BOOK REVIEW / Ultra, Purple, Magic: all crackers: 'The Ultra-Magic Deals & the most secret special relationships 1940-1946' - Bradley F Smith; Airlife Publishing, 17.95

ON the last day of August in 1940, well over a year before a disastrous intelligence error made possible the Japanese raid on Pearl Harbor and so brought the United States into the war, an American officer, Brigadier General George Strong, astonished the British chiefs of staff, at a meeting ostensibly devoted to the technical question of arms standardisation, by proposing a very special exchange of information between the British and American governments. Specifically, he referred to 'certain methods' by which information in the possession of the United States might be placed at Britain's disposal.

What General Strong had in mind was something so extraordinary that in ordinary circumstances it would have been inconceivable: that warring Britain and neutral America share the products of 'cryptanalytic' intelligence: that is, the fruits of broken enemy codes and ciphers.

Before 1940 the governments of Britain and the US were not on particularly good terms; one historian has written, accurately, of 'indifference, suspicion and bitterness' after the United States broke up the First World War alliance, and for generations Americans had been as disapproving as they were envious of Britain's naval and imperial power.

In intelligence terms, Britain in 1940 was the superpower, the United States by far the weaker partner. Quite apart from the information gathered by the armed services and by the agents of the Secret Intelligence Service, Whitehall had the benefit of the large and highly efficient Government Code and Cipher School at Bletchley Park, which was showing promising results in cracking the machine codes in which thousands of enemy messages intercepted every day were veiled.

The United States, however, had one almost priceless bargaining asset. When General Strong made his offer, the American army's small and underfunded signals intelligence service, under the cryptanalytic genius Dr William Friedman, was in the very process of cracking the Japanese diplomatic code known as 'Purple'. To achieve this, Friedman and his team had actually built a replica of what they reasoned a Japanese Purple encryption machine must be like. The result was so valuable that it was called Magic.

In the spring of 1941 Bletchley achieved an even more important success by beginning to crack the ciphers encrypted by the German Enigma machine for naval codes: this material was to be so valuable that it was code-named 'Ultra'. But in late 1940 the Americans were in the position of a trader who may be in a weak position overall, but has one item to sell which he knows the buyer must have.

The initial exchange of information was disappointing. The Americans sent over a small and junior mission. The British showed them round Bletchley, but revealed little that would help them expand their cryptanalytic effort. That triggered the usual Washington suspicion of perfidious Albion. It was not until October 1942 that the first lasting agreement on the exchange of secret material was signed, covering only naval codes.

In June 1943 that was extended to all types of codes. In 1948, after several ups and downs, it seems to have been extended into a permanent, peacetime exchange of material between not only Britain, but also Canada, Australia and New Zealand and the United States. I say 'seems', because even now the subject is blanketed in secrecy.

Gradually, in the secretive world of cryptanalysis and the relevant technologies, as in the atomic partnership and in the Anglo-American relationship as a whole, the balance of power shifted towards the Americans. At the beginning of the war, Britain had a technical lead in the 'bombes' (mechanical, then electrical analytic machines) which speeded up the elimination of codes until the one that was being used was identified). By 1945, the American technology was superior, and - as in the atomic bomb programme - US officials were not slow to exploit their new strength.

When Alan Turing, the legendary cryptanalyst from Bletchley, was sent to look at new scrambler telephones being developed at Bell Laboratories so that the Germans could not listen in to Winston Churchill's phone conversations with Franklin Roosevelt, as they at least once succeeded in doing, Turing was kept cooling his heels.

In part, this was to emphasise to the British that there were limits to American co-operativeness. But lurking under the surface of the Turing incident, as of the whole story of the 'most secret special relationship', was the intense rivalry between the US Army and the US Navy. Indeed, it was almost easier to arrange co-operation between the heirs of George Washington and the heirs of George III than between the squabbling service chiefs on the American side.

Bradley F Smith, an American historian who spends part of his time in Britain, and who has previously written on the Nuremberg trials and on OSS, the American wartime covert action service, has done an excellent job of telling this intriguing and rather important story.

It is intriguing, because the effect on the fortunes of war of the gradual unravelling of enemy codes has its fascination. And it is important, because secret intelligence (especially signals intelligence) and atomic weapons were the two ultra-secret fields in which the wartime 'special relationship' between Whitehall and Washington developed.

Professor Smith is especially evenhanded in understanding the achievements and failures of the two partners. It will be for another historian to trace the legacy, in secretiveness and false illusions of power, left by those wartime partnerships to postwar Britain. At some point we shall have to assess whether, in the end, the generous American saviour did not inflict almost as much damage on Britain's interests as the despicable Nazi enemy or the equally murderous Soviet ally.