Colossus: Bletchley Park's Last Secret

Paul Gannon

• Publisher: Atlantic Books
• Year: 2014
• ISBN:
  9781782394020
• Topic: History
• Subtopic: World History
• Index: History

'Gannon's book contains a mass of utterly fascinating and largely unknown material about an immensely important wartime project, and is very welcome indeed.' Brian Rendell, TES

In 1940, almost a year after the outbreak of the Second World war, Allied radio operators at an interception station in South London began picking up messages in a strange new code. Using science, maths, innovation and improvisation Bletchley Park codebreakers worked furiously to invent a machine to decipher what turned out to be the secrets of Nazi high command. It was called Colossus.

What these codebreakers didn't realize was that they had to fashion the world's first true computer. When the war ended, this incredible invention was dismantled and hidden away for almost 50 years. Paul Gannon has pieced together the tremendous story of what is now recognized as the greatest secret of Bletchley Park.

Contents

A Word about Words
Introduction
PART ONE - INTERCEPTION
1	Wireless War One
2	Codes and Ciphers
3	Between the Wars
4	Wireless War Two
5	A Window on a War
6	NoMo
7	‘If the Wind Meets it’
8	Knockholt
PART TWO - DECRYPTION
9	‘HQIBPEXEZMUG’
10	Fishing the Depths
11	Herring and the Cat’s Whiskers
12	‘Hier ist so traurig’
13	Making the Difference
14	The Robinson Family
15	Inventing the Electronic Computing Machine
16	‘Colossus Arrives Today’
17	Fish Dialects
18	Fish – Landing the Catch
19	A Day in the Life of Fish I
20	A Day in the Life of Fish II
21	The Technology and Organization of Fish
22	Transatlantic Fisheries
23	Siginstitutionalization
24	Colossus – The Legacy
	Epilogue
PART THREE – APPENDICES
A	The Baudot. Code with Shift Function
B	The Vernam Cipher
C	Extended Example of ‘The First Break’
D	Structure of the Lorenz Schluesselzusatzgeraet SZ40/42
E	The Delta Techique
F	Robinson: Block Diagram of Major Function Units
G	Non-machine Problems Identified During the Heath Robinson Period
H	Basic Functional Sub-systems of Colossus
I	Annotated Printout of Colossus 3 run 23/10/44 (Stickleback)
J	Colossus ‘Processing Tree’
K	Sturgeon and Thrasher
L	Bream Message Types and Examples – Early 1944
M	Some Jellyfish Messages from May and June 1944
N	Cribs Used on Dragon
O	Whiting Decode, 5 February 1945
P	‘Hitler as seen by Source’
Q	Newmanry Staffing
	A Note on Sources
	Notes
	Further Reading
	Index

A Word about Words

Any account of the birth of Colossus and the computer age must involve three particularly jargon-ridden subjects – information and communications technology, codebreaking, and military history. I have tried not to be too pedantic in the use of jargon, preferring to use everyday words wherever possible, which has meant simplifying some technical terms. I have also tried to use contemporary rather than modern terminology. Occasional notes in the text will point to variations and possible confusions.

One particularly hard decision was what to call the ‘units’ of the Baudot code (see Chapter two). Telecommunications engineers called them units, but during the Second World War the British codebreakers used the term ‘impulses’. After the war, the term ‘bit’ (from BInary digiT) came into universal use. In many ways, it would have been easiest to use the now familiar ‘bit’, but it creates a false historical impression (in its use before its time). As ‘impulse’ is potentially confusing, I have used ‘unit’ throughout.

Codebreaking, in particular, suffers from problems of terminology. ‘Codebreaking’ is itself a good example. It clearly means breaking a code, but it is also widely used to mean ‘cipher breaking’ (which can also be called ‘cryptanalysis’). Codes and ciphers are generally different things (see Chapter two), but ‘codebreaker’ has come to mean people who crack ciphers as well as codes. By extension, what should properly be called a ‘decrypt’ (i.e. an enciphered message for which the key has been worked out and used to decipher the message to reveal its plain-language content) was usually known at Bletchley Park as a ‘decode’, a term which I have thus adopted. I have also sometimes used the term ‘intercept’ to mean the same as a decode and decrypt. One particularly misleading term is ‘plain language’ (‘plain’, ‘in plain’, etc. and also as ‘clear language’, ‘clear’, ‘in clear’, etc.). The plain-language version of intercepted messages was usually far from plain, consisting of German military terms and abbreviations and often highly formatted reports consisting of figures and punctuation, as a glance at Appendix C – Whiting decode, 5 February 1945 will show. To make matters more difficult, the interaction between the transmission code and the cipher, both of which lay at the centre of the Colossus story, made the plain language even less plain by inserting machine commands (see Chapter two).

Introduction

Towards the end of the Second World War, civilian and military staff who had been employed at Britain’s highly successful codebreaking organization the Government Code and Cipher School (GC&CS), based at Bletchley Park, were all advised that, even after the end of hostilities, they were bound to secrecy about their wartime work. Bletchley Park had been the centre of a massive interception and codebreaking operation that had helped the Allies win the war. Decrypts of German enciphered military messages, revealing strategy and tactics, had been passed to Allied operational commands from as early as 1939 under the name of Ultra. By and large, there were no serious security breaches during the war and, despite the occasional scare, the German armed forces remained wholly unaware of the extent of Allied codebreaking. Now it was essential that tight security be maintained.

The staff were told, ‘The end of the German war is now in clear view. The several strands of keenness, discipline, personal behaviour and security have been admirable and have combined to produce a direct and substantial combination in winning the war.’ However, there were still things to do: the war against Japan had to be won; there was the need to ease the transition from war to peace for everyone; and, most important of all, it was essential ‘to ensure that nothing we do now shall hinder the efforts of our successors… I cannot stress too strongly the necessity for the maintenance of security… At some future time we may be called upon to use the same methods. It is therefore as vital as ever not to relax from the high standard of security that we have hitherto maintained. The temptation to “own up” to our friends and families as to what our work has been is a very real and natural one. It must be resisted absolutely.’¹ Another document was more forceful: ‘All persons concerned must remember that they are bound by honour as well as law to maintain secrecy of Ultra in Peace as well as in time of War.’ Not even their husbands (the majority of GC&CS staff were women) or, for that matter, their wives could be told. The most severe penalties would follow if anyone so much as dared to publicize what they had done or seen. A trial, with a complete ban on any reporting of it, would end with a very long prison sentence and a sorry future.

Naturally, the wartime secrecy was essential – any hint that Britain and the United States were successfully decoding German signals would have led to a tightening of German wireless security and the loss of the intelligence the Allies gleaned from the intercepted messages. Yet the blanket ban was to be held in place for nearly three decades and, during that entire period, histories of the Second World War were written without any awareness of how the Allies managed to achieve some of their most significant victories. The comprehensive history of codes, ciphers and decryption remains Codebreakers, a massive tome written by the American historian David Kahn. Kahn records in conscientious detail the activities of codebreakers through the ages, culminating with the world wars of the twentieth century, and in the first edition of the book, published in 1967, he reported on how, in the Second World War, ‘Some of the most important British communications intelligence resulted, however, not from the scribblings and quiet cogitations of reticent cryptanalysts, but from the explosive sexual charms of a British secret agent in America.’² The reality was otherwise: since the beginning of the twentieth century, spies, regardless of their sexual charisma, have been far less productive than backroom codebreakers in providing intelligence to both the military and governments – a notable exception, perhaps, being the success enjoyed by the Soviet Union in recruiting agents in both Britain and America. Yet such tales of voluptuous agents made more attractive reading and, more importantly, they led the trail away from codebreaking. For interception and decryption operations were undergoing a massive and continuous growth as the Cold War intensified, with nuclear weapons presenting a threat to humanity many times greater than any posed by even Hitler at his most rampant, and those running these operations were determined that the whole subject must remain wholly secret – or see its usefulness destroyed.

But hints that the German Enigma cipher machine had been broken eventually started to come out abroad, beyond the jurisdiction of Britain’s Official Secrets Act. In Poland and France, the few individuals aware of the essential contributions made by Poles and the French to cracking the Enigma machine were no longer willing to remain silent. And a sort of reverse pride was at work in Italy, where, unlike in Britain, you can be prosecuted for libelling the dead. The Italian navy had long simmered with anger at accusations by the Germans that it had allowed a spy to acquire the information that led, in 1942, to a successful attack on Italian naval ships that had been intended to supply German troops in North Africa. The Germans blamed the incident on a spy who used sex to gain information from a fallible Italian admiral, who in due course died. In the early 1970s, when the admiral’s surviving relatives threatened to sue for libel the author of a book that repeated these charges, it seemed likely that the whole codebreaking story would start to emerge. And, indeed, once speculation about whether the Allies had broken Axis codes began to surface in books published in Germany, as well as in France, Poland and Italy, the wall of secrecy built up around the whole affair began to crumble.

The British authorities reluctantly decided that they had no choice but to allow censored versions of the story of wartime codebreaking activities to be published and, in 1974, the total ban on any mention of these activities was relaxed. After all, it had been an astounding success for the British, and an unregulated flood of rumours might not only tarnish that success but also threaten to provide details of areas about which it was essential to remain silent. It was therefore essential to monitor carefully the release of information. So a spate of personal reminiscences, first among them Fred Winterbotham’s pioneering account, The Ultra Secret, published in 1974, alerted the wider world to the existence of Bletchley Park and its cast of eccentric characters and to the breaking of the Enigma cipher machine. Another popular book, Anthony Cave Brown’s Bodyguard of Lies, helped to disseminate the story. (Both these books were marred by numerous errors and are now considered very unreliable.) A rash of other books followed, of varying quality and openness, and the authorities made every effort to ensure they were published under strict guidelines: what could and could not be...