the raiders out with some chance of finding and bombing the right place. The scientists were also quickly drawn into the debate over ways and means to reduce losses. Concentration of attackers in time and space, a notion that, once again, had been debated by flyers and headquarters, could now be given at least preliminary confirmation as a means to overwhelm the defenders and thus reduce losses. Such was the power of science that just months later Dickins would convince Harris that crowding one thousand bombers into the darkness over Cologne would both overwhelm the defences and be virtually without risk of collision. Cologne, too, saw the exploitation of another of the scientists’ innovations. It was the boffins who, in the opening days of 1942, had been given the task by the Commander-in-Chief himself to devise the optimal method for using the new and untested navigation device Gee. The scientists had achieved these results all within their first year in the Headquarters, but the list of their contributions would continue to grow. Problems with the Halifax’s and Lancaster’s escape hatches, and losses due to engine and fuel tank fires, for example, owed their recognition and definition to the scientific observations of the ORS. Precise analysis of each raid began to reveal those techniques for navigation and target identification that were working well or not so well. While supporting, and according to some evidence originating, the notion of a pathfinder force, the scientists warned that even the use of picked crews to mark the aiming point was not an infallible process. Equipment, particularly H2S, had to be held responsible when it let the PFF down, but errors by the crews themselves could not be ignored. The boffins were not reluctant to point out when the bombers were turning home without having come anywhere close to the target, and Dickins’s staff kept close watch over losses, comparing experience levels and particular circumstances to give the leadership clear indication of where problems were to be found. Much of the Section’s work would fall into these sorts of patterns over the first two and a half years of its mandate, and the scientists built up a

228 The Science of Bombing

significant repository of data and observation which would stand them and their leaders in good stead during the last twelve months of the European war. When, in 1944, circumstances obliged the Command first to operate in twilight and then by day, while at the same time taking on small tactical targets requiring precise bombing, the scientists were ready to make recommendations that would maximize the efforts of the crews, while at the same time minimizing the risks they would have to face in these new conditions. While conducting their research the scientists were neither working in isolation, nor even working in their own environment. These were not laboratory studies, nor were the disciplines involved those in which they had long experience and expertise. From the first day they were thrust into a foreign, tribal, professional environment where the conventions, practices, and even the lexicon would have been almost wholly unknown to most of them. As a result, the success or failure of the Section would depend on the scientists’ individual and collective ability first to acclimatize themselves to their surroundings and then develop and demonstrate a mastery of the issues such that their contributions would be acceptable and accepted. It is clear from the evidence that not only were they able to learn the ropes of bombing operations, but that their opinions came to be sought and valued. The evidence shows that time and again it was the scientists who either on their own or at the request or direction of their seniors put practices, personnel, and equipment under intense scrutiny. In doing so, they worked at High Wycombe with a number of specialist staffs as well as the senior leadership, while within the Groups they often worked alongside the commanders. They did not always agree with the service position on an issue, nor was their counsel universally accepted, but the latter circumstance also befell uniformed colleagues. In sum it could be said that they were, but for their tweed and pinstripe, fully integrated into that organism which was a large and complex headquarters. At this juncture, it is important to acknowledge the criticism levelled by some of Dickins’s contemporaries and subsequent writers that he was too much the ‘yes’ man, too pale the civil servant, unwilling or unable to stand up for his beliefs or even to stand up for hard science. One has to ask, however, how these commentators knew such criticisms to be true. Did they personally observe the scientist, read the material that he passed to Harris, participate in the discussions with the CinC, the D/ CinC, and the other seniors, or did they make their judgments on less than complete evidence? It is true that there were meetings where Dick-

Conclusions 229

ins should probably have been a key participant but where the minutes give no indication of his having spoken. Yet how do we know that he was ineffective? When he accompanied Harris or Saundby, his role, like that of any staff officer, would have been to prepare his commander before the session and then to speak only when asked. We are told that Dickins, on at least one occasion, was prepared to debate an issue with Harris, over target selection for the first thousand bomber raid where the scientist is reported to have insisted that the attack be within Gee range. If he was this determined on fundamentally important questions, then he must be seen as a loyal and effective adviser. While not a part of this study, it is important to recognize the limits of the scientists’ influence on strategic issues. Simply stated, the ORS at High Wycombe was not involved in setting the policies which determined how the Command was to be employed. The fact was that Bomber Command had been given a remit for area bombing before the section was established. The scientists’ task was to make this bombing as effective and efficient as possible. Indeed, their efforts, by the end of the war, had contributed to the establishment of a near-precision capability, one regularly as accurate as the precision bombing ability that the Americans had always claimed. Was this precision their intent? Applying logic one would have to conclude that at least implicitly it was, for surely precise bombing maximized one of the two elements of the efficiency equation. There is an element of irony in this coming-full-circle of bombing accuracy. Bomber Command had not set out to be an area bombing force, yet for most of the war this was the best stratagem that could be managed given the technology, personnel, and training available, and conflicting priorities placed on the Command. Aerial bombing had in the early months of the war proved to be no precise rapier, but a rather unwieldy, splayed, and bent pitchfork. Those wielding this makeshift engine had no choice but to adapt to the realities of their situation while at the same time seeking to reforge their weapon into what they had thought it to be. For critics to argue that, because German war industry did not reach its peak production until the last year of the war, Bomber Command’s area bombing was therefore nothing more than a terror campaign is to presume that Harris and his headquarters, along with the political leadership of the Allies, knew all along of Germany’s circumstances. It is also frequently argued that Harris was incapable of dealing with technology and was insensitive to the plight of his crews. Yet there has certainly been evidence to contrary. His was not a campaign of mindless

230 The Science of Bombing

frontal assaults, of sending out forlorn hope after forlorn hope against the German fortress. It has been shown without any possibility of misunderstanding that Harris and his senior leadership understood explicitly, normally thanks to the boffins, the very nature of the challenges facing them and of the solutions available. Where tactics worked they were used as effectively as they could be, always with an eye to the eventuality that the Luftwaffe would develop effective countermeasures. When such countermeasures did materialize the boffins and the commanders were ready to end those practices that had become ineffective. As well, Harris’s headquarters was one where experimentation was generally, if not always in use. Techniques and tactics, as well as equipment, were modified to get the best return on material and human investment. Harris and other senior leaders were not just interested in minimizing the losses that they recognized could not be sustained, they were also mindful of the fact that crews could on occasion be either reluctant to press attacks or incapable of using the resources given them. In dealing with these human issues, as in all matters, it was often the scientists who put the organization’s finger on the causes, as in the case of early returns during the first Battle of Berlin, and who initiated corrective action through the chain of command. Harris might have been a forceful personality, but he was not a ruthless and authoritarian Luddite. Rather he was a commander who, in dealing with an unimaginably difficult and multifaceted set of confounding challenges, was not reluctant to seek and accept assistance. His thoughts on the contribution of the scientists bear repeating, for they can be easily overlooked in even a comprehensive study of his memoirs. ‘Under brilliant young Dr. Dickens [sic], Bomber Command’s Operational Research Section’s investigations always enabled us to know exactly where we stood’ (emphasis added).2 The boffins were, he wrote, ‘a body of brilliant young civilian scientists and technicians at Bomber Command headquarters, who did work of inestimable value in subjecting all aspects of our operations to an impartial scrutiny.’3 But what of Harris’s own powers of analysis and reasoning? And by extension, what do we know about the intellectual capacity of other air commanders? There is much more work to be done to develop a full or even adequate understanding of their thinking and decision making. We need a better investigation of Harris’s operational intentions, for while we have a good understanding of his strategic and political debates with peers and seniors and an equally complete account of how various raids and campaigns played out, we cannot claim to have a good understand-

Conclusions 231

ing of what he planned to do. Equally, in this volume we can only start to see how he adjusted his thinking when confronted with results other than what he anticipated. Indeed, this work has, like some other studies, demonstrated clearly that the bombing campaign, and the battles and raids within it, were every bit as intellectually demanding as any campaign conducted by land or naval forces. The ability to think critically, to deal with the complex and the ambiguous was a precious commodity. That there were not the physical privations and dangers of living in the field or at sea should not be taken to mean that the job of balancing this vast array of confounding issues was not stressful. As Harris has said, there was no option of pulling the Command or the Groups out of the line. On virtually every night there was the possibility of a new engagement; if the weather was bad over the primary objective there was likely another route or target which would allow operations to go ahead and so with little notice the staffs and the flyers would have to prepare for and put on an operation over Germany and Occupied Europe. Small wonder perhaps that there was so little opportunity for the senior leaders either to gain some respite or to reflect on what was going right or wrong, and all the more need for the boffins. Not only did the scientists have the time and skills for data analysis, but they soon developed a sense for what should and should not be coming back to them as results and losses, and what the flyers could achieve. In this way the current study begins to reveal the intense thinking that went on behind the campaign, allowing a glimpse, at least, of the complexity of the mechanism that was Bomber Command Headquarters. The Headquarters has been called both an organism and a mechanism in the preceding paragraphs. Perhaps headquarters are both. They are in one sense a set of processes or a mechanism, but they are at the same time governed by the people who work in them. Indeed the manner in which the processes are, or are not, carried out is determined by the personalities of the staff, and the relationships among those personalities. But we have very little insight into either aspect of military headquarters. In this case, for instance, there is no definitive organization chart to show, formally at least, how the various functions of the Headquarters were linked to the chain of command at High Wycombe. That this is the case for Bomber Command is not too surprising, for there are very few headquarters for which we have this level of detail. Yet without an understanding of how actual headquarters have functioned, succeeded, or failed, it is difficult to understand what made them work then and makes them work today.

232 The Science of Bombing

This is not to say that historians and military professionals do not have an understanding of staff systems – the two commonly recognized models being the continental staff system and the British general (operations) and adjutant/quartering (administration) diarchy. The latter is not that much different from the system employed by the RAF and by Bomber Command; the Senior Air Staff Officer running the operational side of the headquarters while the Air Officer Administration took care of everything else. Indeed, it is not a stretch to assume that the RAF system evolved directly from the system learned by the senior members of the Royal Flying Corps in the previous war. But what of the people who inhabit the various ‘staff appointments’ and their associated duties and responsibilities? The 5 Group comment seen at the beginning of this study that there were no fixed organizations because headquarters structures were subject to the competencies of the personnel occupying the positions is somewhat surprising. We are wont to believe that military organizations are carefully thought out and populated with well-qualified individuals used to carrying out their assigned functions and to working smoothly and effectively within a collective profession. This clearly is not always the case. Military organizations are in this sense no more or less blessed with both strong and weak performers than any other segment of society. Sometimes a commander is lucky and finds the staff well suited to the tasks at hand; at other times the circumstances will be less ideal. This, of course, is not to say that the staff will not sometimes be saddled with a senior who is less than capable. Understanding headquarters and coming to terms with how they work is as critically important today as it was then. While the problems of that war were in no way insignificant, they appear relatively straightforward to commanders and staffs who today are challenged by small wars which contain strategic implications in every skirmish and incident. How best to deal with this level of complexity, how to resolve problems which appear without resolution, is the challenge which these men and women face and it is the challenge at which they cannot fail. If works like the present study are to be of some use to today’s sailors, soldiers, and aviators, then these investigations must seek to illuminate the inner workings and even the social interdependencies of headquarters, to expose and make comprehensible commander-staff dynamics which can be used to combat the fog and chaos of war. The workings of headquarters and the demands on the intellects of senior officers point to one other reality: the military education system which underpins the institution must be capable of producing officers

Conclusions 233

who can think clearly in situations of almost unimaginable complexity. Had Harris and others not had the openness of mind and the ability to visualize and conceptualize which they did, then the performance of Bomber Command would surely have been much worse. As has been pointed out elsewhere, the ability to deal with ambiguity and competing factors is essential; the lesson for military educators, both then and now, is that any process of education must not only impart knowledge, but also take students outside their area of comfort, forcing them to build mental flexibility and giving them the thinking tools to be able to break down these problems into manageable parts, and synthesize viable solutions. While these examples and lessons can be valuable to today’s senior combatants, from the perspective of air force history, this investigation fills a significant gap. We have a broad and deep understanding of the strategic direction of the 1939–45 war, and more specifically of the strategic bombing campaign against Germany. There have been scores of volumes written about the direction and results of the campaign: three official histories, including the American; two official surveys of bombing results; several biographies of the senior commanders, including three of Harris; dozens of studies of the campaign itself and of specific raids or battles; and many critical assessments of the morality of bombing. Taken together these works allow one to appreciate the decisions made at various times that took the RAF through both the build-up to the war and through the various phases of the actual campaign. At the other end of the spectrum there are again scores of biographies, autobiographies, and memoirs of the individual flyers which give us an idea of what they experienced in their personal wars. These accounts allow us to see and occasionally share some individual episode, physical or mental, emotional or even spiritual. Such accounts give us a glimpse of the flyers’ war and permit us to tease out some common threads, but it is often hard to see the link between what went on in any particular aircraft on any given night and the grand strategy which had sent that crew into harm’s way. That link is the story of Bomber Command Headquarters, of its commander and Service staff and more particularly of the ORS. The scientists were the ones who, as the war progressed, came best, next only to Harris and a select few in the Headquarters, to understand the impact that prosecuting the policies could have on the aircrew. They were the staff officers charged with making the campaign as efficient as possible, and in striving to make good on that duty they came to understand, perhaps implicitly, the crucial responsibility of a commander and a head-

234 The Science of Bombing

quarters in balancing the weighty scales of life and death, victory and defeat. They more than anyone else could help Harris achieve his strategic remit while minimizing the losses to his flyers. They, more than anyone else, could understand the stresses of Harris’s command.

Appendices

Appendix 1 Appendix 2 Appendix 3 Appendix 4 Appendix 5 Appendix 6 Appendix 7 Appendix 8

Britain and Northwest Europe: Selected Bomber Command Targets Glossary of Operational Terminology Air Ministry and Bomber Command Headquarters Organization Charts Air Ministry and Bomber Command Appointments Air Staff Directive to Bomber Command, 13 July 1940 Principal Bomber Command Aircraft Bomber Command ORS Senior Personnel Bombing Techniques

This page intentionally left blank

Copyright © 2009. University of Toronto Press. All rights reserved.

Appendix 1: Bomber Command Targets 237

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

APPENDIX 1: BRITAIN AND NORTHWEST EUROPE

This page intentionally left blank

Appendix 2: Glossary

239

APPENDIX 2: GLOSSARY OF OPERATIONAL TERMINOLOGY 5 Group Visual

Target-marking technique. Flares and proximity marking followed by visual dive marking. These markers were offset by several hundred yards so that they were not obscured by smoke and dust; Main Force crews were given settings for their sights to allow for this.

8 Group Visual

Target-marking technique. Similar to controlled Oboe. Master bomber usually assessed Oboe marking and remarked the AP visually.

A/C/M

Air Chief Marshal (equivalent to a four-star general).

A/Cdre

Air Commodore (equivalent to a one-star general).

A/M

Air Marshal (equivalent to a three-star general).

A/V/M

Air Vice Marshal (equivalent to a two star general).

AGLT

Automatic gun laying turret. A rear turret device which identified, tracked, and fired on approaching aircraft using radar.

AI

Airborne Interception Radar. The use of a small radar carried by the fighter to home in on a bomber until it could be seen by the pilot. The radar would be operated by a second crew member. The range of the radar would only be a few miles and the aircraft would need guidance from a ground controller to get close enough to the target to acquire a radar lock-on.

Airborne Cigar

Cigar apparatus mounted in RAF aircraft. See Cigar below.

AOC

Air Officer Commanding. An officer usually of Air Vice Marshal rank who commanded a Bomber Group.

AOCinC

Air Officer Commanding-in-Chief. An officer usually of

240 Appendix 2: Glossary Air Marshal or Air Chief Marshal rank who commanded an RAF Command. Also referred to as CinC. AP

Aiming point

Backers up

Pathfinder aircraft which would drop additional target indicators during the course of a raid so that the main force crews would have a precise aiming point. The backers up would aim their TIs at the centre of the visible TIs; this technique was called Visual Centring.

Benito

German Night Fighter Control System.

Boozer GCI and AI

Warning device which alerted bomber crew that German GCI or AI radar was operating.

Carpet

Device to jam German GCI radar.

CAS

Chief of the Air Staff; the military head of the Royal Air Force.

Catseye fighter

A fighter where the pilot used normal eyesight to make the final approach and attack on a bomber.

Cigar

Jamming by UK ground stations of German VHF (very high frequency) frequencies used for direction of fighters.

CinC

Commander in Chief. In Bomber Command a commonly used abbreviation for AOCinC.

Controlled Oboe

Oboe-placed target markers were assessed by a master bomber who then directed the main force by radio and sometimes backed up the best-placed TIs with additional TIs of a different colour.

Corona

Issuance of false directions to German fighters by RAF personnel broadcasting on German frequencies.

Creepback

The tendency for successive waves of bombers to bomb before reaching the actual aiming point. This was often a result of the target area and TIs becoming obscured by

Appendix 2: Glossary

241

smoke, dust, and fire as the raid developed. As a result, the actual fall of bombs ‘crept back’ towards the direction from which the bomber stream was coming. Crossbow

Attack on German V weapon launch sites.

Dead reckoning navigation

This technique had been used since aerial navigation began and remained the basis for all navigation. The use of two vectors – the course and airspeed of the aircraft and the direction and speed of the wind – to produce a third vector, the track and groundspeed of the aircraft. The first vector was provided by the navigator and flown by the pilot. This was often easier said than done, given the difficulties in flying the heavy aircraft with precision and the evasive action which would throw off both course and speed. The second was forecast by the meteorological staff but was often imprecise, and both values varied with height. The easiest way to calculate actual winds was to compare where the aircraft would have been at a given time in no-wind conditions with its actual position; this was extremely difficult to do at night and in bad weather. If the first two vectors could be confirmed then the navigator had to calculate the third vector in a small ill-lit compartment, working on a vibrating desktop.

ETA

Estimated time of arrival (over a certain navigation point or at a location).

Fishpond

Radar device, operating off of the H2S system, which warned of the presence of other aircraft.

Flensburg

German device allowing a fighter to home in on transmissions made by Monica and Mandrel.

Free-lance fighter

A fighter using neither ground-based nor on-board radar to close with and identify the bomber stream or individuals bombers. These fighters would normally be directed towards the bomber stream by a running commentary from the ground controllers of the attacks progress.

Freya

German early warning radar.

242

Appendix 2: Glossary

G/C

Group Captain (equivalent to army Colonel).

Gardening

Op for laying of anti-shipping mines by Bomber Command aircraft.

GCI (German Würzburg)

Ground Controlled Interception. This refers to the use of ground based radar to direct a fighter towards a particular bomber or the bomber stream. The fighter could be either a catseye interceptor, which would use visual means (eyesight) to sight and attack the bomber, or an AI interceptor with its own integral radar for the final approach to the bomber.

Gee (TR 1335)

First used operationally on 8 March 1941, Gee consisted of a series of three ground transmitters producing signals that would be received by aircraft and portrayed on a cathode ray tube as two position lines called Gee coordinates. Using these and a special map with all Gee lines overprinted on it, experienced navigators could pin-point an aircraft’s position in less than one minute. Accuracy ranged from one-half to five miles, the less accurate fixes obtained as the aircraft approached the system’s maximum range, which varied from 300 to 400 miles. Gee required ‘line of sight’ between the aircraft and the ground stations and thus the curvature of the earth limited the range. The system was good for targets as far east as the Ruhr. Any number of aircraft could use the system simultaneously. Gee was also subject to jamming and this started five months after first operational use. Subsequently the system was not reliable past the Dutch coast or northern France but remained useful for aircraft returning to base.

G-H

First used on 3 November 1943, the device combined H, in effect an Oboe system operating in reverse, with Gee. Aircraft could navigate to the area of the target using Gee and then proceed to a precise bomb release point by means of H. While still limited by line of sight and jamming or interference, it had a number of advantages over Gee and Oboe: one hundred aircraft at a time could make use of the ground stations; range was also greater

Appendix 2: Glossary

243

and accuracy did not diminish as was the case with Gee; and H could be used in combination with H2S. G-H Formation

Daylight attack procedure where aircraft attacked in formation. Leading aircraft bombed using G-H and followers released bombs when the leaders’ bombs were seen falling. Aircraft usually flew in elements of three with each leader using G-H. There was no requirement to see ground.

Grocer/Airborne Grocer

Device for jamming German AI radar.

Groundmarking

The marking of a target using target indicators which ignite just before hitting the ground and continue to burn for some minutes after reaching the ground. Primary TIs were normally red and secondary (backers up) TIs were green.

H2S

First used on 30 January 1943, H2S was an aircraft-mounted ground-mapping radar which was useful for both navigation and blind target identification and bombing. Aircraft were equipped with rotating radar which depicted ground detail on a cathode ray tube, the Plan Position Indicater (PPI), in the navigator’s compartment. Bodies of water and coastlines were easily defined and the outlines of towns could also be seen. H2S was a self-contained system; as such there were no range limitations. However H2S did not give a precise image of built-up areas, so that it was possible to confuse similar looking towns, and over Berlin in particular the size of the city was such that the entire screen would be white. These characteristics significantly degraded the blind bombing efficacy of the system. As well, H2S signals allowed the enemy to determine the location of the bomber stream and also to home in on individual bombers using an airborne device called Naxos. This was not realized until the fall of 1944 at which time the use of H2S was severely restricted. Despite these limitations, H2S was more widely available and used than other systems and was also employed by American forces under the name H2X.

244 Appendix 2: Glossary H2S Paramatta

Target-marking technique. Similar to Musical Paramatta, but using H2S rather than Oboe. As such more primary markers could be dropped and these were used by backers up who aimed at the MPI. The Main Force then bombed the backers up marking and not the primary TIs.

H2S Skymarking

See Wanganui.

Heading

The compass direction in which the aircraft is pointing. It would not be the same as the track of the aircraft across the ground except where the wind was either directly behind or in front of the aircraft. See Dead Reckoning and Track.

IFF (Identification Friend or Foe)

Device used to identify friendly aircraft by ‘interrogating’ a transmitter on the aircraft for a specific coded reply. This system was believed by Bomber Command crews to have some ability to jam ground-based German radars.

Incendiary

Bomber Command principally used two types of incendiary or fire bombs: the 4 lb IB and the 30 lb IB. The purpose of these bombs was to raise a fire on the ground which could completely destroy a target. By comparison, High Explosive (HE) bombs were more likely to damage facilities. In terms of destructive power per ton of bombs the IBs were found to be much more effective.

Main Force

Bombers not involved in finding and marking the target. These aircraft would time their arrival such that the target was illuminated and marked before they began their attacks.

Mandrel

Electronic jamming of German early warning radars.

Master Bomber

A technique first developed in 5 Group in which a designated officer would orbit the target area and issue, via radio telephone, adjustments to the raid order or specific corrections to arriving crews so that the bombing was as accurate as possible. Deputy master bombers were also

Appendix 2: Glossary

245

assigned so that this control could be generally assured in case of the loss of the primary master bomber. Monica

On board radar device to warn of approaching enemy fighters.

MPI

Mean point of impact. The centre of a number of bomb strikes. Could be used to describe the centre of a series of target indicators (see H2S Paramatta) or the results of a raid.

Musical Newhaven

Same as Newhaven but with initial proximity marking done by Oboe.

Musical Paramatta

Target-marking technique. Method of groundmarking using coloured target indicators dropped blindly using Oboe.

Musical Wanganui

Target-marking technique. Method of skymarking using coloured markers dropped blindly using Oboe.

Naxos

German device to detect and home on H2S transmissions.

Newhaven

Target-marking technique. Method of groundmarking using flares dropped blindly using H2S followed, when possible, by visual identification using coloured target indicators.

Oboe

First used in December 1942, Oboe consisted of two ground stations which transmitted pulses which were rebroadcast back from an aircraft. These returns were measured and permitted the ground stations to determine the distance of the aircraft from the station. The ‘cat’ station would coordinate the track of the aircraft along an arc which would bring the aircraft over the target, while the ‘mouse’ would determine the point along the arc at which to drop bombs. Unlike Gee, accuracy of the system was measured in hundreds of yards. Oboe stations could only control one aircraft at a time and for the final run-in to the target this meant ‘handling’ one bomber every ten min-

246 Appendix 2: Glossary utes. In addition, Oboe was limited by ‘line of sight’ and by jamming and interference from other radio signals. To minimize the first limitation, Bomber Command set up three pairs of stations so that eighteen aircraft could be controlled each hour. To counter the range issue Mosquitoes with ceilings of over 30,000 ft were employed so that Oboe had a range of approximately 270 miles. Because of the limited number of aircraft that could be controlled, it was decided that Oboe would be used by aircraft marking the target, while the main force would bomb on these indicators. Oboe Formation

Daylight attack procedure where aircraft attacked in formation. Leading aircraft bombed using Oboe control and followers (usually not more than a dozen) released bombs when the leader’s bombs were seen falling. No requirement to see ground.

OIC

Officer in Charge or Officer in Command (normally associated with rank of Wing Commander).

op

Each night’s raid would be termed an ‘op’ or operation. There would be a separate Operation Order issued which would detail the weather, target, route, bomb load, force size, and any additional details. The term could be interchanged with ‘trip’ or ‘sortie.’

PFF/pathfinder

An aircraft and crew of No 8 Group also called the Pathfinder Force (PFF) who, at the outset of the actual bombing phase of a raid, would identify the target area and mark the aiming point using Target Indicators (described elsewhere) for the Main Force crews which followed. PFF units were initially picked for their accuracy, but later individual crews would be assigned directly to the Group and would be allowed to conduct marking only after they had undergone extensive training and gained experience. Pathfinders occasionally dropped markers along the route but this practice was abandoned when it was discovered that German fighters were using the TIs to find the bomber stream.

Random Error

The bombing error averaged across the errors of all crews. See also systematic error.

Appendix 2: Glossary

247

RDF

Radio Direction Finding. The term first used to describe what became known as radar.

Running Commentary

Technique used by the German air defence system to direct fighters towards the bomber stream. This was employed after the introduction of Window, which made ground-controlled interceptions unworkable.

Samson

Blind bombing using Gee.

Serrate

Radar device allowing RAF fighters to home on AI radar transmissions of German aircraft.

Shaker

Target identification technique using flares dropped by Gee-equipped aircraft and followed by a small force bombing by visual sighting which would drop incendiaries in order to mark the target with fires for the following main force.

Skymarking

The use of TIs dropped by pathfinders which would descend slowly by parachute.

Small Bomb Container (SBC)

SBCs were used to carry incendiary bombs (IBs) in the bomb bay. These were needed because it was not possible to store the large quantities of small IBs in any other fashion. When these bombs were released they would fall away from the bomber in a shower of small projectiles. This was relatively ineffective technique as the IBs were not particularly aerodynamic and could drift. They also filled the sky, creating the possibility of strikes on friendly aircraft. As the war progressed attempts were made to replace the SBCs with cluster bomb units (CBUs), which would fall as a normal bomb and then break apart into individual IBs close to the ground, thus mitigating both problems.

Sortie

Each aircraft taking part in a mission would be counted as one sortie. Crews would have to fly thirty sorties, generally, to complete a tour. The term could be interchanged with ‘trip’ or ‘op.’

248 Appendix 2: Glossary Systematic error

An error in bombing applied to all aircraft. Commonly systematic errors were the result of poor target marking or strong winds, factors which would cause all crews to bomb away from the intended aiming point.

Target Indicator (TI)

A variety of devices were tried in early years but only at the end of 1942 was an effective visual device found to mark the target. The TI consisted of a 250-lb bomb case filled with 60 12-inch pyrotechnic candles. These would be ejected from the case at a preset altitude of 3000 ft, at which time they would ignite and cascade to the ground where they would burn for three minutes. Other variations were subsequently developed.

Tinsel

Jamming of radio transmissions of German fighters.

Tour

A tour (not ‘tour of duty’) would consist generally of thirty completed sorties. Partial credit was given for ‘easy’ targets.

Track

The actual direction that an aircraft travels across the ground. It is the vector of the heading of the aircraft and the winds present at the altitude at which the aircraft is flying. See Dead Reckoning and Heading.

Trip

Each sortie undertaken by a crew would be termed a trip. The origin of the term is unknown, but it continues to be used in RAF and Canadian parlance. The term could be interchanged with ‘op’ or ‘sortie.’

W/C

Wing Commander (equivalent of an army Lieutenant Colonel).

Wanganui

Target-marking technique; also called H2S skymarking. Similar to Oboe Skymarking (Musical Wanganui).

Window

Tin foil strips designed to interfere with German radar frequencies, thus preventing effective direction of fighters and flak.

Source: Organization diagram from Henry Probert, Bomber Harris: His Life and Times (London: Greenhill Books, 2001),

Appendix 3: Organization Charts 249

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

APPENDIX 3: AIR MINISTRY AND BOMBER COMMAND HEADQUARTERS ORGANIZATION CHARTS AND A ‘GENERIC’ POST-WAR COMMAND HEADQUARTERS

Note: There is no reference to an ORS. Source: Organization diagram from Henry Probert, Bomber Harris: His Life and Times (London: Greenhill Books, 2001), redrawn by M. Bechthold.

250 Appendix 3: Organization Charts

BOMBER COMMAND 1942–5

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

Appendix 3: Organization Charts 251

Note: In this early post-war version of a headquarters ‘Research’ is now included as a staff within the Air branch. Source: A.H. Stradling, The Brass Hat (London: Gale and Polden, 1951), redrawn by M. Bechthold.

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

This page intentionally left blank

Appendix 4: Personnel 253

APPENDIX 4: AIR MINISTRY AND BOMBER COMMAND APPOINTMENTS Organization

Incumbent(s)

Dates

Sir Archibald Sinclair Capt Harold Balfour

1940–45 1939–45

Air Ministry Secretary of State for Air Parliamentary Under-secretary of State for Air Permanent Under-secretary Chief of the Air Staff (CAS)

Maurice Dean A/C/M Sir Cyril Newall A/C/M Sir Charles Portal Vice Chief of the Air Staff (VCAS) A/M Sir Richard Peirse A/C/M Sir Wilfrid Freeman A/V/M C.DE.H. Medhurst Deputy Chief of the Air Staff (DCAS) A/V/M R.E.C. Peirse (position redesignated A/CAS Ops A/V/M W.S. Douglas from 1941–3) A/V/M Arthur T. Harris A/V/M Norman H. Bottomley Director of Bomber Operations A/Cdre John Baker (position created 1941) A/Cdre Sidney O. Bufton Deputy Director of Bomber G/C Sidney O. Bufton Operations

1939–45 1937–40 1940–5 1940 1940–2 1942 1937–40 1940 1940–1 1941–5 1941–3 1943–5 1941–3

Bomber Command HQ Air Officer Commanding-in-Chief (AOCinC)

Deputy Commander-in-Chief (D/CinC) Senior Air Staff Officer (SASO) Note: There were two SASOs from February 1944, Walmsley for strategic bombing and Oxland for tactical operations

A/C/M Sir Edgar Ludlow-Hewitt A/M Sir Charles Portal A/M Sir Richard Peirse A/V/M J E A Baldwin A/C/M Sir Arthur Harris A/M Sir Robert Saundby

1937–40 1940 1940–2 Jan–Feb 1942 1942–5 1943–5

A/V/M Norman H. Bottomley 1938–40 A/V/M Robert S. Saundby 1940–3 A/V/M Robert D. Oxland 1943–5 A/V/M Hugh S.P. Walmsley 1944–5

254 Appendix 4: Personnel Organization

Incumbent(s)

Deputy Senior Air Staff Officer D/SASO Air Commodore Operations (A/Cdre Ops) Group/Captain Operations G/C Ops (CNavO) Command Navigation Officer Command Signals Officer (CSO) Group/Captain Radar (G/C Radar) Air Officer Training (AOT) Chief Intelligence Officer (CIO) Chief Armament Officer (CArmO) Chief Engineering Officer (CEngO) Officer-in-Charge Operational Research Section (OIC ORS) ORS 1 – Research into Success of Operations ORS 2 – Research into Bomber Losses ORS 3 – Statistical Section

A/Cdre T.M. Williams 1941–2 A/Cdre Richard Harrison 1942–3 A/Cdre Hugh A. Constantine 1943–5

ORS 4 – General Problems ORS 5 – Research into Radar Aids and Navigation and Bombing ORS 6 – Research into Manpower Economy ORS 1 Group ORS 3 Group ORS 4 Group ORS 5 Group ORS 6 (RCAF) Group ORS PFF 8 Group ORS 100 Group

Dates

G/C Samuel. C. Elworthy

1942–3

G/C H.R. Graham G/C Dalton-Morris G/C D. Saward A/Cdre A.J. Capel G/C L.F. Penred unknown A/Cdre Roach Dr Basil G. Dickins

1942–3

Mr G.A. Roberts Dr B.G. Peters Dr R.J. Smeed Miss J.M.M. Goggin Mr W.D. Richards Mr J.E. Fothergill unknown Mr G.A. Roberts

1941–4 1944–5 1941–5 1941–2 1942–4 1944–5

Mr K.A. Scott

1942–5

Mr A.W. Pratt unknown Mr F.J. Lloyd Mr J. Curry Dr J.W. Hopkins (Cdn) Mr J.A. Jukes Mr L.F. Lemmerton

1944–5 1943–5 1943–5 1944–5 1942–5 1943–5

A/V/M Arthur Wright A/V/M Robert D. Oxland

1939–40 1940–3

1941–5 1941–5 1941–2

1941–5

1944–5

Operational Groups 1 Group

Appendix 4: Personnel 255 Organization

3 Group

4 Group

5 Group

6 (RCAF) Group

8 Group (Pathfinder Force) 100 Bomber Support Group

Incumbent(s)

Dates

A/V/M Sir Edward Rice A/V/M ‘Bobby’ Blucke A/V/M J.E.A. Baldwin A/V/M The Hon R.A. Cochrane A/V/M Richard Harrison A/V/M Sir Arthur Coningham A/V/M Sir Roddy Carr A/V/M Sir John Whitley A/V/M Sir Hugh Walmsley A/V/M A.T. Harris A/V/M Sir Norman Bottomley A/V/M Sir John Slessor A/V/M Sir Alec Coryton A/V/M the Hon Sir Ralph A. Cochrane A/V/M Sir Hugh Constantine A/V/M George Brookes A/V/M Clifford (‘Black Mike’) McEwen A/V/M Donald Bennett A/V/M Edward B. Addison

1942–5 1945 1939–42 1942–3 1943–6 1939–41 1941–5 1945 1945 1939–40 1940–1 1941–2 1942–3 1943–5 1945 1942–4 1944–5 1942–5 1943–5

This page intentionally left blank

Appendix 5: Air Staff Directive 257

APPENDIX 5: AIR STAFF BOMBING DIRECTIVE TO BOMBER COMMAND, 13 JULY 1940* Air Vice-Marshal W.S. Douglas (Deputy Chief of the Air Staff) to Air Marshal Sir Charles Portal Sir, I am directed to inform you that the Air Staff has considered the bombing policy for the coming moon-phase and, in general, has decided that your main offensive should be directed towards objectives the destruction of which will reduce the scale of air attack on this country, and that your only diversion should be a comparatively small effort against communications. 2. In the course of their review, the Air Staff came to the conclusion that attacks on industrial objectives have hitherto been too dispersed and that, in consequence, few objectives have sustained a sufficient damage to put them out of action for any length of time. During this moon phase therefore, you are requested to direct a greater effort of attack on fewer targets with a view to complete destruction rather than harassing effect. 3. I am therefore to request that you will direct your operations in the immediate future in accordance with the following revised directive, which you will observe includes a total of only 15 primary targets. Aircraft Industry 4. To gain the most immediate effect on the output of aircraft, the five aircraft depots and the five airframe assembly factories in List I of Appendix A to this letter should be destroyed. In this connection, your attention is drawn to paragraphs 27–31 of plan W. A. 1, in which the conclusion is drawn that 140 five-hundred lb. bombs should be aimed at each factory in order to complete its destruction. While in no way suggesting that the scale should be rigidly adhered to, it is nevertheless an indication of the weight of attack which the Air Staff has in mind. Oil 5. The reduction of Germany’s oil resources will not have an immediate effect on the activity of the G.A.F., but the effect when felt will be permanent. Moreo-

258 Appendix 5: Air Staff Directive ver, as your operations have shown, oil targets are very vulnerable and do not call for as great an expenditure of effort as factories. While the Alumina plants which are included in the Air Staff directives of 20th June and 4th July are probably of equal significance to the aim in view, it is considered that oil plants are more profitable objectives for night attack during the present phase. 6. Your night objectives in the second priority should therefore be the five oil plants in List II of the Appendix. Communications 7. The undamaged aqueduct of the pair north of Munster is the most important target in this category and the Air Staff feels that a determined effort should be made to put it out of action and thus complete the satisfactory results which have already been achieved against this double objective. Apart from this one objective however the weight of attack on Communications should be light. They might well be included only as ‘last resort’ objectives, and it is desired to specify only the following: The Shaft-Lock at Minden The Aqueduct “

g

Mining 8. The effort of three squadrons should continue to be devoted to these operations. Employment of Medium Bombers 9. The employment of the medium bombers should be coordinated with these main operations as far as possible. Their primary role however must be to attack any large concentrations of barges or shipping which are reported to have been collected for an invasion of this country, and any advance aerodromes which are known to be carrying a sufficiently large concentration of aircraft to make attack worth while. Such effort as can be spared from these primary tasks should be directed against the three oil targets in occupied French territory, enumerated in List III of the Appendix. 10. Finally, I am to request that you will be prepared, at short notice, should the occasion arise, to divert the whole of the bomber force to the attack of an

Appendix 5: Air Staff Directive 259 invading force at the ports of departure and subsequently at sea or at the points of landing in this country. (signed) W.S. Douglas

* Reproduced from C. Webster and N. Frankland, The Strategic Air Offensive against Germany, 1939–1945, vol. 4 (London: HMSO, 1961), 119–21.

This page intentionally left blank

APPENDIX 6: PRINCIPAL BOMBER COMMAND AIRCRAFT

Appendix 6: Aircraft 261

‘Whitley in flight. Date – 25 February 1941.’ (Canadian Forces Joint Imagery Centre PL 3099)

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

262 Appendix 6: Aircraft ‘Hampden aircraft on flight line. 408 Sqn. Date – 30 September 1941.’ (Canadian Forces Joint Imagery Centre PL 4715)

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

Appendix 6: Aircraft 263

‘Wellington Mk II, W5379 in flight. Date – 13 October 1941.’ (Canadian Forces Joint Imagery Centre PL 5391)

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

264 Appendix 6: Aircraft

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

‘Carrying more than five tons of bombs, a giant Stirling goes out in the waning light, course set for Germany. Date – 1 March 1943.’ The slab-sided Stirlings were unfortunately low and slow compared even to the Halifax. (Canadian Forces Joint ImC PL 15526)

‘Handley Page Halifax Mk III of 424 Squadron at Skipton on Swale Yorkshire. “Oscar”. Date – 13 November 1944.’ Compared to the Halifax seen bombing up in the photograph section this aircraft now has the improved and rectangular vertical fins, radial engines which were less vulnerable to damage, and an H2S blister under the rear fuselage. (Canadian Forces Joint Imagery Centre PL 41055)

Appendix 6: Aircraft 265

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

266 Appendix 6: Aircraft

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

‘The first Canadian built aircraft, Lancaster KB-700. Date – 6 August 1943.’ Building British designs in North America had proven extremely challenging, but the Canadian Lancasters were just a capable as their English compatriots.(Canadian F J i I C PL 20419)

Appendix 6: Aircraft 267 ‘DeHavilland Mosquito B Mk 25. Date 25 June 1943.’ The Mosquitos escaped German defences by virtue of their exceptional speed and altitude. (Canadian Forces Joint Imagery Centre PL 14568)

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

This page intentionally left blank

Appendix 7: Bomber Command Personnel 269

APPENDIX 7: BOMBER COMMAND ORS SENIOR PERSONNEL Individual

Duties

Dates

Mr H.L. Beards

Bombing accuracy; weapons effectiveness; i/c ORS 1(d) Losses analysis; later i/c ORS 2(b) (radio counter measures) Radio aids to navigation; losses analysis; ORS 5 Group Theoretical investigations concerning losses O i/c ORS Losses analysis; later i/c ORS 3 Monthly review of bomber losses; i/c ORS 3; later i/c ORS 2(a) ORS 6 Group (Canadian attachment) Radio aids to navigation; ORS PFF; ORS BDU Loose analysis; ORS 4 Group i/c ORS 2(b); later ORS 100 Group Vulnerability research; i/c ORS 2(c) i/c ORS 1(a); later i/c ORS 1 ORS 1 Group i/c ORS 3 i/c ORS 1; later i/c ORS 5 i/c ORS 2 Visual Identification and training problems; later i/c ORS 1(c)

1941–5

Mr S.C. Britton Mr J. Curry Mr F.J. Dyson Dr B.G. Dickins Mr J.E Fothergill Miss J.M.M. Goggin Mr J.W. Hopkins Mr J.A. Jukes Mr F.J. Lloyd Mr L.F. Lammerton Mr E.A. Lovell Dr B.G. Peters Mr A.W. Pratt Mr W.D. Richards Mr G.A. Roberts Dr R.J. Smeed Mr G.W.H. Stevens

1942–5 1942–5 1943–5 1941–5 1946–6

1944–5 1941–5 1942–5 1942–5

1943–6 1942–4 1941–5 1941–7

This page intentionally left blank

Appendix 8: Bombing Techniques 271

APPENDIX 8: BOMBING TECHNIQUES

272 Appendix 8: Bombing Techniques

Appendix 8: Bombing Techniques 273

274 Appendix 8: Bombing Techniques

Appendix 8: Bombing Techniques 275

276 Appendix 8: Bombing Techniques

Diagram Illustrating the Technique of Night Photography Source: Detail from ORS Memorandum No. 71, Relation between the Centre of a Night Photograph and the Fall of Bombs, 23 May 1943 (PRO AIR 14/1756), drawn by M. Bechthold.

Appendix 8: Bombing Techniques 277 Source: Detail from ORS Report 89, Analysis of Navigation – Raid against Kassel 22/23 October 1943, 30 December 1943

Wakelam, Randall Thomas. The Science of Bombing : Operational Research in RAF Bomber Command, University of Toronto Press, 2009. ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/nyulibrary-ebooks/detail.a Created from nyulibrary-ebooks on 2020-11-04 21 38 24

Copyright © 2009. University of Toronto Press. All rights reserved.

278 Appendix 8: Bombing Techniques

Source: Detail from ORS Report 89, Analysis of Navigation – Raid against Kassel 22/23 October 1943, 30 December 1943 (PRO AIR 14/1753).

Appendix 8: Bombing Techniques 279

Note: While relatively well grouped, the raid missed its target completely, despite the use of pathfinders. Source: Detail from Night Raid Report No 176, Bomber Command Report on Night Operations, 15/16 October 1942, Cologne, 31 December 1942 (PRO AIR 14/3409).

280 Appendix 8: Bombing Techniques

Percentage within 3 miles of Aiming Point

Source: The Crucible of War, 1939–1945, vol. 3 of The Official History of the Royal Canadian Air Force. Reproduced with the permission of the Minister of Public Works and Government Services and Courtesy of National Defence, 2008.

Appendix 8: Bombing Techniques 281

Source: Detail from ORS S 91, Night Bomber Losses on German Targets, 1942, 12 April 1943 (PRO AIR 14/1800).

282 Appendix 8: Bombing Techniques

Source: Detail from ORS S 187, The Effect of Experience on Efficiency, 2 May 1944 (PRO AIR 14/1847).

Appendix 8: Bombing Techniques 283

Source: The Crucible of War, 1939–1945, vol 3 of The Official History of the Royal Canadian Air Force. Reproduced with the permission of the Minister of Public Works and Government Services and Courtesy of National Defence, 2008.

This page intentionally left blank

Notes

Introduction: Commanders, Headquarters, and Decision Making 1 Col. Phillip S. Meilinger, ‘Trenchard and “Morale Bombing”: The Evolution of Royal Air Force Doctrine before World War II,’ Journal of Military History 60, no. 2 (April, 1996), 258, 266–7. 2 Scot Robertson, ‘In the Shadow of Death by Moonlight,’ in David Bercuson and S.F. Wise, eds, The Valour and the Horror Revisited (Montreal: McGillQueen’s University Press, 1994), 169–71. 3 Robertson, ‘In the Shadow of Death by Moonlight,’ 165–6. 4 Group Captain Peter W. Gray, RAF, ‘Review Essay: Bomber Harris His Life and Times,’ The Royal Air Force Air Power Review, UK Ministry of Defence, vol. 4 (2001), 1–7. 5 Robin Neillands, The Bomber War: The Allied Air Offensive against Nazi Germany (New York: Overlook Press, 2001), 1. Readers will want to consult all of the works noted to this point as they give what is arguably a fair airing of the concerns about area bombing which many people have had both during and since the war. 6 Sir Charles Webster and Noble Frankland, The Strategic Air Offensive against Germany 1939–1945, vol. I, Preparation; vol. II, Endeavour; vol. III, Victory; vol. IV, Annexes and Appendices (London: Her Majesty’s Stationery Office, 1961). Hereafter SAO. 7 Brereton Greenhous et al., The Crucible of War, 1939–1945: The Official History of the Royal Canadian Air Force Volume III (Toronto: University of Toronto Press, 1994). 8 Air Ministry, The Origins and Development of Operational Research in the Royal Air Force (London: HMSO, 1963). Hereafter Operational Research in the RAF. 9 Air Chief Marshal Sir Arthur T. Harris, Despatch on War Operations 23rd

286

Notes to pages 9–17

February, 1942, to 8th May, 1945, ed. Sebastian Cox (London: Frank Cass, 1995). Hereafter Despatch. 10 Marshal of the R.A.F. Sir Arthur Harris, Bomber Offensive (Toronto: Stoddart, 1990). 11 Basil G. Dickins, Operational Research in Bomber Command (n.d.) Hereafter OR in Bomber Command. 1. Strategic Bombing: When Theory and Practice Do Not Match 1 Webster and Frankland, SAO, I, 129. 2 Webster and Frankland, SAO, III, 124. 3 Wing Commander H.R. Allen, The Legacy of Lord Trenchard (London: Cassell, 1972), 32. See also Col. Phillip S. Meilinger, ‘Trenchard, Slessor, and Royal Air Force Doctrine before World War II,’ in The Paths of Heaven, ed. Col. Phillip S. Meilinger (Maxwell Air Force Base, Alabama: Air University Press, 1997), 45–7. 4 Allen, Legacy, 31–6. 5 Allen, Legacy, 49–58. 6 Webster and Frankland, SAO, I, 89–106. See also Charles Carrington Soldier at Bomber Command (London: Leo Cooper), 1987, 61. 7 Carrington, Soldier, 61. 8 Allen, Legacy, 78. 9 Allen, Legacy, 79. 10 Allen, Legacy, 78–80 (Allan cites Harris, Bomber Offensive). See also Webster and Frankland, SAO, I, 116. 11 Webster and Frankland, SAO, I, 98. 12 Webster and Frankland, SAO, I, 89–106. See also Meilinger, ‘Trenchard, Slessor and the RAF,’ 69–71. 13 See Ronald W. Clark, Tizard (London: Methuen and Co, 1965). 14 Webster and Frankland, SAO, I, 114. 15 Webster and Frankland, SAO, I, 107–25. See also Allen, Legacy, 31–6. 16 Allen, Legacy, 82. 17 Allen, Legacy, 82. See alternatively John Terraine, The Right of the Line: The Royal Air Force in the European War 1939–1945 (London: Hodder and Stoughton, 1985), 81–3. 18 Sir John Slessor, The Central Blue (London: Cassell and Co., 1956), 203–6. See also Greenhous et al., The Crucible of War, 1939–1945, 533. 19 Slessor, The Central Blue, 150–1. 20 Denis Richards, Portal of Hungerford (London: Heinemann, 1977), 139–40. 21 Richards, Portal, 26–70.

Notes to pages 17–22 287 22 Denis Richards, The Hardest Victory: RAF Bomber Command in the Second World War (London: W.W. Norton, 1995), 7. 23 PRO AIR 14/1982 ‘Organization of Air Staff at 5 Group Headquarters,’ n.d. 24 PRO AIR 14/3922 Note to file, covering Loose Memorandum ‘Air Staff Memorandum No. 32 “Operational Research Section,”’ 18 September 1941. Also Loose Document ‘Air Staff Memorandum No. 38 Organization of the Operations Branch,’ 17 January 1944. In fact many of the staff positions were ‘up-ranked’ in this way; for example G/C Trg became Air Officer Training (AOT), with the rank of Air Commodore in 1941. These position titles were a mix between ranks and functions, a relatively confusing system; however, the constant through the war was the functions. 25 Webster and Frankland, SAO, IV, 118–23. 26 Richards, Portal, 160. 27 Webster and Frankland, SAO, I , 190–7. See also Max Hastings, Bomber Command (New York: Dial Press, 1979), 37–58. 28 Webster and Frankland, SAO, I, 202. 29 Webster and Frankland, SAO, I , 190–7. See also Martin Middlebrook and Chris Everitt, The Bomber Command War Diaries: An Operational Reference Book, 1939–1945 (New York: Viking, 1985), 19–21, 40 (hereafter BCWD); Terraine, Right of the Line, 99–105 and 107–13. 30 Webster and Frankland, SAO, I, 204. 31 Webster and Frankland, SAO, I, 204–5. 32 Webster and Frankland, SAO, I, 204–5. See also Malcolm Smith, British Air Strategy between the Wars (Oxford: Clarendon, 1984), 269–81. 33 Webster and Frankland, SAO, I, 211. 34 Webster and Frankland, SAO, I, 205–12. See also Terraine, Right of the Line, 267–9, 274–5. 35 Richards, Portal, 299–300. The same point is made in Richards, The Hardest Victory, 289–90 and Greenhous et al., The Crucible of War, 535–6, 539–41. 36 Webster and Frankland, SAO, I, 216. 37 Webster and Frankland, SAO, I, 216. 38 Webster and Frankland, SAO, I, 215–20. 39 Greenhous et al., The Crucible of War, 561. 40 Webster and Frankland, SAO, I, 226; Greenhous et al., The Crucible of War, 541. 41 Webster and Frankland, SAO, I, 222–9. See also Middlebrook and Everitt, BCWD, 56–8, 92–4, 131–3, 174–5; Air Commodore Henry Probert, Bomber Harris His Life and Times: The Biography of Marshal of the Royal Air Force Sir Arthur Harris, the Wartime Chief of Bomber Command (Toronto: Stoddart, 2001), 99; Slessor, The Central Blue, 366–7; Marshal of the Royal Air Force

288

42 43 44 45 46 47 48

49 50 51

Notes to pages 22–7

The Lord Tedder, Air Power in War: The Lees Knowles Lectures (London: Hodder and Stoughton, 1948), 100. Webster and Frankland, SAO, I, 245–7. Richards, Portal, 164–7. Terraine, Right of the Line, 290–1. Guy Hartcup, The Effect of Science on the Second World War (London: Macmillan Press, 2000), 10–11. Richards, Portal, 303–4. See also Webster and Frankland, SAO, I, 178; Allen, Legacy, 90–3. Terraine, Right of the Line, 292–4. For other descriptions of Butt’s findings see also Webster and Frankland, SAO, I, 178; Greenhous et al., The Crucible of War, 550–2; Francis H. Hinsley, British Intelligence in the Second World War: Its Influence upon Strategy and Operations Volume II (London: HMSO, 1981), 258–62; Charles Messenger, Bomber Harris and the Strategic Bombing Offensive, 1939–1945 (London: Arms and Armour Press, 1984), 48–9; Richards, Portal, 303–4. Webster and Frankland, SAO, I, 183. Richards, Portal, 303. Richards, Portal, 303–4.

2. Operational Research: Finding a Solution through Science 1 Patrick Rivett, Concepts of Operational Research (London: C.A. Watts and Co., 1968), 1. 2 Rivett, Concepts of Operational Research, 2. 3 P.G. Moore, Basic Operational Research (London: Sir Isaac Pittman and Sons, 1968), vi. 4 Moore, Basic Operational Research, 2. 5 Moore, Basic Operational Research, 2. 6 Moore, Basic Operational Research, 6–8. 7 Moore, Basic Operational Research, 2. 8 Max Davies and Michel Verhulst, eds., Operational Research in Practice: Report of a NATO Conference (London: Pergamon, 1958), 2. 9 Davies and Verhulst, eds., Operational Research in Practice, 2. See also 114– 17. 10 J.G. Crowther and R. Whiddington, Science at War (London, HMSO, 1947), 117. 11 Rivett, Concepts of Operational Research, 5–6. 12 Air Ministry, Operational Research in the RAF, xvii. 13 Air Ministry, Operational Research in the RAF, xviii. See also Solly Zuckerman,

Notes to pages 27–34

14 15 16 17 18 19 20

21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38

289

‘The Need for Operational Research,’ in Davies and Verhulst, eds., Operational Research in Practice, 6–8; Hartcup, The Effect of Science, 100–1. Air Ministry, Operational Research in the RAF, 6. Zuckerman, ‘The Need for Operational Research,’ 8. Air Ministry, Operational Research in the RAF, 9. Davies and Verhulst, eds., Operational Research in Practice, 2. P.M.S. Blackett, Studies of War Nuclear and Conventional (New York: Hill and Wang, 1962), 171. Air Ministry, Operational Research in the RAF, xix. Ronald W. Clark, The Rise of the Boffins (London: Phoenix House, 1962), 11–13. ‘Boffins,’ a somewhat whimsical term, was used largely by the RAF to describe scientists working directly within military organizations. Clark, xvii Air Ministry, Operational Research in the RAF, 2–4. See also Hartcup, The Effect of Science, 6–7; Terraine, Right of the Line, 21–2. Air Ministry, Operational Research in the RAF, 5. Air Ministry, Operational Research in the RAF, 6–7. Hartcup, The Effect of Science, 103. Air Ministry, Operational Research in the RAF, 7–8. See also Crowther and Whiddington, Science at War, 91–3. Air Ministry, Operational Research in the RAF, 12–16. See also Rivett, Concepts of Operational Research, 5–6. Air Ministry, Operational Research in the RAF, 19. Air Ministry, Operational Research in the RAF, 17–21. Air Ministry, Operational Research in the RAF, 74–7. See also Rivett, Concepts of Operational Research, 6–11. Rivett, Concepts of Operational Research, 11. Slessor, The Central Blue, 486–7. Slessor, The Central Blue, 524–6. Slessor, The Central Blue, 525. Slessor, The Central Blue, 526. The Nobel Prize in Physics 1948 http://nobelprize.org/physics/laureates/1948/index.html, accessed 15 May 2006. Joseph F. McCloskey, ‘British Operational Research in World War II,’ Operations Research 35, no. 3 (May–June 1987), 462–3. Blackett, Studies of War, 171. Blackett, Studies of War, 176.

3. Boffins at Bomber Command: September 1941 1 Air Ministry, Operational Research in the RAF, 8.

290 Notes to pages 34–8 2 Air Ministry, Operational Research in the RAF, 8. 3 Air Ministry, Operational Research in the RAF, 8. See also Clark, Tizard, 161–2 and 196. A comprehensive account of these issues is contained in chapter 1 of the Bomber Command Operational Research Section manuscript which details the earliest steps in conducting OR on bomber operations. 4 Dickins, OR in Bomber Command, 1. This work is also described in Air Ministry, Operational Research in the RAF, 43. 5 Dickins, OR in Bomber Command, 1. 6 Dickins, OR in Bomber Command, 1. Other scientists had already been working in the headquarters. In April Professor A.O. Rankine had been attached as a liaison officer from Sir Robert Watson-Watt, who was the Scientific Adviser on Telecommunications at the Air Ministry. Rankine was replaced by Mr C.A. Roberts of the Telecommunications Research Establishment in July. 7 Dickins, OR in Bomber Command, 2. 8 Dickins, OR in Bomber Command, 3. 9 Dickins, OR in Bomber Command, 2–6. 10 E-mail Ms Catherine Harpham, Archivist, Imperial College London, to the author, 10 February 2006. 11 ‘Atomic Weapons Chief: Appointment of Dr B.G. Dickins,’ The Times, London, 4 June 1959, p. 16. See also Directory of British Scientists 1964–65 (London: Ernest Benn, 1964) and ‘Aeroplane’ Directory of British Aviation 1967 Edition (London: Temple Press, 1967). 12 David Edgerton, Warfare State: Britain 1920–1970 (Cambridge: Cambridge University Press, 2006), 340. 13 Clark, The Rise of the Boffins, 50–4. 14 Clark, The Rise of the Boffins, 87. 15 Ralph Barker, The Thousand Plan: The Story of the First Thousand Bomber Raid on Cologne (London: Chatto and Windus, 1965), 42. 16 Webster and Frankland, SAO, I, 251. 17 Sir Solly Zuckerman, From Apes to Warlords: The Autobiography (1904–1946) of Solly Zuckerman (London: Hamish Hamilton, 1978), 234. It is also worth noting that either Zuckerman or his editor seemed ill disposed to spell Dickins’s name correctly, referring to him as Dickens. 18 Hartcup, The Effect of Science, 110. 19 Hartcup, The Effect of Science, 110. Quoted from Freeman Dyson, ‘The Flying Coffins of Bomber Command,’ Observer Magazine, 28 October 1979, p. 69 20 Freeman Dyson, Disturbing the Universe (New York: Harper and Row, 1979), 20–31. 21 Dyson, Disturbing the Universe, 21–5.

Notes to pages 38–42 291 22 23 24 25

26 27 28 29 30 31 32 33 34

35 36 37 38 39 40 41

42 43 44 45

Dyson, Disturbing the Universe, 26–8. Dyson, Disturbing the Universe, 26. Directory of British Scientists 1964–65, 463. Carrington, Soldier, 12–16. One can achieve a good sense of the physical layout, working relationships, and even, to some degree, the social life within Bomber Command Headquarters from the observations of Charles Carrington, who joined the Command in 1940. What makes Carrington’s observations particularly useful is the fact that as a soldier and veteran of the First World War, he was not particularly concerned with the possible sensitivities of describing this air force headquarters in any particular fashion, positive or negative. A contemporary of Saundby, with whom he had served in the trenches, and Harris, he was well connected in academic and literary circles. He felt no need to grovel before his seniors. Carrington, Soldier, 13. Carrington, Soldier, 13. Carrington, Soldier, 14. Carrington, Soldier, 14–15. Carrington, Soldier,, 14. Gp Capt A.H. Stradling, The Brass Hat: Being Hints on How to Make the Job Easier (Aldershot, UK: Gale and Polden, 1951), 81, 84–5. Carrington, Soldier, 16. Carrington, Soldier, 51. Messenger, Bomber Harris, 16, 55, 103, 201. For other descriptions of this occurrence see also Hastings, Bomber Command, 246; Norman Longmate, The Bombers: The RAF Offensive against Germany, 1939–1945 (London: Hutchinson, 1983), 148. Harris, Bomber Offensive, 71–2. RAF Museum, Harris Papers, H81, DO Letter, Harris to Portal, 5 July 1942; and H82, DO Letter, Harris to Portal, 14 June 1943. Carrington, Soldier, 19–20. Hastings, Bomber Command, 246–7. Carrington, Soldier, 18. Carrington, Soldier, 18. Carrington, Soldier, 138–9. Carrington recounts that his credibility immediately grew in the eyes of the brigadiers in the crowd after he was publicly greeted by Blackett at a conference that the scientist was chairing. Dickins, OR in Bomber Command, preface. Dickins, OR in Bomber Command, 13. Dickins, OR in Bomber Command, 13. Dickins, OR in Bomber Command, 13.

292

Notes to pages 43–6

46 Dickins, OR in Bomber Command, 13. See also Stradling, The Brass Hat, 8–9, for a description of the intellectual flexibility and openness of mind necessary in a good staff officer 47 Carrington, Soldier, 20. 48 Dickins, OR in Bomber Command, 14. 49 Dickins, OR in Bomber Command,, 14. 50 Criticisms of this organizational model, such as Hartcup’s described earlier, fail to recognize that commanders are also managers and set up their organizations so that they can control them in a style which they find effective. Military organizations are continually being tinkered with and sometimes wholly restructured for this reason. As long as the end structure is effective and people know to whom to speak to get the job done, such criticism is largely irrelevant. An effective organization seems to have been in place at High Wycombe. 51 Carrington, Soldier, 64. 52 Air Ministry, Operational Research in the RAF,, 60–1. 53 Dickins, OR in Bomber Command, 15. 54 Air Ministry, Operational Research in the RAF, 44. 55 PRO AIR 14/3922 Minute 5, OIC ORS to SASO, 17 September 1941; Minute 6, Note to file, covering Letter, HQ BC to Groups, ‘Operational Research Section,’ 18 September 1941. See also Ken Delve and Peter Jacobs, The Six-Year Offensive: Bomber Command in World War II (London: Arms and Armour Press, 1992), 80, 129. 56 PRO AIR 14/3922 Note to file, covering Loose Memorandum ‘Air Staff Memorandum No. 32 “Operational Research Section,”’ 18 September 1941. See also Clark, Rise of the Boffins, 220–1, who describes the integral nature of the section within the HQ: ‘[Dickins’s] brief ... had a deceptively simple air, but it did in fact cover virtually every aspect of Bomber Command operations, from the tactical methods of marking and attack of targets, through the methods of reaching them with the minimum losses, to the use of new navigational and other aids and new weapons that were being constantly introduced. In this ubiquitous role of the Operational Research Section there probably lies one reason for the fact that its work has not been more frequently mentioned. It was in fact regarded as an integral part of the Command and the results of its analysis were available to the Air Staff and were frequently of great assistance in helping them make operational decisions.’ 57 Dickins, OR in Bomber Command, 6. 58 Dickins, OR in Bomber Command, 2–6. 59 Dickins, OR in Bomber Command, 6–12.

Notes to pages 46–50 293 60 PRO AIR 14/3274 Loose Minutes ‘Notes of an Informal Discussion of the Work of the Operational Research Sections at Fighter, Bomber and Coastal Commands and of the A.W.A.S. Held at M.A.P. on Friday 10th October 1941.’ See also Crowther and Whiddington, Science at War, 94–9. 61 PRO AIR 14/3274 Loose Minutes ‘Notes of an Informal Discussion of the Work of the Operational Research Sections at Fighter, Bomber and Coastal Commands and of the A.W.A.S. Held at M.A.P. on Friday 10th October 1941.’ 62 PRO AIR 14/1763 Minute 2, SASO to CinC, 11 October 1941, covering letter Air Ministry to Commands, 2 October 1941. 63 PRO AIR 14/1763 ‘Minutes of the 1st Meeting of the Operational Research Committee, Held in the Air Ministry at 14.45 hrs. on 31st October, 1941’; ‘Operational Research Committee Minutes of the 2nd Meeting Held at the Air Ministry at 11 a.m. on Thursday, 27th November, 1941.’ 64 PRO AIR 14/1763 Minute 3, CinC to SASO, 14 October 1941. 65 PRO AIR 14/3274 Loose Minutes, ‘Minutes of the 1st Meeting of the Operational Research Committee, Held in the Air Ministry at 14.45 hrs. on 31st October 1941.’ 4. Sorting Out Process and Producing Results: September 1941–February 1942 1 Dickins, OR in Bomber Command, 18. 2 Dickins, OR in Bomber Command, 18–19. 3 PRO AIR 14/3922 Loose Memorandum, ‘Research in Relation to Bombing Operations,’ 23 July 1941. As discussed later, the scientists’ desire to take over the collection and analysis of data was accepted by the intelligence staff. Equally interesting and important are Rankine’s comments on the ability of flyers to effectively use complex equipment. In the years to follow there would be episodes of considerable frustration that various navigation and bombing devices were declared unserviceable when no fault could later be found. Equally one wonders how many bombers lost their way or misidentified the target because some piece of equipment had been miscalibrated or mistuned, or because the operator simply was incapable of keeping up with the information coming from the perfectly serviceable black box. Problems like these exist today, even though aircrews are trained over months and years and though more often than not there is no one shooting at them. Why would one expect that the hastily trained personnel of 1941 would have been any more capable? 4 Dickins, OR in Bomber Command, 22. 5 Dickins, OR in Bomber Command, 21–2.

294 Notes to pages 50–3 6 7 8 9 10 11

12 13

14 15 16 17 18

Dickins, OR in Bomber Command, 26. Webster and Frankland, SAO, I, 251. Webster and Frankland, SAO, I, 252–3. Dickins, OR in Bomber Command, 26–7. The sort of difficulty that both bomb aimers and photo interpreters faced can be comprehended from several of the night photos. Dickins, OR in Bomber Command, 28–9. Only this source refers to the CIU as the ACIU. See also Air Ministry, Operational Research in the RAF, 47–8 for a short description of the assistance provided by N Section and R.E. 8. For background on the CIU see Hinsley, British Intelligence in the Second World War, I, 279–81. Dickins, OR in Bomber Command, 30. RE 8 was a section in the Research and Experiments Department of the Ministry of Home Warfare responsible for looking at the effect of bombing. See Zuckerman, From Apes to Warlords, 242; and Air Ministry, Operational Research in the RAF, 47–8 for a short description of the assistance provided by N Section and R.E. 8. Dickins, OR in Bomber Command, 33. Dickins, OR in Bomber Command, 33. Dickins, OR in Bomber Command, 33. Dickins, OR in Bomber Command, 33. The manuscript goes on to list a variety of documents from which relevant data could be gleaned and mentions several specifically, including weather forecasts; the form ‘A’ which was the Bomber Command operation order; ‘B’ forms, which were operation orders issued by groups and later by the Pathfinder Force from which the ORS produced a document called the ‘Plan of Attack.’ In addition there was a Bomber Command Daily Operational Sorties report, which discribed those aircraft despatched and also provided the first indication of missing aircraft; there was a Pathfinder force operations board; Bomber Command Operational Weather Summary; Pathfinder Force Provisional Analysis of Operations; and finally, the Bomber Command Intelligence Summary and Narrative of Operations. These documents provided a wealth of qualitative data, but in true scientific form, the ORS considered that ‘So far as raid analysis is concerned, they are of subsidiary importance in comparison with the following three main sources of information.’ These resources were the Sortie Raid Reports, operational photographs, and the daylight photographic reconnaissance. Dickins warned that ‘limitations inherent in each type of analysis should be fully appreciated.’ However, he continued, it was essential to understand the collection and interpretation of data ‘so that the work of the ORS can

Notes to pages 53–8 295

19 20 21 22 23 24

25 26 27 28

29 30 31 32

33 34 35 36 37

be seen in the right perspective.’ Signals Branch Headquarters Bomber Command, ‘An Historical Account of the Radio Countermeasures Offensive Conducted by the Royal Air Force Bomber Command During the War in Europe 1939–1945,’ Bomber Command Headquarters, High Wycombe, 1945, 19–20. Dickins, OR in Bomber Command, 36. Dickins, OR in Bomber Command, 48–9. Dickins, OR in Bomber Command, 44–5. Dickins, OR in Bomber Command, 46. Dickins, OR in Bomber Command, 15–16. See also McCloskey, ‘British Operational Research in World War II,’ 453. RAF Air Historical Branch, Bomber Command Operational Research Section Reports ‘S’ Series S 3 ‘Investigation of raid on Huls – night of 6/7th September 1941,’ 10 September 1941. Dickins, OR in Bomber Command, 50. Dickins, OR in Bomber Command, 51. Dickins, OR in Bomber Command, 51. See also Air Ministry, Operational Research in the RAF, 48–9. PRO AIR 14/1758 Minute 2, OIC ORS to CinC (through SASO), 20 November 1941, covering ORS Memorandum 19 ‘Preliminary Note on Questionaire [sic] relating to Target Identification,’ 19 November 1941 (this report was subsequently re-designated as ORS S 19; a copy is in the RAF AHB ‘S’ Series papers); Minute 3, CinC to OIC ORS, 24 November 1941. See also PRO AIR 14/1755 Minute 2, OIC ORS to CinC (through SASO), 20 November 1941. PRO AIR 14/1758 ‘Minutes of an Informal Meeting Held at B.C.H.Q. at 10.30 a.m. on December 6th, 1941.’ Dickins, OR in Bomber Command, 52; RAF AHB BC ORS Memo M88 ‘The Success of Night Bombing Attacks, An Appreciation,’ 22 December 1941. PRO AIR 14/1758 Loose Letter 5 Group to HQBC, 9 December 1941. PRO AIR 14/1758 Minute 27, OIC ORS to W/C Nav, 19 February 1942 covering ORS Report 31 ‘The Visual Recognition of Ground Features as an Aid to Target Identification at Night,’ 14 February 1942. PRO AIR 14/1758 Minute 27, OIC ORS to W/C Nav, 19 February 1942 covering Letter BC HQ to Groups, 9 March 1942 (signed by Saundby). PRO AIR 14/1758 Minute 28, W/C Nav to OIC ORS, 20 February 1942. PRO AIR 14/1758 Minute 29, OIC ORS to SASO, 26 February 1942. Dickins, OR in Bomber Command,, 52–3. PRO AIR 14/1755 Minute 10, OIC ORS to CinC (through SASO), 6 December 1941.

296

Notes to pages 58–63

38 PRO AIR 14/1755 Minute 10, OIC ORS to CinC (through SASO), 6 December 1941. The report referred to is unidentified but is probably ORS S22 ‘Analysis of Photographs Taken during Night Bombing Operations, Period 26/7/41–31/10/41,’ nd. 39 PRO AIR 14/1755 Minute 11, SASO to CinC, 7 December 1941. There is no record of Peirse’s response. 40 AHB ORS ‘S’ Series ‘S 23 Operational Use of Gee,’ 1 December 1941, and ‘S 26 Operational Use of Gee – 2. Proposed Experiments for Gee Development Flight,’ 7 January 1942. 41 PRO AIR 14/695 Loose Minute 1, OIC ORS to SASO, 2 December 1941, covering ORS Report No 23 ‘Operational Use of GEE,’ dated December 1941. 42 PRO AIR 14/695 Loose Minute 2, D/SASO to OIC ORS, CSO, 2 December 1941. 43 PRO AIR 14/695 ORS Memorandum No 23 ‘Operational Use of GEE,’ 1 December 1941. [Subsequently re-designated S 23.] 44 PRO AIR 14/695 Minute 2, Dickins to Saundby, 2 December 1941, covering ORS Report No 23 ‘Operational Use of GEE,’ December 1941. 45 PRO AIR 14/1313 ‘Some Rough Notes on Results of 2 Flights to Investigate the Possible Use of T.R. 1335 without Lattice Charts,’ 23 December 1941. 46 PRO AIR 14/1313 ‘T.R. 1335 Development Unit – Interim Report for Period 16.12.’41 to 20.12.’41.’ 47 PRO AIR 14/3922 Minute 39, OIC ORS to SASO, 6 January 1942. 48 PRO AIR 14/3922 Minute 40, SASO to OIC ORS, 30 January 1942. 49 Dickins, OR in Bomber Command, 54. 50 PRO AIR 14/695 Loose Document ORS S 30, ‘The Operational Use of Gee. III The Use of Flares in Conjunction with GEE,’ 24 January 1942. 51 PRO AIR 14/695 Letter HQ BC to AOC 3 Gp, ‘Experimental Use of Flares by Tr. 1335 Aircraft in Order to Assist Aircraft without Tr. 1335 to Locate the Target.’ dated 29 January 1942. 52 PRO AIR 14/1755 Enclosure 26A Letter AOC 5 Group to HQ BC, 16 January 1942. 53 Dickins, OR in Bomber Command, 53–4. See also Terraine, Right of the Line, 473–5. 54 PRO AIR 14/695 AOC 3 Group to HQ BC ‘Exercise Crackers,’ 15 February 1942. 55 Webster and Frankland, SAO, I, 385–8. 56 PRO AIR 14/695 AOC 3 Group to HQ BC ‘Exercise Crackers,’ 15 February 1942.

Notes to pages 63–8 297 57 PRO AIR 14/695 Minute 51, SASO to A/CinC, 16 February 1942. Peirse had been replaced for his mishandling of a raid in early November. The catastrophic attacks of the night of 7/8 November 1941 resulted in an overall loss rate of 9.25 per cent. But when those aircraft not claiming attack were removed from the numbers the losses mounted to 13.85 per cent; for those attacking Berlin the losses were a staggering 28.76 per cent. While the cause was at first attributed to the weather, Peirse was eventually deemed the culprit for not having exercised due diligence as the commander. In these circumstances he was quietly reassigned to the Far East. This led to the interim appointment of A/V/M Baldwin, then AOC of 3 Group, until Harris could be brought back from a temporary mission to the US. See Terraine, The Right of the Line, 459–60. 58 Dickins, OR in Bomber Command, 58. See also Maurice W. Kirby, Operational Research in War and Peace: The British Experience from the 1930s to 1970 (London: Imperial College Press, 2003), 148. 5. Arthur Harris and a New Beginning: February–December 1942 1 Dudley Saward, ‘Bomber’ Harris: The Story of Marshal of the Royal Air Force Sir Arthur Harris (London: Cassell, 1984), 13–19. 2 Saward, ‘Bomber’ Harris, 27–31. See also Barker, The Thousand Plan, 1–3, and Allen, Legacy, 101–4. Allen contends that although Harris was a Trenchard disciple he was, at the same time, a much more intelligent man. Allen notes that the area bombing campaigns that Harris waged were in contradiction to the long-standing British philosophy of minimum force. 3 Messenger, Bomber Harris, 17–24. 4 Messenger, Bomber Harris, 26–40. 5 Probert, Bomber Harris His Life and Times, 95 (from the Papers of Air Chief Marshal Sir Edgar Ludlow-Hewitt, Folder 11). 6 Hastings, Bomber Command, 245. 7 Hastings, Bomber Command, 247. See also 147. 8 Dyson, Disturbing the Universe, 29. 9 Carrington, Soldier, 85–6. 10 Carrington, Soldier, 131. See also 75–6 for a general description of Harris and his relations with subordinates. 11 Probert, Bomber Harris His Life and Times, 95. 12 Probert, Bomber Harris His Life and Times, 104. 13 Harris, Bomber Offensive, 72. See also Norman Longmate, The Bombers: The RAF Offensive against Germany, 1939–1945 (London: Hutchinson, 1983), 148.

298 Notes to pages 68–73 14 Harris, Despatch, 8. See also Allen, Legacy, 107–11. 15 Webster and Frankland, SAO, IV, 143–8. See also Harris Despatch,, 7–8, 192; Webster and Frankland, SAO, I, 322–4; Greenhous et al., Crucible of War, 576. 16 Webster and Frankland, SAO, IV, 148–51. See also Middlebrook and Everitt, BCWD, 238–42, 269–71. 17 Webster and Frankland, SAO, I, 386–8. What the official history does not mention is the extensive role played by the ORS in developing the Shaker concept and in identifying concentration as a valuable means of reducing casualties. 18 Webster and Frankland, SAO, I, 388–95. 19 PRO AIR 14/2017 Letter HQ BC to Groups, 7 March 1942 covering ORS S 31 ‘Statistical Evidence of the Visual Location of Targets by Night,’ 6 February 1942. 20 PRO AIR 14/2017 Minute 2, Air 1 to AOC, SASO, 9 March 1942, covering Letter HQ BC to Groups, 7 March 1942 covering ORS S 31 ‘Statistical Evidence of the Visual Location of Targets by Night,’ 6 February 1942. 21 PRO AIR 14/2017 Minute 3, AOC to SASO, ‘Tactics,’ n.d. 22 Harris, Bomber Offensive, 94. 23 Harris, Bomber Offensive, 94–5. 24 Middlebrook and Everitt, BCWD, 248. 25 Webster and Frankland, SAO, I, 388–95. See also Middlebrook and Everitt, BCWD, 251; 259–61. 26 Webster and Frankland, SAO, I, 395–7. 27 PRO AIR 14/2017 ORS S45 ‘Success of Bombing Operations as Shown by Night Photographs Dec. 1941–Feb. 1942,’ 22 April 1942. 28 PRO AIR 14/2017 Minute 4, Air 1 to AOC, SASO, 13 May 1942, covering Letter HQ BC to Groups, 3 May 1942, covering ORS S45 ‘Success of Bombing Operations as Shown by Night Photographs Dec. 1941–Feb. 1942,’ 22 April 1942; Minute 5, SASO to AOC, 13 May 1942; Minute 6, AOC to SASO, 14 May 1942. 29 Greenhous et al., The Crucible of War, 584–5. 30 PRO AIR 14/1769 ORS B 111 ‘Attack on Essen I – Blind Bombing by T.R. 1335,’ 4 June 1942. 31 PRO AIR 14/1769 Minute 27, OIC ORS to CinC, 5 June 1942, covering ORS S 47 ‘Operational Accuracy of Blind Bombing with T.R. 1335, March and April 1942,’ 18 May 1942. 32 PRO AIR 14/1769 Minute 28, CinC to SASO, 8 June 1942. 33 PRO AIR 14/1769 Minute 37, OIC ORS to SASO (through W/C R.D.F.) 11 July 1942, covering ORS S 55 ‘Accuracy of T.R. 1335 – Observational Errors,’ 3 July 1942.

Notes to pages 73–6 299 34 PRO AIR 14/1769 Minute 39, OIC ORS to CNavO, W/C RDF, G/C Ops 16 July 1942 covering ‘Note on Attack on Essen March 8/9th–June 8/9th,’ 26 July 1942. 35 See Saward ‘Bomber’ Harris. The W/C RDF was in fact Dudley Saward, who in his account had been appointed to the RDF job in at the end of 1941. He says that “at about the same time” the ORS was expanded and that Dickins was appointed as its head. Saward, 112. One wonders at the relationship between Dickins and Saward given that Dickins had been head of the ORS for some months and had had experience with Bomber Command issues for several more. Indeed, though he had worked closely with the scientists, Saward’s account contains only three references to the ORS. Is one to get the sense that there was no love lost between the promoters of radar aids and those who measured their effectiveness? 36 PRO AIR 14/1769 ‘Note on Attack on Essen March 8/9th–June 8/9th,’ 26 July 1942. 37 PRO AIR 14/1769 ‘Note on Attack on Essen March 8/9th–June 8/9th,’ 26 July 1942. 38 PRO AIR 14/1769 Minute 40, CNavO to OIC ORS, 17 July 1942. 39 PRO AIR 14/1769 Minute 41, G/C Ops to OIC ORS, 20 July 1942. 40 PRO AIR 14/1769 Minute 43, SASO to OIC ORS, 24 July 1942. 41 PRO AIR 14/1769 Minute 47, CinC to SASO, 19 August 1942. 42 Michael Cumming, Beam Bombers: The Secret War of No. 109 Squadron (Thrupp, Gloucestershire: Sutton Publishing, 1998), 61; McCloskey, ‘British Operational Research in World War II,’ 458–9. 43 Signals Branch Headquarters Bomber Command, ‘An Historical Account of the Radio Countermeasures Offensive,’ 292. 44 Webster and Frankland, SAO, I, 418. 45 Middlebrook and Everitt, BCWD, 297; Richards, Portal, 306. 46 PRO AIR 14, 516 Letter SASO to Groups, 3 November 1941. There are documents in the same file from Group Captain Bufton, D/DBOps, at the end of November pushing the idea of target-finding squadrons. See also Air Ministry, Operational Research in the RAF, 48–9; Signals Branch Headquarters Bomber Command, ‘An Historical Account of the Radio Countermeasures Offensive,’ 58. 47 Webster and Frankland, SAO, I, 418–27. 48 Webster and Frankland, SAO, I, 422–3. 49 John Maynard, Bennett and the Pathfinders (London: Arms and Armour, 1998), 53–64. See also Greenhous et al., The Crucible of War, 611–13. 50 Webster and Frankland, SAO, I, 428–32. 51 Webster and Frankland, SAO, I, 434–6.

300

Notes to pages 77–82

52 Harris, Bomber Offensive, 131. 53 PRO AIR 14/1804 Minute 2, OIC ORS to CinC, 22 September 1942, covering ORS Loose Memorandum ‘An Assessment of the Success of Operations Led by the P.F.F., 22 September 1942.’ During the fall of 1942 the ORS was able to conduct two reviews of the success of pathfinding. The first of these looked at the results of thirteen raids and found that six had yielded improvements while only two had been failures. A more detailed study of the first twenty-one raids was carried out later in the year. Report M 117 focused on the causes of failure: inadequate numbers of finder aircraft, malfunctioning flares, and the inadequacy of visual marking devices underlining the need for coloured TIs. Overall the report concluded that pathfinding had been successful on one-third of the raids over Germany and partially successful during another third. Over Italy results had been better, with success on seven of nine attacks. Signals Branch Headquarters Bomber Command, ‘An Historical Account of the Radio Countermeasures Offensive,’ 59. 54 PRO AIR 14/1804 Minute 3, CinC to OIC ORS, 23 September 1942. 55 PRO AIR 14/1804 ORS Loose Memorandum ‘Notes on Effectiveness of P.F.F. Operations to 21/22 November,’ n.d. 56 Webster and Frankland, SAO, I, 434–6. 57 PRO AIR 14/900 Minute 1, ORS to SASO, 18 June 1942 covering ORS Report S 53 ‘The Operational Use of Oboe Mark I for Target Location,’ dated 18 June 1942; See also Air Ministry, Operational Research in the RAF, 49–50; Signals Branch Headquarters Bomber Command, ‘An Historical Account of the Radio Countermeasures Offensive,’ 74–7. 58 PRO AIR 14/900 Minute 2, SASO to CinC, 20 June 1942; Minute 3, CinC to SASO, 20 June 1942; Letter HQ BC to Air Ministry, 21 June 1942. 59 PRO AIR 14/900 Letter, Air Ministry to HQ BC, 25 June 1942. 6. Concentration and Other Curatives: February–December 1942 1 2 3 4

Webster and Frankland, SAO, I, 397–400. Webster and Frankland, SAO, I, 402. Harris, Bomber Offensive, 108–9. Messenger, Bomber Harris, 75–6. Messenger’s reference to Dickins had been taken from Alfred Price, Battle over the Reich (London: Ian Allan 1973), 28–9. 5 PRO AIR 14/232 Minute 113, G/C Ops to SASO, 30 August 1941; Minute 115, SASO to OIC ORS, 1 September 1941; Minute 117, OIC ORS to SASO, 4 September 1941. 6 PRO AIR 14/396 Minute 24, G/C Ops to SASO, 8 November 1941, covering

Notes to pages 82–4 301

7

8

9

10

11

12

13 14

Letter D/CAS to AOCinC Bomber Command, 3 November 1941; Minute 25, SASO to AOCinC, 9 November 1941; Minute 27, AOCinC to SASO, 10 November 1941; ORS Report No. 9 ‘Preliminary Note on the Relation between Concentration in Time over a Target and Bomber Losses,’ dated 1 November 1941. Catseye fighters were those which relied on visual means – the eyes – to close with and attack a bomber. AI fighters relied on radar carried in the aircraft. PRO AIR 14/396 Minute 34, Ops 1 (b) to G/C Ops and SASO, 5 December 1941, covering letter D/CAS to AOCinC Bomber Command, 3 December 1941; Minute 35, OIC ORS to SASO, 11 December 1941; Enc 52A, Letter HQ BC to Groups, 3 March 1942. PRO AIR 14/396 BC ORS Report No. 29 ‘An Investigation into the Effect on Bomber Losses of Concentration in Time at the Target and the Duration of the Attack,’ dated 11 March 1942. PRO AIR 14/396 Letter Saundby to USS, 18 March 1942, covering BC ORS Report No. 29 ‘An Investigation into the Effect on Bomber Losses of Concentration in Time at the Target and the Duration of the Attack,’ dated 11 March 1942. PRO AIR 14/396 ORS Report No. 34 ‘Preliminary Note on the Effect on Bomber Losses on Concentration in Space en Route,’ 30 May 42; Minute 62, Letter Bomber Command to Air Ministry, 4 June 1942. PRO AIR 14/396 Minute from A/C Harrison to Dickins, 4 June 1942. In Rise of the Boffins, Clark says that one of the ORS’s great successes was promoting and then optimizing concentration. This was first done, he says, as part of the Cologne raid, but as time went on the process was refined with great attention being paid to the analysis of each raid. ‘More important still, it was possible to estimate not only what had happened but why, thus providing a pointer to future operations that became more important as operational tactics and scientific equipment grew in complexity.’ Clark, The Rise of the Boffins, 221. PRO AIR 14/363 Minute 93, CIO to Int 3, 4 February 1942; Minute 94, Int 3 to CIO, 6 February 1942; Minute 95, CIO to SASO (through ORS), 6 February 1942. Marginal note; Minute 96, OIC ORS to SASO, 9 February, 1942 PRO AIR 14/795 Letter DBOps to HQ BC SASO, 11 April 1942; Minute 66, SASO to OIC ORS, 12 April 1942. PRO AIR 14/795 ORS Report No. 44 ‘Comparison of Losses of Manchester and Stirling Aircraft,’ 3 June 1942; Minute 67, OIC ORS to SASO, 29 May 1942; Minute 68, SASO to OIC ORS 30 May 1942; Letter HQ BC to DBOps, 4 June 1942 covering ORS Report No. 44 ‘Comparison of Losses of Man-

302 Notes to pages 84–6

15

16

17

18 19 20 21

22

23

24

chester and Stirling Aircraft,’ 3 June 1942; Minute 70, D/SASO to CinC, 7 June 1942. PRO AIR 14/1773 ORS Report 28 ‘Preliminary Note Comparing the Number of Casualties Not Due to Enemy Action of Aircraft on Operational Sorties during the months October–December, 1940 and 1941,’ 24 January 1942. PRO AIR 14/1773 Loose Minute 3, W/C Eng 3 to CEngO, 14 February 1942. The Wellington was the last of the two-engined bombers to be used in the Command. See appendix 7. PRO AIR 14/1773 ORS 33 ‘Note on Bomber Command Aircraft Reported Missing during the Period May 1941–April 1942,’ 5 April 1942. The copy of the report in this file is a draft with pencil corrections. Dickins refers to it as Report 33 but the final document was issued as S 43 on 22 April 1942 and is found in the RAF AHB catalogue of ORS Reports. Minute 7, OIC ORS to SASO, 10 April 1942, covering ORS 33 ‘Note on Bomber Command Aircraft Reported Missing during the Period May 1941–April 1942,’ 5 April 1942. PRO AIR 14/1773 Minute 9, AOA to SASO, 14 April 1942; Minute 10, AOT to AOA, SASO, 16 April 1942. PRO AIR 14/1773 Minute 24, OIC ORS to CinC (through SASO), 22 May 1942; Minute 25, CinC to SASO, ORS, 28 June 1942. PRO AIR 14/1773 Minute 32, Note to file on fuel shortage investigation, n.d. PRO AIR 14/1773 ‘Bomber Command Flying Accident Review No. 9 April, 1944,’ n.d. Of note in April 1944 No. 6 (RCAF) Group had the best record for operational groups in the command with a rate of 0.0016 per cent. PRO AIR 14/1777 Minute 6, OIC ORS to CEngO, 2 May 1942, covering ORS Report 42 ‘The Effect of Icing Conditions on the Percentage of Aircraft Missing,’ 24 August 1942; Loose Minutes, Eng 3 to CEngO (through Eng 4) 4 May 1942; Eng 4 to CEngO 11 May 1942; Minute 11, OIC ORS to SASO (through CEngO), 15 [May] 1942; Minute 14, OIC ORS to CinC, 24 July 1942; Minute 14, OIC ORS to CinC, 24 July 1942. PRO AIR 14/1777 Minute 32, OIC ORS to SASO (through CEngO), n.d., covering ‘Effect of Icing on Bomber Operations Return for the Month of October 1942,’ n.d. From this point on the file is generally quiet. There is some fragmentary discussion involving the scientists about the use of ‘Killfrost,’ which appears to have been a de-icing fluid for wings and tail surfaces, but other than a brief engineering minute about its not being of use on the Lancaster or Halifax there is little indication of its employment. PRO AIR 14/1799 Minute 46, OIC ORS to SASO (through Int 3), 27 April

Notes to pages 86–91 303

25

26

27

28 29 30 31

32 33

34

35 36 37

38 39

40 41

1942, covering ORS Draft Report ‘Note on Bomber Command Aircraft Reported Missing during the Period May 1941–March 1942.’ PRO AIR 14/1799 Minute 52, OIC ORS to SASO, 22 May 1942; Loose Table (Enclosure 30A) ‘Comparison of Losses of Bombers Having Air & Liquid Cooled Engines: Operational Flying by Night,’ n.d. PRO AIR 14/1794 Minute 2, OIC ORS to CinC (through SASO), 12 August 1942, covering ORS Report 48 ‘Losses of Halifax Aircraft July 1941–June 1942,’ 30 July 1942. See also Greenhous et al., The Crucible of War, 607, 683–5. PRO AIR 14/1794 Minute 3, CinC to OIC ORS, 13 August 1942. The use of the question mark was Harris’s; perhaps he was postulating a hypothesis for the scientists to examine rather than stating a fact. PRO AIR 14/1794 Minute 5, OIC ORS to CinC, 17 August 1942. PRO AIR 14/1794 Minute 12, OIC ORS to SASO, 6 October 1942. PRO AIR 14/1794 Minute 12, OIC ORS to SASO, 6 October 1942. PRO AIR 14/1846 ORS B 116 ‘The Incidence of the Experience of Pilots on the Missing Rate,’ 3 October 1942 (the document was originally drafted as Report 55 but the number was changed at the draft stage). Minute 2, OIC ORS to SASO, 1 October 1942, covering ORS B 116 ‘The Incidence of the Experience of Pilots on the Missing Rate,’ 3 October 1942. PRO AIR 14/1846 Minute 4, SASO to AOT, 4 October 1942; Minute 5, AOT to SASO, 13 October 1942; Minute 9, OIC ORS to CinC, 2 November 1942; Minute 10, CinC to OIC ORS, 4 November 1942. ‘Conversion to type’ refers to the training necessary for a crew to convert to and become proficient at flying a new aircraft type. PRO AIR 14/1846 Letter AOC 4 Group to HQBC, 14 November 1942; Minute 13, SASO to CinC, 17 November 1942, covering Letter AOC 4 Group to HQBC, 14 November 1942; CinC to SASO, 17 November 1942. PRO AIR 14/1846 Minute 19, OIC ORS to SASO, 16 November 1942; Minute 20, SASO to CinC, 17 November 1942. PRO AIR 14/1846 Minute 21, CinC to SASO, 17 November 1942. PRO AIR 14/1846 Minute 26, OIC ORS to CinC (through SASO), 24 December 1942, covering ORS B 119 ‘The Effect of Pilots’ Operational Experience on the Casualty Rate in 5 Group,’ 24 December 1942. Clark, The Rise of the Boffins, 222–3. PRO AIR 14/1790 Minute 41, OIC ORS to SASO (through CEngO), 13 December 1942, covering ORS S 74 ‘A Note on the Fire Risk in Bomber Aircraft,’ 22 December 1942. PRO AIR 14/1790 Minute 42, CEngO to SASO, 13 December 1942. PRO AIR 14/1790 G/C Eng to CEngO, SASO, 29 December 1942.

304 Notes to pages 91–6 42 PRO AIR 14/1790 SASO to OIC ORS (through Ops 1 (e)), 1 January 1943. 43 PRO AIR 14/1790 ‘Explosion and Fire Risk in Petrol Tanks of Bomber Aircraft,’ n.d. 44 PRO AIR 14/1790 Minute 124, CEngO to D/CinC, 12 April 1943, covering ‘Explosion and Fire Risk in Petrol Tanks of Bomber Aircraft,’ n.d. 45 PRO AIR 14/1800 Minute 3, OIC ORS to CinC (through SASO), 21 December 1942, covering ORS Draft Report ‘Investigation into the Bomber Losses Sustained on the Night of 17/18th December, 1942,’ n.d. 46 PRO AIR 14/1800 Minute 5, CinC to SASO, OIC ORS, 28 December 1942. Emphasis in original. 47 PRO AIR 14/69 Letter Harris to Under-Secretary of State and CAS, dated 24 March 1942 ‘Air Navigation – General Policy.’ 48 PRO AIR 14/516 BC ORS Report No 31, ‘The Visual recognition of Ground Features as an Aid to Target Identification at Night,’ dated 14 February 1942. 49 Dickins, OR in Bomber Command, 52–3. 50 PRO AIR 14/508 ‘Minutes of Conference Held at Headquarters Bomber Command on Thursday, April 16th 1942.’ 51 PRO AIR 14/232 Minute 140, SASO to ORS, 30 May 42; Minute 142, Dickins to SASO, 5 June 1942; Minute 141, covering ORS Memo ‘The Effect of Directness of Route on Bomber Losses,’ dated 5 June 1942. 52 Harris, Bomber Offensive, 108–9. 53 Barker, The Thousand Plan, 44–8. 54 PRO AIR 14/276. The operation order for the raid did list Hamburg as the primary target, with Cologne as the alternate. Of note, the attack at Hamburg was planned for sixty minutes (16.7 aircraft per minute for 1000 raiders) while Cologne was scheduled for ninety minutes (11 aircraft per minute). BC Operation Order No 148, dated 26 May 1942. 55 Barker, The Thousand Plan, 44–8 56 Saward, Bomber Harris, 142. 57 Webster and Frankland, SAO, I, 403–9. 58 Harris, Bomber Offensive, 120–1. 59 Webster and Frankland, SAO, I, 410–12. 60 Harris, Bomber Offensive, 108–9, 120–1. 61 Webster and Frankland, SAO, I, 413–16. 62 PRO AIR 14/1814 ‘Minutes of a Meeting Held in CAS’s Room on 30th June 1942 to Consider the Use of Bomber O.T.U’s in Operations,’ dated 30 June 1942. 63 PRO AIR 14/1814 ORS Report S 70 ‘The Use of O.T.U’s on Operations,’ dated 23 October 1942.

Notes to pages 96–104 305 64 Webster and Frankland, SAO, I, 417. 65 PRO AIR 14/363 ORS Report 18 ‘Investigation into Our Bomber Losses Sustained on the Night of 7/8th November,’ dated 14 November 1941. 66 Harris, Bomber Offensive, 122–3. 67 Greenhous et al., The Crucible of War, 607–9. 68 PRO AIR 14/363 ORS Report S 64 ‘Investigation into the Bomber Losses Sustained on the Nights of 27/28th and 28/29th August, 1942,’ September 1942. 69 Harris, Bomber Offensive, 133. 70 PRO AIR 14/1761 ORS B 120 ‘On the Routeing of Night Bombers through Areas Defended by Fighters,’ 18 December 1942. 71 PRO AIR 14/1761 ORS B 120 ‘On the Routeing of Night Bombers through Areas Defended by Fighters,’ 18 December 1942. 72 PRO AIR 14/1798 Minute 3, OIC ORS to SASO (through AOT), 27 August 1942, covering ‘Proportions of Our Aircraft Intercepted Attacked Damaged and Missing on Night Operations,’ n.d.; W/C Trg 1 to SASO, 28 August 1942. 73 Harris, Bomber Offensive, 90. 7. Putting Concepts to the Test: January–February 1943 1 Webster and Frankland, SAO, II, 90–7. See also Middlebrook and Everitt, BCWD, 334–6, 362–4; Terraine, Right of the Line, 515–17 2 Webster and Frankland, SAO, II, 98–101. 3 Webster and Frankland, SAO, II, 101–2. See also Harris, Bomber Offensive, 159–61. The CinC was fairly happy with the results achieved by skymarking. 4 See Lovell, Echoes of War, 151–3, for an assessment of the initial performance of the device from the perspective of its creators. Theirs was a somewhat more enthusiastic interpretation. 5 Webster and Frankland, SAO, II, 101–2. 6 Webster and Frankland, SAO, II, 102–7. See also Kirby, Operational Research in War and Peace, 152–3, for discussion of modifications of Newhaven over the course of 1943. 7 PRO AIR 14/ 2984 ORS B 123 ‘Report on Attack on Wilhelmshaven 18/19th February, 1943,’ 2 March 1943. 8 PRO AIR 14/ 2984 ORS B 123 ORS B 124 Report ‘The Attack on Wilhelmshaven, 19/20 February, 1943,’ n.d. 9 PRO AIR 14/1858 D/CinC to CinC, 6 March 1943; Minute 34, D/CinC to SASO, 6 March 1943. 10 Greenhous et al., Crucible of War, 685–6.

306 Notes to pages 105–10 11 PRO AIR 14/1855 Minute 5, OIC ORS to CNavO, T2(N), 3 January 1943; Minute 6, OIC ORS to SASO (through AOT), 18 January 1943. 12 PRO AIR 14/4545 ORS Report 64, ‘A Review of the Problem of Visual Identification of Targets and Landmarks at Night,’ 31 January 1943. (This is the date of the final version of the report.) 13 PRO AIR 14/4545 ORS Report 64, ‘A Review of the Problem of Visual Identification of Targets and Landmarks at Night,’ 31 January 1943. (This is the date of the final version of the report.) 14 PRO AIR 14/1855 Minute 7, AOT to SASO, 23 January 1943; Minute 8, SASO to OIC ORS, 29 January 1943; Minute 16, CinC to SASO, 16 February 1943. 15 PRO AIR 14/1855 Minute 34, OIC ORS to D/CinC, 5 March 1943, covering ‘Minutes of the Meeting Held at Headquarters, Bomber Command at 1500 Hours on 1/3/43 to Discuss Certain Points Arising Out of O.R.S.(B.C.) Report No. 64,’ 4 March 1943. Roberts was section head for ‘Success on Operations’ while Stevens worked on ‘visual identification’ as part of that team. 16 PRO AIR 14/1855 Minute 52, OIC ORS to ORS 1, 24 March 1943, covering Letter MAP SR 1 to OIC ORS ‘ORS/BC Report Number 64,’ 20 March 1943, covering Memorandum ‘Progress Report in Connection with O.R.S.B.C. Report No. 64 on “Visual Identification of Targets at Night,”’ 16 March 1943. 17 PRO AIR 14/451 ‘Minutes of Navigation Conference Held At Headquarters Bomber Command on the 14th January 1943,’ n.d. 18 PRO AIR 14/516 Letter SASO to Groups, ‘Timing of Bombing Attacks,’ 27 January 1943. 19 PRO AIR 14/1819 Minute 7, OIC ORS to SASO, 2 January 1943; Minute 8, SASO to CinC, 4 January 1943; Minute 9, CinC to SASO, Dr Dickins, 5 January 1943; Minute 10, OIC ORS to SASO, 21 January 1943; SASO to OIC ORS, 23 January 1943. 20 PRO AIR 14/1819 Letter OIC ORS to OIC OR Centre, 25 January 1943 21 PRO AIR 14/364 ORS Report G 66 ‘Review of Sorties, Losses and Interceptions of Bomber Aircraft in Night Operations during the Period 1st Aug., 1941–31st. October, 1942,’ dated 13 January 1943. 22 PRO AIR 14/364 ORS Report G 66 ‘Review of Sorties, Losses and Interceptions of Bomber Aircraft in Night Operations during the Period 1st Aug., 1941–31st. October, 1942’ dated 13 January 1943. 23 PRO AIR 14/206 TC 31, ‘Tactical Countermeasures to Combat Enemy A.A. Searchlights and Guns,’ February 1943. This was an updated version of TC 22. 24 PRO AIR 14/206 Untitled Minute Ops.1(e) to G/C Ops, 19 February 1943. 25 PRO AIR 14/1979 5 Group file: Minute 3, SASO to AOC, 3 February 1943.

Notes to pages 110–16 307 26 PRO AIR 14/1979 5 Group file: Minute 30, ORS to AOC, 11 February 1943. 27 PRO AIR 14/1979 5 Group file: Minute 31, AOC to SASO, 11 February 1943. 28 PRO AIR 14/1357 Minute 62, OIC ORS to SASO (through CSO), 12 January 1943, covering ORS B 119 ‘The Introduction of “Monica” in Bomber Command,’ 12 January 1943. 29 PRO AIR 14/1357 Minute 68, CSO to D/CinC, 2 March 1943. 30 PRO AIR 14/1357 Minute 72, OIC ORS to D/CinC, 31 March 1943. 31 PRO AIR 14/1357 Minute 72, OIC ORS to D/CinC, 31 March 1943; marginal note, 31 March 1943. 32 PRO AIR 14/1357 Minute 79, D/CinC to CinC, 25 May 1943; marginal note, 25 May 1943. 33 PRO AIR 14/1803 Minute 12, OIC ORS to SASO, 14 January 1943, covering Letter Prof Mott, ORG to OIC ORS, HQBC, 8 January 1943; Minute 13, SASO to OIC ORS, 20 January 1943. 34 PRO AIR 14/1803 Minute 20, OIC ORS to SASO (through AOT), 31 January 1943, covering ORS S 77 ‘Casualties among Aircrew Personnel Directly Due to Enemy Action on Night Operations,’ 17 February 1943. This is the date of the final version of the report and reflects the changed header described in this exchange of minutes. The sequence of minutes has been verified, as they do appear out of order by number. Minute 19, SASO Note to file regarding ORS S77, 7 February 1943. 35 PRO AIR 14/1803 Minute 25, OIC ORS to SASO, 23 February 1943. 36 PRO AIR 14/1786 Minute 29, OIC ORS to SASO (through Ops1 (e)), 8 February 1943; Minute 30, SASO to Ops 1(e), 9 February 1943. 37 PRO AIR 14/1786 Minute 62, OIC ORS to Ops 1 (e), SASO, D/CinC, 22 March 1943; Minute 63, Ops 1 (e) to OIC ORS, 22 March 1943; Minute 66, Ops 1 (e) to SASO, 25 March 1943. 38 PRO AIR 14/1746 Letter OIC Operational Research Centre to OIC BC ORS, 17 February 1943; Letter OIC ORS to Operational Research Centre, 23 February 1943, covering ‘Note on the Immobilization of Turrets Due to Damage by Enemy Action to the Hydraulic System in Lancasters, Stirlings and Wellingtons,’ 23 February 1943. 39 PRO AIR 14/1746 Letter OIC Operational Research Centre to OIC BC ORS, 26 February 1943, covering Air Ministry OR2 ‘Engine Armour Protection in Bomber Aircraft,’ 24 February 1943. 40 PRO AIR 14/1746 Letter OIC ORS to OIC OR Centre, 12 March 1943. 8. New Concepts, Old Targets – The Ruhr: March–June 1943 1 Webster and Frankland, SAO, II, 102–10.

308

Notes to pages 116–22

2 Webster and Frankland, SAO, II, 114–18. See also Messenger, Bomber Harris, 111. 3 Harris, Bomber Offensive, 145–6. 4 PRO AIR 14/1812 Minute 25, OIC ORS to D/CinC, covering ORS S 80 ‘Comparative Effect of 4 lb & 30 lb Incendiary Bombs Attacks on Genoa 22/23 October, 6/7 November & 15/16 November, 1942,’ 20 March 1943; Minute 26, D/CinC to OIC ORS, 23 March 1943; Minute 27, OIC ORS to D/CinC, 27 March 1943. 5 PRO AIR 14/1812 Letter OIC ORS to OIC Operational Research Centre, 7 October 1943. 6 PRO AIR 14/3939 ORS Report No. 76 ‘The Relative Efficiencies of Bombloads for Attack on German Towns,’ 27 June 1943. 7 High Capacity bombs had a relatively thin casing and a significant amount of high explosives (HE) packed in. They did their damage by blast rather than shrapnel. The General Purpose bomb had a thicker casing and created damage by a combination of HE blast effect and flying shrapnel. 8 PRO AIR 14/3939 ORS Report No. 76 ‘The Relative Efficiencies of Bombloads for Attack on German Towns,’ 27 June 1943. 9 PRO AIR 14/3939 Minute 15, CArmO to OIC ORS (through G/C Ops), 30 June 1943; Minute 16, G/C Ops to OIC ORS (through CArmO), 6 July 1943; Minute 19, OIC ORS to D/CinC, 15 August 1943. 10 Webster and Frankland, SAO, II, 119–20. 11 Webster and Frankland, SAO, II, 118–24. 12 Webster and Frankland, SAO, II, 129. 13 PRO AIR 14/1858 Loose Minute W/Cdr RDF to CSO, 8 April 1943. 14 Dickins, OR in Bomber Command, 71. 15 PRO AIR 14/1858 Minute 55, CSO to D/CinC, 9 April 1943. 16 PRO AIR 14/1858 Minute 56, D/CinC to CSO, 11 April 1943. 17 PRO AIR 14/1858 Minute 61, OIC ORS to D/CinC, 14 April 1943, covering ORS B 130 ‘Attack on Munich 9/10 March 1943,’ and ORS B 131 ‘Attack on Stuttgart, 11/12 March 1943. 18 PRO AIR 14/1804 Minute 15, OIC ORS to D/CinC, 20 April 1943. See also Dickins, OR in Bomber Command, 60–1. Minute 16, DCinC to CinC 22 April 43. 19 PRO AIR 14/1804 ORS B 133 ‘Effectiveness of the P.F.F. Phase II – January to March, 1943,’ 21 April 1943. 20 PRO AIR 14/1858 Minute 67, OIC ORS to D/CinC, 29 April 1943. See also Dickins, OR in Bomber Command, 70–1. 21 PRO AIR 14/1858 Minute 93, OIC ORS to Ops 1(b), 27 June 1943. 22 PRO AIR 14/1858 Minute 94, Ops 1(b) to SASO, 28 June 1943. 23 PRO AIR 14/3025 Minute 17, OIC ORS to D/CinC, July 1943, covering

Notes to pages 123–7 309

24 25

26 27

28 29

30 31

32

33

34 35 36

ORS B 151 ‘Review of H2S Groundmarking Raids on Germany February– April, 1943,’ n.d. See also Dickins, OR in Bomber Command, 65–8. Greenhous et al., The Crucible of War, 667. PRO AIR 14/804 ‘Minutes of the Meeting Held at Headquarters, Bomber Command, 1100 Hours, 23rd April, 1943 to Discuss Tactical Aspects Arising from Recent Operations,’ 26 April 1943. PRO AIR 14/3935 CNavO Loose Note ‘Two Theories on Some of the Causes for Bombs Falling Short of the Target,’ n.d. PRO AIR 14/3943 Loose Minute OIC ORS to SASO, 11 June 1943 covering ‘Two Theories on Some of the Causes for Bombs Falling Short of the Target,’ n.d. PRO AIR 14/3943 Minute 2, SASO to OIC ORS, 12 June 1943; SASO to D/ CinC, 15 June 1943. PRO AIR 14/3943 Minute 7, ORS 1(c) to OIC ORS, n.d., covering Loose Note ‘Notes on the Reasons for Undershooting in Night Operations,’ n.d. See also Cumming, Beam Bombers, 104–5. PRO AIR 14/3943 Minute 11, n.d., covering ‘Notes on the Reasons for Undershooting in Night Operations,’ 30 July 1943. PRO AIR 14/3943 Minute 12, ORS 1(c) to ORS 1, 19 August 1943. Interestingly, the file does not show what happened to these recommendations; the next entry deals with only the cross-wind problems and what to do about them – these reached the SASO who felt they could not be dealt with at the moment. PRO AIR 14/1846 Minute 2, OIC ORS to CinC (through D/CinC, SASO, A/Cdre Ops, G/C Ops, CIO), covering ORS Memorandum 71 ‘Relation between the Centre of a Night Photograph and the Fall of Bombs,’ 24 May 1943. (Note: the General series switched back and forth between Report and Memorandum for some reason. This document is part of that series.) See also Air Ministry, Operational Research in the RAF, 29–30. PRO AIR 14/2017 Minute 20, SASO to AOC, 6 June 1943, covering Letter HQ BC to Groups, 4 June 1943, covering ORS Memorandum 71 ‘Relation between the Centre of a Night Photograph and the Fall of Bombs,’ June 1943; Minute 21, Photo to SASO, 7 June 1943; Minute 22, SASO to Photo, 11 June 1943. PRO AIR 14/2017 Minute 24, Air 1 to ORS, 4 November 1943; Minute 25, ORS to Air 1, 4 November 1943. PRO AIR 14/ 1746 Letter Deputy Director Scientific Research 3, MAP to OIC ORS, ‘Directions of Attack by German Night Fighters,’ 3 March 1943. PRO AIR 14/ 1746 Letter OIC ORS to DDSR 3 MAP ‘Directions of Attack by German Night Fighters,’ 12 March 1943.

310

Notes to pages 127–32

37 PRO AIR 14/ 1746 Letter DDSR 3, MAP to OIC ORS, 10 July 1943; Letter OIC ORS to DDSR 3, MAP, 15 July 1943. 38 PRO AIR 14/ 1746 Letter OIC ORS to DDSR 3, MAP, 27 August 1943. 39 PRO AIR 14/1800 Letter ORS 5 Group to OIC ORS, 15 February 1943, covering ORS 5 Group Report ‘9 Squadron’s Missing Rate since the Commencement of Their Operations in 5 Group,’ n.d.; OIC ORS to ORS 2, 25 February 1943. 40 Webster and Frankland, SAO, II, 125–37. 41 Webster and Frankland, SAO, II, 141–6. 42 PRO AIR 14/364 Minute 18, OIC ORS to ORS Report S 91 ‘Night-Bomber Losses on German Targets, 1942,’ 12 April 1943. 43 PRO AIR 14/1800 Loose Minute 1, OIC ORS to D/CinC, 8 May 43, covering ORS S91 ‘Night-Bomber Losses on German Targets, 1942,’ 12 April 1943; Loose Minute 2, D/CinC to OIC ORS, 7 July 1943; Letter OIC ORS to Distribution List, 7 August 1943. 44 PRO AIR 14/1800 Minute 54, OIC ORS to DCinC, 14 April 1943, covering ORS Draft Report ‘A Note on the Trend of Lancaster Losses,’ 29 March 1943; Minute 55, DCinC to OIC ORS, 15 April 1943. 45 Reuben Jacob Smeed was born on 1 September 1909 and died on 3 September 1967. He attended Imperial College, earning a BSc in mathematics in 1931 and a PhD in Aerodynamics in 1933, after which he joined the faculty. In 1939 he joined the RAE and did operational research on radio and radar systems. He became head of Bomber Command ORS in 1945 before returning to civilian academic life in 1947, where he specialized in road and traffic safety. 46 PRO AIR 14/1800 Minute 62, ORS 2 to OIC ORS, 20 April 1943, covering ORS Report 74 ‘An Investigation into the Effect of Gee on Casualties,’ 15 March 1943. 47 PRO AIR 14/1800 Loose Minute OIC ORS to D/CinC, 30 June 1943, covering ORS Report 74 ‘An Investigation into the Effect of Gee on Casualties,’ 15 March 1943; D/CinC Marginal note, 16 July 1943, on Loose Minute OIC ORS to D/CinC, 30 June 1943. 48 PRO AIR 14/1800 ORS B141 ‘A Note on the Losses of the P.F.F.,’ 27 May 1943. 49 PRO AIR 14/1791 Minute 1, CEngO to D/CinC, 23 June 1943. 50 PRO AIR 14/1791 Minute 3, OIC ORS to D/CinC (through CEngO), SASO, Ops 1(d), 9 August 1943; Minute 4, D/CinC to OIC ORS, 14 August 1943. 51 PRO AIR 14/1791 Minute 8, ORS 2 to OIC ORS, 15 September 1943. 52 ORS S 74 ‘A Note on the Fire Risk in Bomber Aircraft,’ 22 December 1942, discussed previously.

Notes to pages 133–40 311 53 PRO AIR 14/1791 Minute 9, CEngO to SASO, 9 September 1943; Minute 19, CEngO to D/SASO, 21 September 1943; Minute 20, D/SASO to SASO (through CEngO), 22 September 1943; Minute 21, OIC ORS to SASO (through CEngO), 5 November 1943; Minute 22, CEngO to SASO, 8 November 1943; Minute 24, D/CinC to SASO, 14 November 1943. The file shows the debate over redundancy and weight continuing into early 1944 with CEngO against overcomplicating life. What he had against maximizing the survival chances of the crews is not discernible from the minutes on this and related files. He seems to have been ready to believe hard fact and physical evidence, as evinced by his earlier change of heart on using inert gas, but such was most certainly not the case when it came to Methyl Bromide tank bay extinguishers. 54 PRO AIR 14/1746 Letter Director of Servicing and Maintenance, Air Ministry to OIC ORS ‘Vulnerability of Aero Engines against Enemy Action – Comparison of In-Line and Radial Engines,’ 8 June 1943; Letter OIC ORS to DSM, Air Ministry ‘Vulnerability of Aero Engines against Enemy Action – Comparison of In-Line and Radial Engines,’ June 1943. 55 PRO AIR 14/3958 ORS B 140 ‘ Note on the Number of Incendiary Bombs Falling on Bomber Aircraft in Concentrated Raids,’ 30 June 1943. 56 PRO AIR 14/1794 Minute 13, OIC ORS to DCinC, 30 June 1943; Minute 14 DCinC to SASO, 3 July 1943. 57 PRO AIR 14/872 Minute 1, AOT to SASO, 13 May 1943; Minute 2, D/SASO to AOT 4 June 1943; Minute 3, unknown to SASO, n.d. (text missing). See also Messenger, Bomber Harris, 124. 58 PRO AIR 14/872 Letter HQ BC to AOCs 1, 3, 4, 5, 6, PFF Groups, 5 May 1943. See also Air Ministry, Operational Research in the RAF, 64–5. 59 PRO AIR 14/477 ‘Bomber Command Tactical Memorandum - Evasive Action at the Target,’ 24 May 1943. 60 What they were talking about, indirectly, was something that pilots and others call ‘the big sky little bullet theory.’ In this popular construct it is argued that there is almost no chance of hitting some object that occupies a relatively infinitesimal volume of the airspace around the aircraft. The same theory was in play when it came to the chances of being hit by falling bombs or colliding with another aircraft. 61 PRO AIR 14/3958 Minute 89, OIC ORS to DCinC, 30 June 1943, covering Letter Bomber Command HQ to Groups, 13 July 1943. 62 Webster and Frankland, SAO, II, 141–6. 9. Hamburg and New Science: July–October 1943 1 Webster and Frankland, SAO, II, 146–50.

312 Notes to pages 140–8 2 3 4 5 6 7 8

9 10 11

12 13

14

15

16 17

18 19 20

Webster and Frankland, SAO, II, 151. Middlebrook and Everitt, BCWD, 410–11. Webster and Frankland, SAO, II, 157–8. Webster and Frankland, SAO, II, 158–62. See also Lovell, Echoes of War, 180–1. PRO AIR 14/1753 ORS Report 82 ‘Analysis of Navigation – Raid against Nuremburg – August 27/28th 1943,’ 7 October 1943. PRO AIR 14/1753 ORS Report 82 ‘Analysis of Navigation – Raid against Nuremburg – August 27/28th 1943,’ 7 October 1943. PRO AIR 14/1753 ORS Report 86 ‘Analysis of Navigation – Raid against Hannover – Sept. 22/23rd 1943,’ 17 November 1943. See also Messenger, Bomber Harris, 139. PRO AIR 14/1753 ORS Report 89 ‘Analysis of Navigation – Raid against Kassel October 22/23rd, 1943,’ 30 December 1943. PRO AIR 14/1770 ORS 1b to OIC ORS, 22 October 1943, covering ORS S120 ‘G-H Attack on Düsseldorf 3rd November, 1943,’ 7 November 1943. PRO AIR 14/1770 OIC ORS to D/CinC (through G/C Radar), 7 November 1943, covering ORS S120 ‘G-H Attack on Düsseldorf 3rd November, 1943,’ 7 November 1943. See also Air Ministry, Operational Research in the RAF, 53–5. Harris, Bomber Offensive, 172–5. This passage gives some indication of the various issues with which the CinC was dealing at the time. PRO AIR 14/364 ORS Report G 77 ‘Monthly Review of Losses and Interceptions of Bomber Command Aircraft in Night Operations, July 1943,’ 9 August 1943 PRO AIR 14/1800 Minute 92, OIC ORS to D/CinC, 19 August 1943, covering ORS S 98 ‘The Effect of “Window” on Bomber Operations, (24/25 July–10/11 Aug.),’ n.d. PRO AIR 14/1800 Minute 92, OIC ORS to D/CinC, 19 August 1943, covering ORS S 98 ‘The Effect of “Window” on Bomber Operations, (24/25 July–10/11 Aug.)’; D/CinC to SASO 20 August 1943; SASO to D/CinC 24 August 1943. (The file contains interesting diagrams regarding waves.) Greenhous et al., Crucible of War, 678–86. PRO AIR 14/1800 Loose Minute 1, OIC ORS to CinC, 19 July 1943, covering ORS B147 ‘A Note on 6 Group Losses,’ 10 July 1943; Loose Minute 3, D/CinC to OIC ORS, 12 July 1943; Letter CinC to AOC 6 Group, 17 July 1943. PRO AIR 14/1800 Minute 89, OIC ORS to D/CinC, 26 July 1943. PRO AIR 14/1794 Loose Minute 1, OIC ORS to DCinC and CEngO, 2 July 1943. PRO AIR 14/1794 Loose Minute 1, OIC ORS to DCinC and CEngO, 2 July 1943; Loose Minute 3, CEngO to DCinC, 5 July 1943.

Notes to pages 148–56 313 21 PRO AIR 14/1794 Letter Dickins to A/M R.S. Sorley, CRD, Ministry of Aircraft Production, 2 July 1943. 22 PRO AIR 14/1875 Loose minute, OIC ORS to D/CinC, 17 September 1943. 23 PRO AIR 14/1875 OIC ORS to OIC OR Centre, 3 October 1943. 24 PRO AIR 14/1875 DD Science to OIC ORS ‘Comparison of Aircraft Types,’ 29 April 1944. See also Greenhous et al., The Crucible of War, 603–6; 754–7. 25 PRO AIR 14/1880 Minute 1, CinC to OIC to ORS, 23 September 1943; Minute 2, OIC ORS to CinC, 29 September 1943. 26 PRO AIR 14/1880 Minute 3, CinC to D/CinC, n.d.; Minute 4, D/CinC to OIC ORS, 1 October 1943; Minute 5, OIC ORS to D/CinC, 7 October 1943. 27 Dyson, Disturbing the Universe, 26–9. 28 PRO AIR 14/1880 Minute 6, OIC ORS to D/CinC (through AOT), 31 January 1944, covering ORS S 122 ‘An Examination of Emergency Escape Arrangements from Operational Aircraft,’ 28 January 1944. 29 PRO AIR 14/1880 Minute 7, T1 to AOT, 6 February 1944; Minute 8, AOT to D/CinC, 8 February 1944; Minute 12, A/Cdre Ops to AOT, 10 March 1944. 30 PRO AIR 14/1880 Minute 13, CEngO to AOT, 15 March 1944; Minute 14, CEngO to AOT, 15 March 1944. 31 PRO AIR 14/1746 ‘Minutes of a Meeting between O.R.S.B.C. and R.A.E./. to Discuss Bransom Fuel Tanks and the Need for Self-Sealing Material in Oil Tanks,’ 8 October 1943; ‘Minutes of a Meeting at M.A.P. on the 23rd December, 1943 to Discuss Self-Sealing Fuel Hoses,’ 29 December 1943. 32 PRO AIR 14/1801 Minute 2, ORS 2 to OIC ORS, 5 November 1943, covering Loose Memorandum ‘The Losses of Squadrons of Bomber Command, May– September, 1943,’ n.d. See also Greenhous et al., The Crucible of War, 718. 33 PRO AIR 14/477 ‘Tactical Countermeasures to Combat Enemy Night Fighter, A.A. Searchlight and Gun Defences Report,’ dated August 1943. 34 PRO AIR 14/1872 Minute 1, OIC ORS to D/CinC, 27 August 1943; ORS B163 ‘Enemy Defensive Tactics against the Operation on Berlin of August 23/24th, 1943,’ 26 August 1943. 35 PRO AIR 14/1872 Marginal notes 29 and 30 August to Minute 1, OIC ORS to D/CinC, 27 August 1943. 36 PRO AIR 14/1801 Minute 1, OIC ORS to D/CinC, 13 September 1943, covering ORS Report 80 ‘Some Notes on the Defensive Tactics Now Used by the Enemy,’ 11 September 1943. 37 PRO AIR 14/1301 ‘Minutes of a Meeting Held 14th August 1943 at Headquarters Bomber Command, to Discuss the Operational Application of Airborne Cigar,’ 18 August 1943. 38 PRO AIR 14/639 ORS S 104 ‘Report on the Operational Use of A.I. Boozer,’ 14 September 1943.

314 Notes to pages 156–60 39 PRO AIR 14/638 Minute 15, OIC ORS to SASO (through Ops. 1(e)), 29 June 1943. Minute 16, SASO to DCinC, 30 June 1943. Marginal note to Minute 16, SASO to DCinC, 30 June 1943. 40 PRO AIR 14/638 Minute 25, OIC ORS to SASO (through Ops. 1(e)), 9 August 1943. Marginal note to Minute 25, OIC ORS to SASO (through Ops. 1(e)), 9 August 1943 10. Science versus Attrition – Berlin: November 1943–March 1944 1 Webster and Frankland, SAO, II, 190–5. For other interpretations see also Allen, Legacy, 117–18; Middlebrook and Everitt, BCWD, 417–18, 446–51; and, Greenhous et al., Crucible of War, 730–2. 2 Harris, Bomber Offensive, 186–90. Harris offers considerable discussion to back up this pithy statement. It is instructive to see how the battle was recalled in the memory and perspective of the commander. 3 Harris, Bomber Offensive, 166–71, 182–6. By comparison, Lovell in Echoes of War contends, as might be expected given his work in fielding the system, that the bugs had been largely worked out by the time of the attack on Hamburg. He was displeased that the success of Kassel was ignored for three months. See 151–3; 175–8. 4 PRO AIR 14/3025 Minute 34, OIC ORS to D/CinC, 5 December 1943, covering ORS S 111 ‘Accuracy of H2S as a Blind-bombing Device,’ n.d. See also Air Ministry, Operational Research in the RAF, 51–3. 5 PRO AIR 14/3025 Minute 46, OIC ORS to D/CinC, 2 January 1944, covering ORS S 115 ‘The Use of H2S as an Aid to Navigation,’ 4 December 1943. 6 PRO AIR 14/1753 Letter HQ BC to Groups ‘Navigation Procedure – Broadcasting of Winds,’ 14 February 1944. 7 PRO AIR 14/1753 Letter HQ BC to Groups ‘Operational Navigation – Madeburg, January 21/22nd 1944,’ 18 March 1944. 8 Dickins, OR in Bomber Command, 147–8, 158–9. 9 PRO AIR 14/1753 Minute 5, ORS 1 to OIC ORS. 10 PRO AIR 14/1879 ORS 4 Group ‘Interim Report on Erratic Bombing,’ 19 December 1943. 11 PRO AIR 14/1879 Minute 5, OIC ORS to ORS 1, 13 January 1944, covering ORS 4 Group ‘Interim Report on Erratic Bombing,’ 19 December 1943; Minute 6, ORS 1 to OIC ORS, ORS 1(a), 15 January 1944. 12 PRO AIR 14/1884 Minute 1 ORS 1(c) to ORS 1, 10 January 1944, covering ‘The Visibility of Target Indicators with Special Reference to the Colour Vision Problem,’ 9 January 1944.

Notes to pages 161–8 315 13 14 15 16 17 18 19 20 21 22

23 24

25 26

27

28 29 30 31

Dickins, OR in Bomber Command, 62–3. PRO AIR 14/1884 Minute 2, ORS 1 to ORS 1(c), 12 January 1944. PRO AIR 14/1884 Minute 15, ORS 1(c) to ORS 1, 1 March 1944. PRO AIR 14/1884 Minute 21, T2B to A/Cdre Ops (through T 1 and Ops 1(e)), 5 April 1944. PRO AIR 14/1884 Minute 22, T2B to T1, 25 April 1944 covering Loose Minute T 2 B to AOT, 23 April 1944; Minute 23, T1 to T2B, 26 April 1944. PRO AIR 14/1884 ORS Report 99 ‘The Visibility and Recognition of Target Indicators,’ 6 May 1944. PRO AIR 14/1884 Minute 35, OIC ORS to T1, 7 July 1944, covering Loose Minute ORS 4(d) to OIC ORS, 1 July 1944. PRO AIR 14/1884 Minute 48, OIC ORS to A/C Ops (through T1), 28 November 1944, covering Loose Note, 25 November 1944. PRO AIR 14/1884 Minute 49, A/C Ops to OIC ORS, 3 December 1944. PRO AIR 14/1884 Minute 51, ORS 4(d) to OIC ORS ,19 January 1945, covering ‘Commentary for Use with Transparencies for the Air Ministry Bombing Teacher ... Issue of January 1945,’ n.d. PRO AIR 14/1804 ORS Loose Memorandum ‘A Criticism of Planning and Execution of Target Marking by the Pathfinder Force,’ 30 March 1944. PRO AIR 14/3939 Minute 21, ORS 1 (a) to ORS 1, 1 November 1943, covering ORS Draft Memorandum ‘A Review of Bomb-loads Carried on Recent Operations,’ 1 November 1943. PRO AIR 14/3939 Minute 22, ORS 1 to OIC ORS, n.d.; Minute 23, OIC ORS to ORS 1, 12 November 1943. PRO AIR 14/3939 Minute 24, ORS 1 to G/C Ops (through OIC ORS), 26 November 1943, covering ORS Draft Memorandum ‘A Review of Bombloads despatched to BERLIN and MANNHEIM on the Night of 18/19th November 1943,’ n.d. PRO AIR 14/3939 Minute 26, OIC ORS to G/C Ops, 4 December 1943, covering ORS Draft Memorandum ‘A Review of Bombloads Despatched to BERLIN and MANNHEIM on the Night of 18/19th November 1943,’ n.d. PRO AIR 14/3939 Minute 27, Note to file, 29 December 1943. PRO AIR 14/557 Letter, HQ BC to Groups ‘Operational Performance Data of Aircraft,’ 1 January 1944. PRO AIR 14/3939 ORS B 195 ‘Report on Investigation into Petrol & Bomb Loads,’ 3 February 1944. PRO AIR 14/3922 Minute 113, OIC ORS to ORS 1, ORS 2, 3 November 1943, covering Letter OR Centre to Commands, ‘Operational Research Reports,’ 30 October 1943.

316 Notes to pages 169–73 32 PRO AIR 14/3922 Minute 114, OIC ORS to CinC (through DCinC) 11 November 1943, covering Loose Table ‘The Principal Items on the Programme of O.R.S. (B.C.) October 1943,’ n.d. 33 PRO AIR 14/1794 Draft Report ‘A Note on the Losses of Halifaxes Fitted with Modified Rudders Sept. 5/6th to Dec. 3/4th, 1943,’ n.d.; Minute 24, OIC ORS to CEngO, 30 December 1943; Minute 25, Eng 3 to OIC ORS (through CEngO), 4 January 1944. 34 These comments, more particularly the last, may seem curious, but like most vehicles, be they ships or bicycles, aircraft do seem to have personalities and unique tendencies. The author has seen identical aircraft with different performance and handling characteristics. 35 PRO AIR 14/1794 Minute 26, ORS 2 to OIC ORS, 8 January 1944; Minute 27, OIC ORS to ORS 2, 10 January 1944. 36 PRO AIR 14/1794 ORS S 114 ‘A Note on the Losses of Halifax Aircraft with Modified Rudders,’ 24 February 1944 37 PRO AIR 14/1795 Minute 1, OIC ORS to CinC (through D/CinC), 22 February 1944. 38 PRO AIR 14/1795 CinC to OIC ORS, 24 February 1944. 39 PRO AIR 14/1795 OIC ORS to CinC (through D/CinC), 22 March 1944. 40 PRO AIR 14/1795 Letter, Harris to A/C/M Sir Wilfrid Freeman, 28 March 1944. 41 Webster and Frankland, SAO, II, 195–9. 42 PRO AIR 14/1885 Loose Report ‘Report on Early Returns,’ 9 December 1943. 43 PRO AIR 14/1803 Minute 35, OIC ORS to D/CinC, 29 January 1944, covering ORS S 120 ‘A Further Note on Casualties among Personnel Directly Due to Enemy Action on Night Operations,’ n.d. 44 PRO AIR 14/1803 Marginal note to ORS S 120 ‘A Further Note on Casualties among Personnel Directly Due to Enemy Action on Night Operations,’ n.d. 45 PRO AIR 14/631 ORS Report S 113, ‘A Note on Offensive Defence against Night Fighters,’12 November 1943; Letter, 5Gp to HQBC, ‘Defence against Night Fighters,’ 6 February 1944. Minute 51, W/C Ops 1 (e) to A/Cdre Ops (thru Oi/c ORS), 9 February 1944; Minute 52, OIC ORS to A/Cdre Ops 14 February 1944; Letter, HQ BC to HQ 5Gp ‘Combat Policy,’ 25 February 1944. 46 PRO AIR 14/1847 Letter, HQ 5 Group to HQ BC, ‘Effect of Experience on Casualty Rates,’ 21 December 1943; Minute 47, OIC ORS to AOT, D/ CinC (through SASO), 23 December 1943 covering Letter, HQ 5 Group to HQBC, ‘Effect of experience on Casualty Rates,’ 21 December 1943; Minute

Notes to pages 173–9 317

47

48 49 50 51 52 53

54

55

56 57 58

59

48, SASO to D/CinC, 27 December 1943; Minute 49, D/CinC to SASO, 27 December 1943; Minute 50, G/C T1 to AOT, 30 December 1943; Minute 51, AOT to D/SASO, 31 December 1943. PRO AIR 14/1847 Minute 52, D/SASO to OIC ORS, 2 January 1944; Minute 59, OIC ORS to D/CinC, SASO, A/C Ops, AOT, 7 February 1944; Minute 60, OIC ORS to D/CinC, SASO, A/C Ops, AOT, 26 February 1944; Minute 62, SASO to AOT, 2 March 1944; Minute 64, ORS 2 to OIC ORS, 16 March 1944; Minute 65, OIC ORS to D/CinC, SASO, A/C Ops, AOT, 20 March 1944. PRO AIR 14/1847 ORS S 187 ‘The Effect of Experience on Operational Efficiency,’ 2 May 1944. PRO AIR 14/1847 Minute 70, OIC ORS to G/C Ops 1 (through AOT), 7 December 1944; Minute 72, G/C Ops 1 to SASO, 13 December 1944. Webster and Frankland, SAO, II, 201–2. PRO AIR 14/3484 ZIP/BMP Report No 500 ‘G.A.F. Fighter Activity on the Western Front for Period Dawn 22.10.43–Dawn 23.10.43,’ n.d. PRO AIR 14/1801 ORS Report 88 ‘Review of Defensive Tactics in Bomber Command Night Operations 3/4th September to 19/20th November 1943.’ PRO AIR 14/1801 Minute 4, OIC ORS to D/CinC, 29 December 1943; Minute 5, D/CinC to SASO, 30 December 1943; Minute 6, SASO to OIC ORS (through D/CinC) 31 December 1943. PRO AIR 14/233 5 Group letter ‘Investigation into New Methods of Deceptive Routing,’ 14 January 1944; Minute 83, Wing Commander Ops 1(e) to SASO (through O.I.C. ORS and A/Cdre Ops), 17 January 1944; Minute 84, OIC ORS to SASO, 20 January 1944; Minute 86, C Nav O to G/C Plans 25 January 1944; Minute 87, G/C Plans to D/CinC, 26 January 1944; Minute 88, D/CinC to G/C Plans, 27 January 1944. PRO AIR 14/233 Minute 91 OIC ORS to D/CinC, 31 January 1944, covering ‘ORS Report B 194, Note on Tactical Suggestion from 5 Group – Force to Proceed past the Target and Then Return,’ n.d.; Minute 92, D/CinC to CinC, 20 February 1944; Minute 93, CinC to D/CinC, 22 February 1944; Letter Bomber Command to 5 Group ‘Deceptive Routing,’ 29 February 1944. Webster and Frankland, SAO, II, 199–207. PRO AIR 14/3948 ORS B 197 ‘Review of Bomber Losses in Relation to Enemy Defensive Tactics in Night Operations,’ 7 February 1944. PRO AIR 14/3948 Minute 8, OIC ORS to D/CinC, 10 February 1944, covering ORS B 197 ‘Review of Bomber Losses in Relation to Enemy Defensive Tactics in Night Operations,’ 7 February 1944. PRO AIR 14/733 Marginal note to Loose Minute Ops 1 (d) to A/Cdre Ops through G/C Plans, 9 February 1944.

318 Notes to pages 179–84 60 PRO AIR 14/733 ‘Minutes of a Conference Held on February 20th 1944, Held at Headquarters, Bomber Command, to Discuss New Tactics to Be Employed to Confuse the Enemy,’ n.d. 61 The Canadian official history argues that Harris was against reducing the size of his raids because this would contradict ‘his determination to deliver the greatest weight of bombs possible to every target.’ The evidence presented here does not refute that argument but does show that there were a number of reasons why, up to this point, there had been a desire, both at the strategic level and for practical reasons at the tactical level, to send out large single target attacks. See Greenhous et al., The Crucible of War, 719–21. 62 PRO AIR 14/733 Loose Minute Ops 1 (d) to A/Cdre Ops through G/C Plans, 7 February 1944; ‘Minutes of a Conference Held on February 20th 1944, Held at Headquarters, Bomber Command, to Discuss New Tactics to be Employed to Confuse the Enemy,’ 21 February 1944. See also PRO AIR14/1453. 63 PRO AIR 14/1801 Minute 10, D/CinC to CinC, 24 February 1944. 64 PRO AIR 14/1801 Minute 11, CinC to D/CinC, 25 February 1944. Harris did not forget the requirement to increase the size of the PFF and in September asked the scientists to revisit the question. They found that the training model for the PFF was highly inefficient; a new introductory course was proposed. Dickins, OR in Bomber Command, 62. 65 PRO AIR 14/1801 Letter AOC 5 Group to A/V/M Walmsley HQ BC, 4 March 1944; Letter 100 Group Addison to A/V/M Walmsley HQ BC, 6 March 1944; Letter AOC 3 Group to A/V/M Walmsley HQ BC, 7 March 1944; Letter AOC 4 Group to A/V/M Walmsley HQ BC, 8 March 1944; Letter AOC 8 Group to A/V/M Walmsley HQ BC, 9 March 1944. 66 PRO AIR 14/1801 Letter SASO to Groups, 22 March 1944. 67 Webster and Frankland, SAO, II, 207–11. See also Greenhous et al., The Crucible of War, 766. 68 Webster and Frankland, SAO, II, 195–9. See also Greenhous et al., The Crucible of War, 796–7. 11. Debates Both Political and Technical: April 1944–May 1945 1 Webster and Frankland, SAO, III, 6. See also Greenhous et al., The Crucible of War, 790–5. The Canadian history also adds to the list indicating that the Air Ministry decided in June that the Luftwaffe itself needed to be knocked out. See Greenhous, The Crucible of War, 824–5. 2 Webster and Frankland, SAO, III, 6–7; Greenhous et al., The Crucible of War, 790–5.

Notes to pages 184–94 319 3 Webster and Frankland, SAO, III, 124. 4 Webster and Frankland, SAO, III, 124–31. 5 PRO AIR 14/734 Loose Memorandum, CinC to D/CinC, 31 December 1943; Loose Minute, D/CinC to CinC, 4 January 1944. 6 PRO AIR 14/734 Memorandum by AOCinC Bomber Command on Paper entitled: ‘OVERLORD’ – Employment of Bomber Forces in relation to the Outline Plan. (Reference AEAF/MS.22007/Air Ops. Dated 12th February, 1944), 14 February 1944. 7 PRO AIR 14/734 Appendix C to Memorandum by AOCinC Bomber Command on Paper entitled: ‘OVERLORD’ – Employment of Bomber Forces in relation to the Outline Plan. (Reference AEAF/MS.22007/Air Ops. Dated 12th February, 1944), 14 February 1944. 8 See also Messenger, Bomber Harris, 163. 9 Harris later wrote that the plan was well founded, the product of a brilliant intellect, that of A/C/M Sir Arthur Tedder. Harris, Bomber Offensive, 197. 10 Dickins, OR in Bomber Command, 125–7. 11 Harris, Bomber Offensive, 200–1. See also Saward, ‘Bomber’ Harris, 251. Saward says that with expert advice from the scientists and Radar staff there was never any doubt in Harris’s mind that precision targets could be bombed effectively. 12 Webster and Frankland, SAO, III, 151–3. 13 Zuckerman, From Apes to Warlords, 234. See also Kirby, Operational Research in War and Peace, 163. Kirby both sides with Zuckerman that Dickins lacked a strong personality and also at the same time acknowledges that it was not Dickins’s place to debate policy. 14 Zuckerman, From Apes to Warlords, 242–3. 15 Zuckerman, From Apes to Warlords, 246–7. 16 Zuckerman, From Apes to Warlords, 252–54. 17 Harris, Bomber Offensive, 201–2. 18 PRO AIR 14/3936 Minute 1, OIC ORS to D/CinC, 9 June 1944. 19 PRO AIR 14/3936 Minute 2, D/CinC to CinC, 15 June 1944. 20 RAF Air Historical Branch, Bomber Command ORS S Series Reports, S 159 ‘Summary of 15 Oboe Groundmarking Attacks on Marshalling Yards 6/7th March–10/11th April, 1944,’ 15 June 1944. 21 See also Air Ministry, Operational Research in the RAF, 55–7. The optimization of technology and tactics that Zuckerman took for granted did not come into being much before the close of hostilities in May 1945. 22 Messenger, Bomber Harris, 177. 23 Webster and Frankland, SAO, III, 154–5. See also Middlebrook and Everitt, BCWD, 489–90. 24 Webster and Frankland, SAO, III, 156–8.

320 Notes to pages 195–201 25 Webster and Frankland, SAO, III, 151–62. See also Dickins, OR in Bomber Command, 118–19 and Middlebrook and Everitt, BCWD, 527. 26 PRO AIR 14/733 ‘Minutes of 2nd Meeting of Bomber Command Tactical Planning Committee Held at 1430 Hours on 9th April 1944.’ 27 See also Air Ministry, Operational Research in the RAF, 61–2. 28 Lovell, Echoes of War, 211–16. Lovell was at the meeting as part of the TRE delegation. 29 PRO AIR 14/848 ‘Policy in Regard to Future Use of H.2.S. Minutes of a Meeting Held in the Air Council Room, Whitehall, at 1500 Hours on Saturday, 22nd April, 1944,’ n.d.; Minute 24, D/CinC to CinC, 21 April 1944; Minute 25, SASO to CinC, 29 April 1944, covering ‘Policy in Regard to Future Use of H.2.S. Minutes of a Meeting Held in the Air Council Room, Whitehall, at 1500 Hours on Saturday, 22nd April, 1944,’ n.d. (see also PRO AIR 14/1296); AIR 14/848 CinC, Marginal note to ‘Policy in Regard to Future Use of H.2.S. Minutes of a Meeting Held in the Air Council Room, Whitehall, at 1500 Hours on Saturday, 22nd April, 1944,’ n.d. 30 PRO AIR 14/848 ‘Bomber Command Operation Order No. 187,’ 16 May 1944; Minute 29, D/CinC to SASO, 13 May 1944; Minute 37, D/CinC to CinC, 6 August 1944; Minute 38, CinC to D/CinC, 6 August 1944; Minute 41, A/Cdre Ops to D/CinC (through SASO), 7 August 1944; Minute 45, CinC to D/CinC, 21 August 1944; Minute 46, OIC ORS to D/CinC, 11 September 1944; Letter D/CinC to ACAS; ‘Night Raid on Brunswick,’ 14 September 1944; Minute 47, G/C Ops to SASO, 29 September 1944. 31 PRO AIR 14/1804 Loose Minute OIC ORS to CinC, 1 September 1944, covering Loose Memorandum ‘Recommendations for Increasing the Efficiency of the P.F.F.,’ 25 August 1944. 32 PRO AIR 14/848 Minute 56, D/CinC to A/Cdre Ops, 3 December 1944. 33 PRO AIR 14/848 Minute 57, D/CinC to SASO, 31 December 1944. 34 PRO AIR 14/3956 Minute 39, OIC ORS to ORS 2, n.d.; Minute 40, ORS 2 to OIC Ors, 11 May 1944. 35 PRO AIR 14/1753 Minute 8, ORS 1 to CNavO, 7 April 1944; Minute 9, CNavO to ORS, 11 April 1944; ORS B 206 ‘Analysis of Navigation – Berlin March 24/25th 1944,’ April 1944. 36 PRO AIR 14/1753 Minute 30, OIC ORS to A/Cdre Ops (through CNavO), 7 April 1944. 37 PRO AIR 14/1753 ORS S 139 ‘Analysis of Navigation – Berlin – March 24/25th 1944, n.d. 38 PRO AIR 14/1753 ORS S 169 ‘Performance of Windfinding Aircraft,’ 25 July 1944. 39 PRO AIR 14/3026 Minute 1, ORS 1(b) to OIC ORS, 25 April 1944, covering

Notes to pages 202–7 321

40

41

42

43 44

45 46

ORS S 157 ‘A Suggested “Reference Point” Method of Bombing by Means of H2S Mk. III,’ 19 April 1944; Minute 2, OIC ORS to G/C Radar, 26 April 1944; Minute 3, Unknown to OIC ORS (through R1R and G/C Radar), 6 May 1944; Minute R1R to OIC ORS (through G/C Radar), n.d.; ORS 1 (b) to OIC ORS, 11 May 1944. PRO AIR 14/3026 ORS S 155 ‘H2S Navigation,’ 5 May 1944; Minute 6, ORS 1 (b) to OIC ORS 11 May 1944, covering ORS S 155 ‘H2S Navigation,’ 5 May 1944; Minute 7, OIC ORS to CNavO, 12 May 1944; Minute 8, CNavO to OIC ORS 13 May 1944. PRO AIR 14/3026 ORS S 189 ‘H2S Blind Bombing Accuracy – Oct. 1st 1943–Apr. 30th, 1944,’ 13 August 1944; Minute 13, G/C Radar to OIC ORS (through CNavO), 21 July 1944; Minute 14, OIC ORS to ORS 1(b), 23 July 1944; Minute 13, G/C Radar to OIC ORS (through CNavO), 21 July 1944; Minute 14, OIC ORS to ORS 1(b), 23 July 1944; Minute 17, OIC ORS to D/ CinC, 14 August 1944, covering ORS S 189 ‘H2S Blind Bombing Accuracy – Oct. 1st 1943–Apr. 30th, 1944,’ 13 August 1944. PRO AIR 14/1753 Letter HQ BC to Groups ‘Operational Navigation – Timekeeping,’ 20 March 1945, covering ORS 122 ‘Analysis of the Hannover Raid 5/6th January and of the Munich Raid 7/8th January 1945,’ n.d. PRO AIR 14/451 ‘Minutes of Navigation Conference Held at Headquarters Bomber Command on Tuesday 30th January 1945,’ n.d. PRO AIR 14/517 Letter HQ BC to Gps ‘Operational Navigation – Timekeeping,’ 20 March 1945, covering ORS Report 122 ‘Analysis of the Hanover Raid 5/6th January and the Munich Raid 7/8th January 1945,’ n.d. PRO AIR 14/517 HQ 5 Gp to 5 Group Bases 53–6 ‘O.R.S. Navigational Report - Munich (7/8th.1.45),’ 6 April 1945. PRO AIR 14/517 HQ 5 Group to HQ BC ‘O.R.S. Navigational Report – Munich (7/8th.1.45),’ 8 April 1945.

12. Round-the-Clock Bombing: April 1944–May 1945 1 2 3 4

Webster and Frankland, SAO, III, 163–72. Greenhous et al., The Crucible of War, 840–2. Webster and Frankland, SAO, III, 166–7. See Dickins, OR in Bomber Command, 83–7 for commentary on the scientists’ contribution to these tactics. 5 PRO AIR 14/1801 Minute 18, ORS 2 to OIC ORS, 17 June 1944. 6 PRO AIR 14/733 Loose Minute OIC ORS to A/Cdre Ops (copy to DCinC) 17 June 1944; Loose Minute 2, D/CinC to CinC, 18 June 1944; Loose Minute 3, CinC to D/CinC, 19 June 1944.

322 Notes to pages 207–13 7 PRO AIR 14/733 Loose Memorandum ‘Bomber Tactics during Very Light Nights,’ 17 June 1944. 8 PRO AIR 14/1453 ‘Minutes of the Fifth Meeting of Bomber Command Tactical Planning Committee Held on 24th June 1944,’ 26 June 1944. 9 Webster and Frankland, SAO, III, 166–7. 10 Webster and Frankland, SAO, III, 168–70. 11 Webster and Frankland, SAO, III, 185–6. 12 Webster and Frankland, SAO, III, 183. 13 Webster and Frankland, SAO, III, 187. See also Richards, Portal, 322. 14 PRO AIR 14/3950 Minute 4, OIC ORS to D/CinC, 29 November 1944, covering ORS Loose Memorandum ‘Estimation of the Effort Required against Oil Targets in Western Germany,’ 29 November 1944. 15 PRO AIR 14/3950 Minute 8, CinC to D/CinC, 11 December 1944; Minute 9, D/CinC to CinC, 12 December 1944. 16 PRO AIR 14/3936 ORS S 197 ‘Bombfall Distribution in Night Area Attacks on German Cities,’19 December 1944. 17 Webster and Frankland, SAO, III, 184–90. 18 PRO AIR 14/3950 Minute 12, OIC ORS to D/CinC (through G/C Plans), 5 January 1945. 19 PRO AIR 14/3950 Minute 18, D/CinC to CinC, 14 January 1945, covering ORS Loose Memorandum ‘Estimation of Effort Required against German Marshalling Yards,’ 12 January 1945. 20 PRO AIR 14/3950 HQBC to Air Ministry DBOps, 18 January 1945, covering ORS B 231 ‘Estimation of Effort Required against German Marshalling Yards,’ 12 January 1945. 21 PRO AIR 14/3950 Letter Air Ministry DBOps to HQ BC, 22 January 1945. 22 PRO AIR 14/3950 Minute 22, OIC ORS Memo to file, 26 January 1945; Minute 23, ORS 1 (d) to OIC ORS (through ORS 1), 1 February 1945. 23 PRO AIR 14/3950 Minute 24, OIC ORS to G/C Plans, 3 February 1945; Minute 25, ORS 1(d) to G/C Plans, 8 February 1945. 24 PRO AIR 14/3950 Minute 26, G/C Plans to D/CinC, 9 February 1945. 25 PRO AIR 14/3936 ORS S 184 ‘Bombing Accuracy in Bomber Command against Lightly Held Targets,’ 20 September 1944. 26 PRO AIR 14/3936 ORS S 197 ‘Bombfall Distribution in Night Area Attacks on German Cities,’ 19 December 1944. 27 PRO AIR 14/1389 Letter HQ BC to Groups ‘Assessment of Bombing Accuracy on Daylight Operations’ 22 September 44, covering ORS Report B224 ‘The Assessment of Bombing Accuracy on Daylight Operations: Notes for the Guidance of Bombing Leaders & Intelligence Officers’ 11 September 1944.

Notes to pages 213–18 323 28 PRO AIR 14/1389 Letter HQ 4 Group to HQ BC ‘Assessment of Individual Bombing Accuracy on Daylight Operations,’ 26 October 1944. 29 PRO AIR 14/1389 Letter HQ 6 (RCAF) Group to HQ BC ‘Assessment of Bombing Accuracy on Daylight Operations,’ 2 November 1944. 30 PRO AIR 14/1389 Letter HQ 1 Group to HQ BC; 28 October 1944; Letter HQ 5 Group to HQ BC, 28 October 1944; Letter HQ 8 Group to HQ BC, 30 October 1944. 31 Webster and Frankland, SAO, III, 172. 32 PRO AIR 14/1771 ORS Memorandum ‘Interim Statement on the Operational Accuracy of G-H and the Concentration of Bombing Achieved on G-H Raids,’ 10 December 1944. 33 PRO AIR 14/1771 Letter 3 Group to BC ORS (Dr Peters), 25 February 1945. 34 Webster and Frankland, SAO, III, 136–9; Greenhous et al., The Crucible of War, 822. 35 Webster and Frankland, SAO, III, 141–51. 36 PRO AIR 14/638 ORS S 133 ‘The Operational Record of FISHPOND and MONICA III (Visual MONICA) October, 1943–February 1944,’ March 1944. 37 PRO AIR 14/1300 Loose Minute OIC ORS to D/CinC, 19 July 1944. 38 PRO AIR 14/1300 Loose Minute CSO to D/CinC, SASO, A/Cdre Ops, OIC ORS, ‘Radar Silence Approaching the Enemy Coast,’ 22 July 1944. See also Lovell, Echoes of War, 233–7. Lovell contends that the CSO, G/C DaltonMorris, was a ‘well known antagonist of H2S,’ but his thoughts on making the best of these circumstances are not obviously indicative of this. 39 PRO AIR14/1823 ‘Appreciation of the Effect of the Use Made by the Enemy of Transmissions from Bombers,’ 1 August 1944. 40 PRO Air/14 638 ‘Appreciation of the Effect of the Use Made by the Enemy of Transmissions from Our Aircraft,’ 7 September 1944. ORS ‘S 175 Trials of Flensburg and S.N. 2 against a Bomber Stream,’ 7 September 1944. 41 PRO AIR 14/638 Minute 36, CSO to DCinC (through SASO), 7 September 1944; Minute 37, SASO to DCinC, 9 September 1944; Minute 38, DCinC to CinC, 11 September 1944; Marginal note, CinC to D/CinC, 11 September 1944, to Minute 38; Minute 39, D/CinC to SASO, 12 September 1944. 42 PRO AIR 14/632 Letter Air Ministry to HQ BC, 12 September 1944. 43 PRO AIR 14/632 Minute 1, Ops 1.e to ORS 2, 15 September 1944; Minute 2, ORS 2 to Ops. 1 (e) 11 October 1944; ‘The Evidence for the Use of Fighters Equipped with Upward Firing Guns against Night Bombers,’ 1 October 1944. 44 PRO AIR 14/632 Minute 3, Ops. 1(e) to SASO, 12 October 1944. 45 PRO AIR 14/632 Minute 5, DCinC to SASO, 13 October 1944; Letter HQ

324 Notes to pages 218–23

46 47 48 49 50 51 52 53 54

55

56

57 58

59 60 61 62 63 64 65

BC to Groups, ‘Tactical Note on Upward Firing Guns,’ dated 25 October 1944; Letter HQ BC to Groups, ‘The Use of Fishpond as a Warning Aid to Detect Fighters Approaching from below the Horizontal,’ 27 October, 1944. PRO AIR 14/632 Letter DAT to HQ BC, 12 November 1944. PRO AIR 14/1801 Loose Minute, OIC ORS to D/CinC, 14 November 1944; Letter, Harris to Under-secretary of State, 24 November 1944. Webster and Frankland, SAO, III, 173–5. Webster and Frankland, SAO, III, 176–8. Webster and Frankland, SAO, III, 173–80. Webster and Frankland, SAO, III, 171–2. PRO AIR 14/3932 Loose Memorandum ‘Formation Flying in Daylight,’ 18 August 1944. PRO AIR 14/3932 Minute 20, ORS 2 to OIC ORS, 21 August 1944, covering Loose Memorandum ‘Formation Flying in Daylight,’ 18 August 1944. PRO AIR 14/3932 Minute 21, OIC ORS to A/Cdre Ops, 12 October 1944, covering ‘Losses and Flak Damage – Heavy Bombers on Daylight Raids June–September 1944,’ 11 October 1944. PRO AIR 14/3932 Minute 24, OIC ORS to D/CinC (through Ops 1 (e), G/C Ops and SASO, 28 November 1944, covering ‘Bomber Command A.A. Losses in Daylight Operations,’ 17 November 1944. PRO AIR 14/3932 Minute 25, Ops 1(e) to D/CinC (through G/C Ops 1, A/Cdre Ops and SASO), 3 December 1944; Minute 26, A/Cdre Ops to D/CinC, 6 December 1944. PRO AIR 14/3932 Minute 27, SASO to D/CinC, 8 December 1944. PRO AIR 14/3932 Minute 28, Note of action 13 December 1944. See Carrington, Soldier, 136–7. A/V/M H.S. Walmsley had come to the headquarters in February 1944 when it became apparent that the work requirements of Overlord and the bombing campaign would each require a separate SASO. PRO AIR 14/1803 Minute 36, ORS 2 to OIC ORS, 22 April 1944. PRO AIR 14/1803 Minute 38, OIC ORS to D/CinC, SASO, AOT, 25 August 1944. PRO AIR 14/1018 Letter CinC to USS ‘Bomber Command L.F.S., H.C.U. and O.T.U. Requirements – 1944–1945,’ 9 August 1944. PRO AIR 14/1018 Minute 10, OIC ORS to CinC, 15 September 1944. PRO AIR 14/1018 Letter CinC to USS, ‘Bomber Command L.F.S., H.C.U. and O.T.U. Requirements – 1944–1945,’ 16 September 1944. PRO AIR 14/1880 ORS S188 ‘Note on Ditching Drill and the Incidence of Ditching Consequent on Bomber Command Operations,’ 19 October 1944. PRO AIR 14/1880 Minute 17, ORS 2 to OIC ORS, 21 August 1944; Minute 20, CASR to ORS 2 (through Sigs Plans) 13 September 1944; Minute 22,

Notes to pages 223–34 325

66

67 68 69 70

OIC ORS to D/CinC (through CSO) 20 October 1944, covering ORS S188 ‘Note on Ditching Drill and the Incidence of Ditching Consequent on Bomber Command Operations,’ 19 October 1944; Marginal note to Minute 22, OIC ORS to D/CinC (through CSO) 20 October 1944. PRO AIR 14/3958 ORS M 84 ‘Report on Collisions between Bombers at Night, October 1943–March 1944,’ n.d.; Loose Report ‘Report on Collisions between Bombers on Operations, April 1944–Jan 1945,’ 16 February 1945. PRO AIR 14/3958 Letter Air Ministry DD Science to OIC ORS, 6 October 1944. PRO AIR 14/3958 Letter ORS BC to Air Ministry DD Science, 25 October 1944. PRO AIR 14/1979 Letter HQ BC to Groups ‘Measures to Defeat Interception by Enemy Aircraft,’ 7 December 1944. Webster and Frankland, SAO, III, 197.

13. Conclusions 1 Greenhous et al., The Crucible of War, 528. 2 Harris, Bomber Offensive, 133. 3 Harris, Bomber Offensive, 131.

This page intentionally left blank

Bibliography

Primary Sources Air Historical Branch, Royal Air Force Bomber Command Operational Research Section reports. Churchill Archives A/V/M Sydney Bufton Papers. Directorate of History and Heritage, Department of National Defence PRO AIR 14 files (copies). Public Record Office AIR 14, Public Record Office. War Office 291, Public Record Office. Royal Air Force Museum Harris Papers. Shephard Military Operational Research Archive, Laurier Centre for Military Strategic and Disarmament Studies Dickins, B.G. Operational Research in Bomber Command, manuscript, n.d. Bomber Command Operational Research Section reports. The Times ‘Obituary Professor R.J. Smeed.’ The Times (London), 8 September 1976, 14g. ‘Atomic Weapons Chief.’ The Times (London) 4 June 1965, 16. Reports Signals Branch Headquarters Bomber Command. ‘An Historical Account

328

Bibliography

of the Radio Countermeasures Offensive Conducted by the Royal Air Force Bomber Command during the War in Europe 1939–1945.’ Bomber Command Headquarters, High Wycombe, 1945. British Bombing Survey Unit. The Strategic Air War against Germany 1939–1945 Report of the British Bombing Survey Unit. Edited by Sebastian Cox. London: Frank Cass and Co., 1998. Harris, Air Chief Marshal Sir Arthur T. Despatch on War Operations 23rd February, 1942, to 8th May, 1945. Edited by Sebastian Cox. London: Frank Cass, 1995. Official Histories Air Ministry. The Origins and Development of Operational Research in the Royal Air Force. London: HMSO, 1963. Greenhous, Brereton, et al. The Crucible of War, 1939–1945. The Official History of the Royal Canadian Air Force Volume III. Toronto: University of Toronto Press, 1994. Hinsley, Francis H. British Intelligence in the Second World War: Its Influence Upon Strategy and Operations Volume I. New York: Cambridge University Press, 1979. – British Intelligence in the Second World War: Its Influence upon Strategy and Operations Volume II, London: HMSO, 1981. Webster, Sir Charles and Noble Frankland. The Strategic Air Offensive against Germany 1939–1945. Vol. I, Preparation; vol. II, Endeavour; vol. III, Victory; vol. IV, Annexes and Appendices. London: Her Majesty’s Stationery Office, 1961. Memoirs Carrington, Charles. Soldier at Bomber Command. London: Leo Cooper, 1987. Dyson, Freeman. Disturbing the Universe. New York: Harper and Row, 1979. Harris, Marshal of the R.A.F., Sir Arthur. Bomber Offensive. Toronto: Stoddart, 1947. Slessor, Sir John. The Central Blue. London: Cassell and Co., 1956. Zuckerman, Solly. From Apes to Warlords: The Autobiography (1904–1946) of Solly Zuckerman. London: Hamish Hamilton, 1978. Secondary Sources Books Allen, Wing Commander H.R. The Legacy of Lord Trenchard. London: Cassell, 1972.

Bibliography 329 Andrews, Allen. The Air Marshals: The Air War in Western Europe. New York: William Morrow, 1970. Barker, Ralph. The Thousand Plan. London: Chatto and Windus, 1965. Bashow, David. No Prouder Place: Canadians and the Bomber Command Experience 1939–1945. St Catharines, Ontario: Vanwell, 2005. Bercuson, David, and S.F. Wise, eds. The Valour and the Horror Revisited. Montreal: McGill-Queen’s University Press, 1994. Biddle, Tami Davis. Rhetoric and Reality in Air Warfare: The Evolution of British and American Ideas about Strategic Bombing, 1914–1945. Princeton: Princeton University Press, 2002. Blackett, P.M.S. Studies of War Nuclear and Conventional. New York: Hill and Wang, 1962. Clark, Ronald W. The Rise of the Boffins. London: Phoenix House, 1962. – Tizard. London: Methuen and Co., 1965. Collier, Basil. A History of Air Power. London: Weidenfeld and Nicolson, 1974. Connelly, Mark. Reaching for the Stars: A New History of Bomber Command in World War II. London: I. B. Tauris, 2001. Crowther, J.G., and R. Whiddington. Science at War. London: His Majesty’s Stationery Office, 1947 (reprinted 1948). Cumming, Michael. Beam Bombers: The Secret War of No. 109 Squadron. Thrupp, Gloucestershire: Sutton Publishing, 1998. Davies, Max, and Michel Verhulst, eds. Operational Research in Practice Report of a NATO Conference. London: Pergamon, 1958. Delve, Ken, and Peter Jacobs. The Six-Year Offensive: Bomber Command in World War II. London: Arms and Armour Press, 1992. Director of British Scientists 1966–67. London: Ernest Benn, 1967. Edgerton, David. Warfare State: Britain 1920–1970. Cambridge: Cambridge University Press, 2006. Falconer, Jonathan. The Bomber Command Handbook, 1939–1945. Stroud: Sutton Publishing, 1998. Frankland, Noble. Bomber Offensive: The Devastation of Europe. New York: Ballantine, 1970. Gooch, John, ed. Airpower: Theory and Practice. London: Frank Cass, 1995. Gray, Peter W., and Sebastian Cox, eds. Airpower and Leadership: Theory and Practice. London: The Stationery Office, 2002. Hall, R. Cargill, ed. Case Studies in Strategic Bombardment. Washington: U.S. Government Printing Office, 1998. Hartcup, Guy. The Effect of Science on the Second World War. London: Macmillan Press, 2000. Hastings, Max. Bomber Command. New York: Dial Press/James Wade, 1979. Howard, John. Bennett and the Pathfinders. London: Arms and Armour, 1996.

330

Bibliography

Kirby, Maurice W. Operational Research in War and Peace: The British Experience from the 1930s to 1970. London: Imperial College Press, 2003. Lewis, Peter. The British Bomber since 1914. London: Putnam, 1967. Lindsey, George R., ed. No Day Long Enough: Canadian Science in World War II. Toronto: Canadian Institute of Strategic Studies, 1997. Longmate, Norman. The Bombers: The RAF Offensive against Germany, 1939–1945. London: Hutchinson, 1983. Lovell, Sir Bernard. Echoes of War: The Story of H2S Radar. Bristol: Adam Hilger, 1991. MacIsaac, David. Strategic Bombing in World War Two: The Story of the United States Strategic Bombing Survey. New York: Garland Publishing, 1976. Maitland, Squadron Leader Andrew. Through the Bombsight. London: William Kimber and Co., 1986. Maynard, John. Bennett and the Pathfinders. London: Arms and Armour, 1998. Messenger, Charles. Bomber Harris and the Strategic Bombing Offensive, 1939–1945. London: Arms and Armour Press, 1984. Middlebrook, Martin. The Battle of Hamburg: Allied Bomber Forces against a German City in 1943. New York: Scribner 1981. – The Berlin Raids: RAF Bomber Command, Winter 1943–44. London: Lane, 1982. – The Peenemunde Raid, the Night of 17–18 August 1943. Harmondsworth: Penguin, 1988. Middlebrook, Martin, and Chris Everitt. The Bomber Command War Diaries: An Operational Reference Book, 1939–1945. New York: Viking, 1985. Moore, P.G. Basic Operational Research. London: Sir Isaac Pittman and Sons, 1968. Musgrove, Gordon. Pathfinder Force: A History of 8 Group. London: Macdonald and Jane’s Publishers, 1976. Neillands, Robin. The Bomber War: The Allied Air Offensive against Nazi Germany. New York: Overlook Press, 2001. Meilinger, Col. Philip, ed. The Paths of Heaven: The Evolution of Airpower Theory. Maxwell Air Force Base, Alabama: Air University Press, 1997. Overy, Richard. Bomber Command, 1939–1945. London: HarperCollins, 1997. – Why the Allies Won. London: Pimlico, 1995. Probert, Air Commodore Henry. Bomber Harris His Life and Times, The Biography of Marshal of the Royal Air Force Sir Arthur Harris, the Wartime Chief of Bomber Command. Toronto: Stoddart, 2001. Richards, Denis. The Hardest Victory: RAF Bomber Command in the Second World War. London: W.W. Norton, 1995. – Portal of Hungerford. London: Heinemann, 1977. Rivett, Patrick. Concepts of Operational Research. London: C.A. Watts and Co., 1968.

Bibliography 331 Saundby, Sir Robert. Air Bombardment: The Story of Its Development. London: Chatto and Windus, 1961. Saward, Dudley. ‘Bomber’ Harris The Story of Marshal of the Royal Air Force Sir Arthur Harris. London: Cassell, 1984. Smith, Malcolm. British Air Strategy between the Wars. Oxford: Clarendon, 1984. Stradling, Gp Capt. A.H. The Brass Hat: Being Hints on How to Make the Job Easier. Aldershot, UK: Gale and Polden, 1951. Tedder, Marshal of the Royal Air Force The Lord. Air Power in War: The Lees Knowles Lectures. London: Hodder and Stoughton, 1948. Terraine, John. The Right of the Line: The Royal Air Force in the European War 1939–1945. London: Hodder and Stoughton, 1985. Thiesmeyer, Lincoln R., and John E. Burchard. Combat Scientists [Unnumbered Volume Within the Office of Scientific Research and Development ‘Science in World War II’ Series]. Boston: Little Brown and Co., 1947. Wagner, Harvey M. Principles of Operations Research - With Applications to Managerial Decisions. Englewood Cliffs, New Jersey: Prentice Hall, 1969. Wells, Mark K. Courage and Air Warfare: The Allied Aircrew Experience in the Second World War. Essex: Frank Cass, 1995. Articles and Chapters Cox, Seb. ‘Harris and the Air Ministry.’ In Airpower and Leadership: Theory and Practice, edited by Peter W. Gray and Sebastian Cox. London: The Stationery Office, 2002. Group Captain Peter W., RAF. ‘Review Essay: Bomber Harris His Life and Times.’ In The Royal Air Force Air Power Review, Vol 4. London: UK Ministry of Defence 2001, 1–7. Jacobs, W.A. ‘The British Strategic Air Offensive against Germany in World War II.’ In Case Studies in Strategic Bombardment, edited by R. Cargill Hall, 91–182. Washington, DC, U.S. Government Printing Office, 1998. McCloskey. ‘British Operational Research in World War II.’ Operations Research 35, no. 3 (May–June 1987): 453–69. Meilinger, Col. Phillip S. ‘Trenchard and “Morale Bombing”: The Evolution of Royal Air Force Doctrine before World War II.’ Journal of Military History 60, no. 2 (April 1996), 243–70. – ‘Trenchard, Slessor, and Royal Air Force Doctrine before World War II.’ In The Paths of Heaven, edited by Col. Phillip S. Meilinger, Maxwell Air Force Base, Alabama: Air University Press, 1997, 41–78. Robertson, Scot. ‘In the Shadow of Death by Moonlight.’ In The Valour and the

332

Bibliography

Horror Revisited, edited by David Bercuson and S.F. Wise. Montreal: McGill Queen’s University Press, 1994, 153–79. Solandt, O.M., Dr. ‘Operations Research Address By Dr. O.M. Solandt at the Closing Exercises of C.N.R. Staff Training Course, Lennoxville, Que., August 3rd, 1956.’ Lennoxville, Quebec, 3 August 1956. Zuckerman, Solly. ‘The Need for Operational Research.’ In Operational Research in Practice: Report of a NATO Conference, edited by Max Davies and Michel Verhulst, 6–16. London: Pergamon, 1958. Websites ‘Air of Authority – A History of RAF Organisation.’ http://www.rafweb.org/ Index.htm, accessed 20 May 2006. The Nobel Prize in Physics 1948 http://nobelprize.org/physics/laureates/ 1948/index.html, accessed 15 May 2006 ‘Royal Air Force Bomber Command 60th Anniversary.’ http://www.raf.mod.uk/ bombercommand/group.html, accessed 20 May 2006

Index

The following abbreviations have been used throughout the index: OR for operational research and ORS for Operational Research Section (ORS). accidents, flying, 84–5, 172, 302n21 Aeroplane & Armament Experimental Establishment (A & AEE), 87 Addison, A/V/M Edward B., 180, 255 Admiralty, 22, 31, 33 AGLT (automatic gun laying turret), 216, 239 AI (airborne) radar, 145, 154–6, 184, 239 aiming point (AP), 5, 12, 19, 22, 44, 50, 52–5, 59, 63, 71, 73–7, 99, 103– 4, 116, 118, 120–5, 134, 140–1, 143, 160, 162–4, 187, 193–4, 197–8, 203, 212–13, 227, 240, 246 Air Defence of Great Britain, 17 air defences, British, 14 air defences, German: effectiveness declines, 109, 214; countermeasures, 137; efficiency increases, 80–1, 92, 97–8, 102, 153–4, 176–82, 217–19; ORS study of countermeasures, 98, 109, 127–8. See also flak; ground-controlled interception; night fighters, German

Air Fighting Development Unit (AFDU), 99 air gunners, 14, 135, 150, 153, 172 Air Historical Branch, 10 Air Member for Supply and Research, 28 Air Ministry: bombing strategy, 20–1, 205; committees, 34, and Gee, 61; laboratory at Imperial College, 29; mentioned, 15, 36, 41, 65–6, 220; and navigation problems, 16; role, 19; and operational research, 32, 35, 44, 47, 49, 73, 75, 79, 82–6, 89, 91–2, 97–8, 100, 105, 133, 149, 171, 174, 182, 187, 194, 196–8, 210, 218, 222–4; Origins of Operational Research in the Royal Air Force, 9, 26; organization, 17–19; 249, 253 Air Ministry directives: to Bomber Command, 7, 19, 69, 257–9; to Commands, 19 Air Ministry Psychology Unit, 163 Air Officer Administration (AOA), 18

334

Index

air raids. See individual cities and targets Air Sea Rescue Service, 222–3 Air Staff, 15–16, 18, 76, 187, 190 Airborne Cigar, 155, 174, 215, 239–40 airborne interception radar. See AI Aircraft Production, Ministry of, 27, 45–6, 113, 127, 148, 162, 170 aircraft type. See Halifax bomber; Hampden bomber; Lancaster bomber; Manchester bomber; Mosquito bomber; Mustang fighter; Stirling bomber; Wellington bomber; Whitley bomber aircrew: and air discipline, 224; experience and loss rate, 38, 88–9, 134, 160, 172–4; 282–3; morale, 73; number of tours, 221–2; surviving ditching, 222–3; training, 76, 293n3. See also bomb aimers; navigators; pilots Anti-Aircraft Command, 32, 44 anti-aircraft artillery. See flak anti-submarine warfare, 30–2 area bombing: British policy, 5–7, 22, 229; morality of, 5–6, 9; Freeman’s views, 38, 66; Harris’s views, 66; and Transportation Plan, 37, 183; in attacks on Germany, 213 Army Operational Research Group, 90 ASV (air-to-surface) radar, 30, 49 Augsburg, 177 backers up, 119, 140,164, 185–6, 240 Baker, A/V/M John W., 83, 253 Balfour of Inchrye, Lord, 253 Baldwin, A/V /M J.E.A., 63, 253, 255, 297 Baldwin, Stanley, 226

Battle of Berlin: assessment of, 121, 154, 157, 165–6, 182; loss rate, 146, 157, 174, 177, 181, 230, 297n57; German fighter defences, 146, 154; groundmarking, 102; H2S, 120, 141, 163, 200–2, 253; Harris’s views on, 158, 170, 230; mentioned, 101– 2, 116, 156; outside range of Gee and Oboe, 102; Window, 207 Battle of Biscay, 31 Battle of Britain, 29–30 Battle of Hamburg: described, 139– 40, 144–5; Harris’s view, 158; loss rate, 139–40, 152, 157; mentioned, 101–3, 137, 146, 185; use of H2S, 140; use of TIs, 140; and Window, 137, 144 Battle of the Ruhr: described, 116–17; bombing effectiveness, 128; casualty rate, 128; and Gee, 242, mentioned, 93, 101–2, 137, 139; losses to night fighters, 110, 112, 128–9, 156; loss rates, 116, 139, 145, 157; and Oboe, 116; and Window, 207 Beards, H.L., 269 Benito, 175, 240 Bennett, A/V/M Donald: commands Pathfinder Forces, 76, 255; and Harris, 65; ORS recommendations, 106; impact of loss rates on aircrew performance, 170; ORS report on fighter evasion tactics, 180 Berlin, 23, 102, 166. See also Battle of Berlin Billancourt, 69 Blackett, C.M.S.: chief of Coastal Command ORS, 30–2; 38; career, 32; on advisory role of OR scientists, 28; wins Nobel Prize for Physics, 32, 47

Index 335 blind bombing: and bombing accuracy, 103; development of concept, 60–1, trial of, 73, effectiveness of, 103, 198–9; and Gee, 60–1, 71; and H2S, 158, 195, 196, 202; Harris’s opposition to H2S blind bombing, 196; and Oboe, 78, 188, 194 bomb aimers, 37, 294; assessment of, 123, 125, 160–2; day bombing technique, 206; need identified, 71; selection of, 161; training of, 71, 105, 107, 161–3 bomb types: comparison of, 117–18; 4-lb incendiary, 117; 30-lb incendiary, 117; 500-lb, 186; 1000-lb General Purpose, 118; 2000-lb High Capacity, 168; 4000-lb Cookie, 117; 8000-lb High Capacity, 117; 22,000lb Grand Slam, 212 Bomber Command: effectiveness, 20–2, 48, 75, 80, 94, 107, 144, 175, 181,184, 187, 189, 204, 225, 229; established, 17; historiography, 9–11; interwar history, 12–15; mentioned, 3, 8 passim; organization of, 17–19, 250, 253–5; ORS established, 23–4; role of directives, 6 Bomber Command Headquarters: integration of ORS and HQ staff, 228–9, 292n56; mentioned, 49, 68, 81, 97, 102, 114, 132, 154, 160, 172, 174, 182, 190, 196, 225; organization, 8, 39–2, 231–3, 292n50; ORS organization, 18, 38–42, 55, 231–3, 250; scientific officer attached to, 35 bombing accuracy: and bombsights, 211–12; recognized as a problem, 15, 20–2, 44; Butt report on 23; G-H, 143, 208–9; Gee, 69–70; Harris’s views, 72, 127; H2S, 119, 164;

increased effectiveness, 184; master bomber concept, 194; methodologies for reporting, 49–50, 54–5, 213; in night operations, 23, 280; Oboe, 119, 192–3, 208–9; ORS role in analysing, 44, 55, 58, 72, 98, 102, 129, 187, 211–13, 227, 229; and Pathfinder Force, 76–9, 198; in attacks on Berlin, 157; in attacks on Germany, 160, 209; in attack on Kassel, 143, 160; in pre-invasion bombing, 186, 207 Bombing Development Unit, 16, 201, 218 bombing policy: British policy, 21–22; See also area bombing, daylight bombing, precision bombing; strategic bombing Bombing Research Unit, 46 bombing techniques. See Shaker, Newhaven, Paramatta, Wanganui bombsights: Mk XIV, 123, 211–12; mounting problems, 106; Stabilized Automatic Bombsight (SABS), 212; tracking problems of, 125 Bonn, 214 Booth, S/L, 162 Boozer radar warning device, 155–6, 240 Boscombe Down, 147 Bottomley, A/M Norman, 81, 196, 253, 255 Bremen, 95 Britton, S.C., 269 Brookes, A/V/M G.E., 146–7, 255 Brunswick, 194, 197 Bufton, A/C Sidney O., 75–6, 210–11, 253, 299, 327 Butt, David M., 23, 25, 34, 37, 44, 50, 108, 147, 226, 288n48

336

Index

Butt report, 23, 34, 226, 288n48 cameras, bombsight, 21, 50–1, 197, 213 Capel, A/C A.J., 105, 151 Capon, Mr, 56 carpet jamming device, 240 Carr, A/V/M Sir Roderick, 134, 180, 255 Carrington, Lt-Col. Charles, 39, 67, 291n25 casualties, aircrew, 19, 21, 84–6, 89, 109–13, 128, 130–5, 145–50, 157, 171–3, 175, 181, 190, 208, 209, 218, 222, 298n17 casualties, civilian, 14, 19–20, 190 catseye fighters, 82, 94, 98,110–11, 123; countermeasures, 154, 175–9; and daylight bombing, 207; description, 240, 242 Central Interpretation Unit, 51, 117 Cherwell, Lord, 22, 28, 58 Chief Armament Officer (CArmO), 18, 45,167 Chief Engineering Officer (CEngO): mentioned, 254; role of, 18; and ORS studies, 85, 87, 90–1, 132, 148, 151 Chief Intelligence Officer (CIO): mentioned, 254; transfer of analysis of loss rates to ORS, 83; role of, 18 Chief Navigation Officer (CNav): navigation failures and bombing accuracy, 107, 119, 124, 158–9, 202–3; value of evasive manoeuvres, 176 Chief Signals Officer (CSO): mentioned, 254; and ORS reports, 112, 119–20, 217; role of, 18 Chiefs of Staff Committee, 6, 22, 137

Churchill, Sir Winston, 21–3, 226 Clausewitz, Karl von, 6 Coastal Command: established, 17; and OR, 24, 30–3 Cochrane, A/V/M Sir Ralph, 65, 172, 180, 255 Coggin, K.M.M., 35 Cologne: attacks on, 71, 81, 279; mentioned, 69, 279, 301n11; Operation Millennium (thousand bomber raid), 94–6, 227, 304n54 Combined Chiefs of Staff, 32 Committee for the Scientific Survey of Air Defence, 29, 34 Committee for the Scientific Survey of Air Offence, 15, 34 Committee for the Scientific Survey of Air Warfare, 34 Communication Plan, 210 concentration: concept and assessment, 80–6, 92, 110, 227; decreased effectiveness, 184, 195; and Gee, 136–7; German countermeasures, 97–8, 139, 153–4, 157, 174–8, 184, 207 ; and flak, 135–6, 200; impact of weather on, 108, 159; impact on loss rate, 86, 133–5; mentioned, 52–3, 69, 71, 74, 76; and ORS, 81–2, 92–6, 100, 107, 123, 131, 174–80, 195, 220, 301n11; risk of collision, 224; and Pathfinder Force, 76–7, 104; thousand plane raids, 94–5; views of Harris, 80–1, 95, 123; in attack on Bremen, 95; in attack on Cologne, 93–4; in attack on Essen, 95; in attack on Munich, 194 Coningham, A/V/M Sir Arthur, 20, 22, 255 Constantine, A/C Hugh, 151, 179, 195, 206–7, 220, 254–5

Index 337 corkscrew manoeuvre. See evasive action Corona, 174, 241 Coryton, A/V/M Sir Alec, 72, 110, 255 Cracker I and II Gee trials, 62 Crawford, G/C C.C., 118 creepback, 140, 144, 179, 194, 240 Crossbow, Operation, 241 course reversal. See evasive action Curry, J., 110, 126, 128, 161, 172–3, 254, 269 daylight bombing: by Luftwaffe, 30; Harris’s opposition to, 205; loss rates, 19, 66, 206, 220–1; mentioned, 12, 15, 51–2, 65, 109, 198; resumed 1944, 205; and ORS 45–6, 206–8; Portal’s support for, 21, 195, 205, 215; and pre-invasion bombing, 193; switch to in 1944, 205–10, 212–15, 219–25; by USAAF, 184 Dewdney, D.A.C., 21, 24, 37 Dickins, Sir Basil: biographical sketch, 29, 35–6; and bombing accuracy, 126; on bomb effects, 117–18; on casualties, 150–1, 222; on concentration, 81–83, 94, 136–7; criticism of, 228–9; Dyson’s criticism of, 38; early radar research, 29, and evasive tactics, 123, 175–9; on experience and loss rates, 88, 173; on exhaust dampening, 113; importance of data collection and analysis, 48–9, 55 ; on flak casualties, 112, 220–1; and G-H, 143; and Gee, 59–61, 73, 130; on German night fighter tactics, 127–8,154, 179; on H2S, 120, 122, 196–202; on Halifax losses, 87–8, 134, 147–50; on icing problems, 85–6; on in-flight fires, 90,

133; Lancaster Halifax comparison, 170; on Lancaster losses, 130, 152; on liquid cooled engines and losses, 86; on losses to night fighters, 129; and Manchester bomber loss study, 84; mentioned, 62–3, 105–6, 114, 124, 171; and Monica, 111–12, 156; and navigation issues, 57,105–7, 159; and night photography, 50, 58, 125–6; and No 6 (RCAF) Group losses, 147; and non-combat loss rate, 85; and oil campaign, 209–10; Operational Research in Bomber Command, 9–10; on OR methodology, 53–5; on ORS reports, 168–9; ORS reports to Saundby, 37, 43; on role of ORS in Bomber Command, 43–4, 226; on Pathfinder Force, 120–1; on pre-invasion bombing plan, 185, 187–90, 225; Tactical Planning Committee, 195; target identification solutions, 56–7; on thousand plane raids, 94; on timed runs, 121–2; valued by Harris, 36–7, 227–30; on effect of weather on operations, 58, 108–9; on Window, 145; Zuckerman’s criticism of, 37–8 Dieppe Raid, 187 Director of Bomber Operations. See Bufton Directorate of Air Tactics, 217–18 Directorate of Operational Requirements, 152 Directorate of Scientific Research, 28–9, 34–5, 56–7, 127 Dortmund, 135 Douhet, Guilio, 5 Dowding, A/C/M Sir Hugh, 28–30 Duisburg, 69, 224 Düren, 143

338

Index

Düsseldorf, 69, 160, 312 Dyson, Freeman: critical of Dickins and ORS, 38; critical of Harris, 38, 66–7; criticism of Dickins and ORS refuted, 150, 152; mentioned, 269 Eisenhower, General Dwight D., 183 Elworthy, G/C Samuel C., 82, 94, 106, 254 Essen: ineffectiveness of early attacks on, 69–70; losses due to flak, 145; ORS report on early attacks, 72–4; thousand plane raid, 95; use of Oboe, 116, 140; use of Pathfinder Force, 102; use of Window, 140 evasive action: corkscrew manoeuvre, 97, 153; course reversal, 176; sidestep, 176, utility of, 97, 123, 134–6, 138, 153–4, 175–7 Fighter Command: established, 17; mentioned, 34, 154; and OR, 24, 29–30, 35, 44, 220 fires, in-flight, 90–2, 132–3, 137–8, 181 First World War, 5, 7, 12, 15–16, 22, 32, 38 Fishpond, 215–16, 218, 241 flak, 6, 80–1, 90, 92, 97–8, 112–13, 123, 128–9, 131, 134–6, 138, 144–5, 153–4, 171, 175–6, 179, 181, 200, 206, 215, 219–21, 226 flares, 56–7, 60–2, 69, 70–5, 94, 102–3, 106, 118, 197–8, 212, 239 Flensburg homing device, 88, 216–17, 241 Flying Training Command, 57 Fothergill, J.E., 254, 269 Frankfurt, 82

freelance fighters. See catseye fighters Freeman, A/C/M Sir Wilfred, 170, 263 Freya early warning radar, 215, 241 G-H: capabilities and assessment of, 143–4, 214; description of, 242–3; formation technique, 208, 213, 243; No 3 Group equipped with, 213–14; in attack on Bonn, 214; in attack on Düren, 143 gardening. See minelaying Gee: capabilities and assessment, 59–61, 73–4, 77–8, 227; description, 242; effectiveness of, 69–71, 74–5, 92–3, 104; first use of, 69; German countermeasures, 76; Harris’s views on, 69–71; impact on casualties, 130–1; jamming of, 76; mentioned, 100–2, 136, 208; and thousand plane raids, 82, 94–5, 229; trials of, 61–3; in attack on Billancourt, 69; in attack on Bremen, 95; in attack on Essen, 69–70 Gelsenkirchen, 194 Glasgow, 120 Goggin, J.M.M., 254, 269 Graham, G/C H.R., 107, 254 Grocer radar jamming device, 243 ground-controlled interception (GCI), 92, 97–9, 111, 153–4, 174–5, 184–5, 207, 215, 219, 242 groundmarking: description of, 243; effectiveness of, 119–20, 144, 163, 212–14; first use, 102; and H2S, 120; and Oboe, 120, 122, 144, 188, 208; visual technique for, 193, 214. See also Newhaven, Paramatta Group Captain Operations (G/C Ops), 18–19

Index 339 Group Captain Plans (G/C Plans), 18, 211 Groups: distribution of ORS reports, 58, 69, 74, 84–5, 105, 125–6, 130; organization of, 18; ORS field detachments, 46; No 1, 164–6, 172; No 3, 61–2, 88, 93, 163, 214; No 4, 19–20, 65, 81, 88–9, 93, 135–6, 146– 7, 156, 160, 180, 213; No 5, 18–20, 50, 57, 62, 66, 72, 89, 91, 110, 122, 126–8, 164, 172, 175–7, 193–4, 204, 208; No 6 (RCAF), 101, 146–7, 164–5, 188, 213, 302n21; No 91, 95; No 100, 180, 215 H2S: assessment and capabilities of, 77–8, 102–4, 119–23, 137, 139, 186–7, 195–6; description of, 243; comparison with Oboe, 119; introduction of, 79, 101; use by Pathfinder Force, 140, 197–8; technical issues, 201–3, 225; in attack on Berlin, 120, 158, 200; in attack on Hamburg,103, 119, 140; in attack on Nuremberg, 118; in attack on Wilhelmshaven, 103–4 H2S skymarking. See Wanganui Hagen, 224 Halifax bomber: aircraft performance, 87–8, 134, 138, 147–8, 151, 160, 169, 182; aircrew experience, 86, 88, 134, 146, 160, 173; aircrew training, 134; bomb loads, 118, 149, 168–70; comparison with Lancaster, 114, 148–9, 170, 184; escape from, 150–1, 227; engine type and loss rates, 86–7, 138; flame dampers, 86–7, 113–14, 147; Halifax III, 147– 8, 167, 170; Halifax Mk VI, 151; loss rate, 86, 88, 95, 110, 144–6, 149,

173, 175, 216; mentioned, 65, 265; survivability,150; tail modifications, 169–70; use by No 6 (RCAF) Bomber Group, 146–7; use by Pathfinder Force, 131; in attack on Berlin,157; in attack on Hamburg, 140, 153 Hamburg, 94, 103, 119, 129, 304n54. See also Battle of Hamburg Hampden bomber, 262 Handley Page Aircraft Company, 87–8, 147–8, 170 Handley Page, Sir Frederick, 147 Hannover, 142–3, 160, 312, 321 Harris, Sir Arthur: biographical sketch, 7–8, 64–8; on blind bombing, 73; Bomber Offensive, 9; on Bufton and the Air Ministry, 75–6; and bombing policy, 4–7; Carrington’s view of, 67–8; commander Bomber Command, 34, 64, 99, 180; on concentration, 80–1, 94–6, 227; criticism of, 7, 229–30; on daylight bombing, 205–6; Despatch on War Operations, 9, 68, 80; Dyson’s criticism of, 66–7; on flying accidents, 85; on flying discipline, 92; on Gee and target identification, 70–1, 93, 107; on German fighter tactics, 97, 110, 154, 176–7, 180, 218; on H2S, 123, 158, 196–7, 199, 203; on Halifax loss rate, 87–9; Lancaster Halifax comparison, 170; on loss rates, 88–9, 92–3, 123, 127, 144, 146, 150–1, 170, 181, 221–2, 230; Ludlow-Hewitt’s view of, 66; on Monica, 112, 217; mentioned, 4, 10, 69, 86, 98–9, 101, 104, 106, 108, 116, 121, 126, 135, 137, 181–3, 196; on navigation problems, 92–4, 104, 108, 203; and No 6 (RCAF)

340

Index

Group, 146–7; on Oboe, 74, 79; and oil campaign dispute with Portal, 183, 193, 209–10; initial opposition to Pathfinder Force, 75–6; on Pathfinder Force, 75–7, 93, 106, 120, 123; and photo interpretation, 125; on pre-invasion bombing plan, 137, 183, 185–94, 225; on Saundby, 40–1; on thousand plane raids, 94–5; on use of OTU aircrews, 96; and tours, 222; and values ORS, 63, 76–7, 98, 100, 137, 169, 182, 230; views on command, 68; views on Dickins, 36–7, 227–30; views on Ludlow-Hewitt, 14; views on preinvasion bombing plan, 37, 183, 185–94, 225 Harrison, A/C Richard, 83, 94, 180, 254–5 Heavy Conversion Units, 95–6 High Wycombe. See Bomber Command Headquarters Home Security, Ministry, 179 Home Warfare, Ministry of, 52, 294n1 Hopkins, J.W., 254, 269 Huls, 55 Imperial College, 15, 29, 35–6, 310n45 incendiary bombs: description of, 244; effectiveness of, 117–18, 70–1; mentioned, 51–2, 91, 133, 185, 247; small bomb containers and bomb loads, 167–8; and target identification, 61–2; in attack on Bremen, 95 identification friend or foe (IFF) systems, 30, 49 intelligence: reliability of, 49, 56; transfer of loss rate analysis to ORS, 83, 293n3; use of data by ORS, 44, 57

Isle of Man, 61 Jukes, J.A., 60–1, 161, 254, 269 Kassel, 50, 95, 97, 141–3, 160, 174, 277–8 Krupp munitions works, 70, 73 Larnder, Harold, 29–31, 46 Lammerton, L.F., 155, 254, 269 Lancaster bomber: aircraft performance, 88, 110, 114, 146, 149; aircrew casualties, 150; aircrew experience, 173; bomb loads, 118, 149, 164–8; comparison with Halifax, 114, 148–9, 170, 184; escape from, 38, 150–1, 227; engine type and loss rates, 133; flame dampers, 113; fuel tanks, 152; loss rate, 89, 110, 130, 144–6, 149, 170, 175, 216; mentioned, 65, 102, 150, 276; nose optimization, 106–7; superiority of 101, 148–9, 170, 184; survivability, 150–1; use by Pathfinder Force, 106; in attack on Berlin, 157, 171; in attack on Hamburg, 140; in attack on Ruhr, 156 Laurier Centre for Military Strategic and Disarmament Studies, 4 leaflet dropping raids, 14, 96 Leipzig, 177 Lille, 214 Lindemann, Frederick. See Cherwell, Lord Lloyd, F.J., 160–1, 254, 269 London: German bombing in First World War, 12, 64; TI trials, 119–20 loss rates: Bomber Command, 19, 34–5, 38, 88, 98, 109, 128–9, 131–2, 146–7, 181, 214, 144, 152–3, 205,

Index 341 215, 219, 221–3; in attack on Augsburg, 177; in attacks on Berlin, 141–2, 146, 157, 177, 181; in attacks on Hamburg, 139–40, 145, 157; in attack Leipzig, 177; in attack on Mailley-le-Camp, 214; in attack on Nuremburg, 181, in attacks on the Ruhr, 116, 139, 145, 157; in attack on Schweinfurt, 177; in attack on Stuttgart, 177 Lovell, Sir E.A., 196, 269 Lübeck, 71 Ludlow-Hewitt, A/M Sir Edgar: assessment of Bomber Command deficiencies, 13–14, 17, 65; commander Bomber Command, 13, 254; views on Harris, 66 Luftwaffe, 13, 97, 139, 146, 153–4, 174–7, 181, 183, 214, 216, 226, 230 Madeburg, 159 Mailly-le-Camp, 214 Manchester bomber, 65, 83–4, 86 Mandrel electronic jamming device, 215, 219, 241, 244 Mannheim, 21, 97, 165 marker bombs, 52, 71, 74, 77–8. See also Target Indicator (TI) bombs. master bomber: concept proposed 121–2; role of 188, 193–4, 211–12, 239–40, 244–5 Maud Project, 36 mean point of impact (MPI), 76, 124–5, 161–3, 211, 245 minelaying, 89, 96, 108–9, 201, 242 Mitchell, General Billy, 5 Monica radar device: capabilities and assessment, 111–14, 156, 215–18, 235; countermeasures, 241; mentioned, 99

morale: of aircrew, 73–4, of German civilians, 5, 7, 19, 22, 69, 118; in No 4 Group, 175; in No 6 (RCAF) Group, 146 Mosquito bomber: bomb load of, 149; comparison with other bombers, 149; and groundmarking, 193; loss rates, 116, 149; mentioned, 102, 262; and skymarking, 102; use in feints and demonstration attacks, 116, 178; use of Oboe, 194, 246; in attack on Berlin, 157 Munich, 120, 129, 194, 202, 204 Münster, 258 Mustang fighter, 214 National Physical Laboratory: Radio Department, 29 navigation: and Harris’s views, 65, 68, 70, 92–3; importance of, 5, 14–17; mentioned, 10, 60, 179, 241; and night bombing, 21; problems, 20–2, 104–8, 123, 140; ORS, 55, 57–9, 61, 92–3, 128, 137, 141–3, 182, 201–3, 226–7 navigation aids: mentioned, 18, 195; and Harris’s views, 65, 68, 70, 92–3; technical challenges, 20, 27; requirement for, 15, 20, 92–3; and ORS, 42, 74, 83, 85; Portal advocates, 22–3; research, 27, 29; See also flares; G-H; Gee; H2S, and Oboe navigators: assessment of, 159, 201–3; problems with navigation, 15, 51, 56–9, 92–3, 104–5, 137, 141–2, 162–3, 200; training of, 15–16, 20, 57–8, 77, 93, 160; use of Gee, 59– 61, 78, 242; use of H2S, 158, 200–2; wind measurement problems, 143, 158–9, 200–4

342

Index

NAXOS, 216, 243, 245 Newall, A/M Sir Cyril, 14, 253 Newhaven, 103, 210, description of, 245, 272–3, 305; Musical Newhaven, 194 night bombing: effectiveness of, 23, 99, 160,186, 195, 295n30, 296n38; loss rates, 183, 129–30, 219, 281; by Luftwaffe, 137; switch to, 21; in First World War, 17, 19, 65; by USAAF, 206 night fighters, British, 30, 154 night fighters, German, 97–8, 127–8, 146, 153–4, 157, 176–7, 181, 183, 214–15. See also catseye fighters Nuremberg, 97, 118, 141, 157 Oboe: capabilities and assessment, 74–5, 78, 120–1, 139, 210, 212; comparison with G-H, 214; controlled Oboe, 194, 212, 240; impact on bombing accuracy, 78, 212, 186; introduction of, 101; range of, 102, 140; use of in daylight bombing, 208; use by Pathfinder Force, 198; in attack on Essen, 74, 116–17; in attack on the Ruhr, 119; in preinvasion bombing raids, 188 Observer Corps, 36 oil, attack on: in pre-war plans 13; Portal-Harris dispute over, 193–4, 208–9; Portal’s support for, 19, 183; in pre-invasion bombing plan, 183, 193–4, 208–9; sorties estimated by ORS, 209, 225 O’Loughlin, Mike, 38 operational research: definitions of, 24–8; and RAF, 8–9, 11, 24, 28–9, 34–5, 46–7; value of, 4, 9, 26–8 Operational Research Centre, 47, 114, 117–18, 153

Operational Research Committee, 47, 190 Operational Research Section, Bomber Command: data collection, and analysis, 10, 48–50, 55; establishment of, 24, 33–5, 226; Dickins appointed as head, 35–6; Harris’s views on, 42–3, 76, 98, 188, 230; integration into Bomber Command Headquarters, 42–3; loss rate analysis transferred to, 83; mentioned, 44, 134, 190; mandate of, 44–5; organization of, 45–6, 254, 287n24, 292n50,n56; personnel, 35, 45, 47, 254, 269; Portal proposes creation of, 9; records of, 4; relations with other Command ORS, 47; reporting relationship of, 37, 43; role of, 29, 45; Slessor’s views on, 62; summation of contribution, 226–34 Operational Research Section, Bomber Command reports: ORS report series, 55; change in format, 168 Operational Research Section, Bomber Commands reports on: Air Sea Rescue Service, 222–3; Airborne Cigar, 155; aircrew experience and loss rates, 38, 134, 172–4; Battle of Berlin, 154, Battle of Hamburg, 144; bomb aimers, 105; body armour, 171–2; bomb loads and bomb efficiency, 117–18, 164–8 ; bombing accuracy, 21–2, 44, 55–6, 211–13; bombing effectiveness against German targets 1944, 212–13; bombing error, 162–3; Boozer, 155–6; casualties, 84–5,150; collisions, 223–4; concentration, 80–3, 136; engaging unknown aircraft, 172; engine dampening and

Index 343 protection, 113–15; evasion tactics, 134–5, 175–82; flak, 112–13, 135–6, 220–1; fires, in-flight, 90–2; Gee, 59–61,104, 130–1; German fighter tactics, 144, 175, 217–19; H2S, 104, 137, 158, 163, 196–203; G-H, 143; Halifax bombing accuracy, 160; Halifax loss rates, 86–7; Halifax Mk III, 147–8; Halifax tail assembly, 169–70; icing, 85–6; Lancaster comparison with Halifax, 149, 170; Lancaster fuel tanks, 152; loss rates, 44; Monica, 111, 156, 215–17; loss rates and daylight bombing, 206–8; loss rates and fighter operations, 109–10, 128; loss rates and incendiary bombs, 133; loss rate and night bombing, 129–30; loss rates and routing, 94; loss rates, 97; Manchester loss rate, 83–4; navigation, 93, 141, 158–9, 200–4; night photography, 72, night vision and bomb aimers’ capabilities, 160–1; No 4 Group aborted sortie rate, 171; No 6 (RCAF) Group loss rate, 146–7; Oboe, 74–5, 78–9; ORS activities, 169; Pathfinder Force, 63, 76–8, 100, 103–4, 120–1, 125, 131–2, 137, 163–4, 300n53; photo interpretation, 72; pre-invasion bombing estimates, 209, 187–90, 210–11; routing, 93–4, 98–9, 175–80; target indicators, 120–5; target identification, 55–6; 69–70, 72; Transportation Plan sortie analysis, 192–3; visibility and target markers, 161; effect of weather on operations, 108–9; Wilhelmshaven, 103–4; Window, 145–6, 154–5 Operational Training Units (OTU):

in thousand plane raids, 94–6; mentioned, 64, 161 Operation Overlord, 37, 181, 194, 197, 199, 206–7; Dickins’s and bombing plan, 185, 187–90, 225; Harris’s view of bombing plan, 7,185–94 Oxland, A/V/M Robert D., 106, 124, 126, 156, 173, 175, 187, 253 Paramatta technique, 198, 244–5, 274 Pathfinder Force: aircrew selection and training, 164, 197–8; assessment of, 53–4, 76–8, 101, 104, 120– 3, 180, 197, 227, 294n18, 300n53; description of, 246; established, 76; groundmarking, 163, 193–4, 198; H2S, 102–4, 118, 120–1, 140, 163–4, 197–8; Harris’s views on, 75–6; Lancaster optimization, 106–7; loss rate, 131, 170; mentioned, 65, 101, 211; and Oboe, 77–9, 120–1, 198, 214; ORS reviews effectiveness of, 77, 103–4, 120–1, 137, 163–4, 300n; proposed, 53, 63, 23, 57, 63, 75, 227; route marking, 141; skymarking, 102–3, 121, 198, 247; target identification, 50; target marking, 102–4, 106, 161–2, 193; in attack on Berlin, 163, 170; in attack on Hamburg, 119, 140; in attack on Hannover, 142–3; in attack on Kassel, 50; in attack on Nuremberg, 118 Peirse, A/M Sir Richard: approves formation of ORS, 35, 42; approves bomb sight cameras, 50; biographical sketch, 253; as commander, Bomber Command, 40; mandate of ORS, 44, 169; opposes centralization of ORS, 47; relationship with

344

Index

Dickins, 56; replaced as commander, Bomber Command, 96, 296n57; study of bombing accuracy and navigation, 55, 58–9; values ORS, 63 Pendred, G/C L.F., 83 Peters, B.G., 254, 269 photographic interpretation, 22–3, 51–3, 55–8, 69, 72, 76, 93, 102, 125–6 photographic reconnaissance: need for, 15–16; review of, 22–3, and night photography, 50–2, 60, 70, 72, 136–7, 276 pilots: as staff officers, 40; navigation skills, 15, 78, 93, 125; pre-war production problems, 15; post-raid reports, 50; survivability and aircraft handling, 87; survivability and experience, 88–9, 165, 169, 173; survivability and flak, 112–13; survivability and Lancaster bombers, 150–1 Plan Position Indicators (PPIs), 103, 163 Portal, M/RAF Sir Charles: biographical sketch, 17; and bombing directives, 19, 253, 257; commander, Bomber Command, 17, 40; on Butt report, 23; mentioned, 66, 253; oil campaign and dispute with Harris, 183, 193, 208–9; proposes creation of Bomber Command, ORS, 23; switch to day bombing, 195, 205, 215; use of OTU aircrew in thousand plane raids, 95; views on strategic bombing, 21–2 Pratt, A.W., 254, 269 precision bombing: and Bomber Command, 12, 20, 184, 204, 229; and improved bombsights, 212; British bombing strategy, 7; and

daylight bombing, 19, 206; and G-H, 143, 213; H2S, 119, 123; Harris’s views, 65; and navigation aids, 79; and night bombing, 14, 65; and Oboe, 101, 119, 141, 208; and Pathfinder Force, 76; Portal’s support for, 21; pre-invasion bombing plan, 183, 187–8, 194, 209, 213 Public Records Office, 10 qualitative analysis, 10, 53, 294n18 quantitative analysis, 10, 54 radar, 18, 26–9, 36; German antiaircraft defences, 98, 128–9, 140; ORS, 45–6, 53, 66, 73, 119–20, 158. See also A.I.; ground-controlled interception; H2S; Monica Radar Research Section. See Stanmore Research Section railways, 13, 185–6, 189, 192, 210 random error, 163, 212, 246 Rankine, A.O., 49 Remscheid, 141 Renault Works. See Billancourt Rhineland, 13 Richards, W.D., 254, 269 Roach, A/C, 91, 132, 148, 254 Roberts, G.A., 29–30, 106, 159–60, 165–8, 200, 254, 269 Rostock, 71 route markers, 141–3, 177–8, 180 routing, 60, 80, 93–4, 98–9, 195 Rowe, A.P., 27, 29, 36, 75, 196 Royal Air Force (RAF), 6–10, 12–13 passim; headquarters organization, 251; Royal Air Force War Manual AP 1300, 6; Strategic Air Offensive against Germany, 9. See also Air Defence of Great Britain; Bomber Command;

Index 345 Coastal Command; Fighter Command; Flying Training Command; Groups; Squadrons Royal Aircraft Establishment, 36, 57,132, 152 Royal Canadian Air Force, 9 Royal Flying Corps, 64, 232 Ruhr, 23, 69, 70–1, 78–9, 95. See also Battle of the Ruhr running commentary, 122, 140, 145, 153–4,174–5, 241, 247 Saar, 13 Saarbrucken, 97 Salmond, A/M Sir John, 65 Samson, 247 Saundby, A/M Sir Robert: biographical sketch, 40–1; on bomb effects report, 117–18; Deputy CinC Bomber Command, 40; and concentration, 81–3; and Dickins, 36–7, 42–3, 61, 63; and use of Fishpond, 218; and flak casualties report, 112–13; and use of flares, 61; and Gee trials, 60–3, 73; on German night fighter tactics, 180; on H2S, 120, 122, 196, 198, 203, 214; on Halifax loss rates, 89; and Harris, 45, 65, 67; Harris’s view of, 41; and Manchester and Whitley bomber study, 83–6, 89; mentioned, 85, 86, 91, 105–6, 108, 113, 124, 126, 129–30, 143, 145, 147, 149–50, 154, 171, 173, 176; and Oboe, 74, 78–9; opposes centralization of ORS, 47; and ORS mandate, 45; ORS reports to, 37, 43–4; support of ORS, 47, 56, 58–9; on Pathfinder Force, 75, 78, 104, 120; and pre-invasion bombing plan, 185, 192; requests routing study, 93–4;

on thousand plane raid, 94; on visual target recognition, 106 Saward, G/C Dudley, 119, 196–7, 201, 254, 299n35 Schonland, Brigadier, 90 Schweinfurt, 177, 194 Scott, K.A., 254 searchlights, 20, 49, 110, 140, 144–5, 153, 175 Secretary of State for Air, 28 Senior Administration Officer (SAO), 40, 232 Senior Air Staff Officer (SASO): role of, 18, 40–1, 232; and OR, 46 Senior Intelligence Officer (SIO): role of, 18 Serrate, 247 Shaker technique: description of, 247, 271–2, 298n17; effectiveness of, 69, 71, 73–5, 99; trial of, 62–3; in attack on Billancourt, 69; in attack on Cologne, 94–5; in attack on Essen, 95 sidestep. See evasive action Sinclair, Sir Archibald, 253 skymarking: assessment of 120–1, 144, 195, 198; comparison with G-H, 144; comparison with groundmarking, 119; description of, 102–3, 244–5, 247–8; in attack on Wilhelmshaven, 103 Slessor, A/V/M Sir John: commander, Coastal Command, 31; commander, No 5 Group, 62, 255; on prewar bombing policy, 16; views on operational research, 31–2, 42, 62, 70, 126 Small Bomb Container (SBC): description, 247; effectiveness, 117– 19; shortage of, 167–8

346

Index

Smeed, R.J: biographical sketch, 301n45; mentioned, 179, 254, 269; member Tactical Planning Committee, 179, 195; ORS studies,130, 132, 152–3,155, 169, 199–200, 206, 217, 220–1 Smith, W/C, 169 SN2 radar, 217 Sorley, A/M R.S., 148 Sortie Raid reports, 49–50, 53, 294n18 Squadrons: No 3, 17; No 9, 128; No 16, 17; No 31, 65; No 39, 64; No 44, 64; No 45, 65; No 58, 65; No 191, 64, No 210, 65; No 424 (RCAF), 265; No 434 (RCAF), 153; No 617, 193, 195, 212 Stabilized Automatic Bombsight (SABS). See bombsights Stanmore Research Section, 30 Stettin, 56 Stevens, G.W.H., 106, 269 Stirling bomber: bomb loads, 118, 166; engines, 84, 133; fuel cells, 84; handling characteristics, 88; loss rate, 86, 131, 145–6; mentioned, 65, 89, 167, 170, 264; in attack on Berlin, 146, 157; in attack on Hamburg, 140, 145 Stuttgart, 120, 177 strategic bombing: origins of concept, 5–6, 8–9; Harris’s views, 37, 40, 99; mentioned, 225–6, 233; Portal’s views, 22; technical challenges, 15, 21–2; switch from strategic to tactical bombing, 181, 189 Supreme Headquarters Allied Expeditionary Force (SHAEF), 187, 189 systematic error, 124, 163, 200, 209, 212, 247 Tait, W/C, 114

Tactical Planning Committee, 179–80, 195, 206–7, 219 target identification and recognition: difficulty in night operations, 20; problem of, 5, 15, 23; need for target identification squadrons, 22–3, 63, 75–6; weather as a factor, 58–9, 69–70; and ORS, 53–7, 63, 69–71, 105, 236–7 Target Indicator bomb (TI): description, 248; effectiveness of, 103; introduced, 52, 71, 76–8, 101; requirement for, 71; and ORS, 78, 99, 160–4, 119–25, 211–12; use of by Pathfinder Force, 102–3, 140, 160–4 target marking. See groundmarking; flares, marker bombs; Newhaven; Oboe; Paramatta; Pathfinder Force; Target Indicator bombs; Shaker; skymarking; visual marking; Wanganui Tedder, A/C/M Sir Arthur, 183, 189–90 Telecommunications Research Establishment (TRE), 27, 35, 61, 196 Thomson, Sir George, 220–1 thousand plane raids: Cologne selected as first target, 94; mentioned, 72, 80; proposed, 93–4, 227; ORS review, 95–6; attack on Bremen, 95; attack on Cologne, 94–5; attack on Essen, 95 timed run, 74, 121–3, 208 Tinsel jamming device, 174, 248 Tizard, Sir Henry, 15–16, 29, 34–6, 46, 56, 75 tours, number of, 222 TR 1335. See Gee Transportation Plan: Bomber Command’s role, 183, 189–90; Harris’s opposition to, 37, 190, ORS analysis of first raids, 192–3

Index 347 Trenchard, M/RAF Sir Hugh, 5, 12–14, 17 United States Army Air Forces (USAAF), 108, 184, 206, 208 United States Army Air Forces (USAAF) Eighth Air Force, 211 V weapons, 205, 220, 241 visual marking, 55, 69, 94, 103–6, 118, 122, 140–1, 163–4, 193, 207–8, 211, 214–15, 239, 240, 300n5 Walmsley, A/V/M Hugh, 173, 180, 221, 224, 253, 324n58 Wanganui: description, 244–5, 248, 275 Wellington bomber: description, 263, engine problems, 84, 86, 320n16; loss rate, 97; in attack on Hamburg, 140; in attack on Mannheim, 97 weather and influence on flying operations: and aircrew experience and training, 16, 96; Bomber Command ORS priority, 47; and bombing effectiveness, 20, 23, 58–9, 72, 108–9; effect of winds on navigation, 97, 107–8, 125, 140–2, 158–9, 164–6, 200–4; and flying accidents, 85, 109; and Gee, 60–1, 70, 104; and German fighter operations, 97; and G-H, 208, 214; and H2S, 77, 104, 120; icing, 85–6, 97,165,171, 302n23; mentioned, 19, 68, 74–5, 82, 108, 158, 165, 209, 231; and navigation challenges, 105, 107; and navigation aids requirement, 51–2, 103; and Oboe, 117, 186, 198, 208; ORS studies, 56, 58, 105; and Pathfinder Force, 75, 103, 105, 120,

137, 170, 193–4, 198; and Shaker, 73, 75, 94; and Transportation Plan, 190–2, 211; in attack on Berlin, 163, 175, 200; in attack on Bonn, 214; in attack on Billancourt, 69; in attack on Cologne, 94–6; in attacks on Essen, 70, 73; in attacks on Hamburg, 140; in attack on Hannover, 142–3; in attack on Kassel, 142–3; in attack on Nuremburg, 181; in attacks on the Ruhr, 117 Whitley bomber, 85–6, 96–7, 261 Wilhelmshaven, 103–4 Williams, E.C., 27, 38 Williams, A/C T.M., 59, 254 Wimperis, H.E., 28–9 Window: assessment of, 139–41, 144– 5, 154–6, 169, 174, 179, 182, 185, 203, 207, 219; capabilities of, 98, 102, 112, 137; description of, 247–8; German countermeasures to, 153– 4, 175, 179, 207; mentioned, 101; in attack on Berlin, 207; in attack on Essen, 140; in attack on Hamburg, 137, 139–40 Wing Commander Navigation (W/C Nav): role of, 18; report on Gee, 60 Wing Commander Photos (W/C Photos): role of, 18 Wing Commander Radio Direction Finding (W/C RDF): role of, 18 Woodward-Nutt, A.E., 36 World War, First, 5, 7, 12, 15–16, 22, 32, 38 Wurzburg. See ground-controlled interception Zeppelins, 12 Zuckerman, Sir Solly, 189–93, 207, 213; views on Dickins, 37