The First War of Physics (27 page)

Kurchatov wrote two memoranda for the State Defence Committee summarising his interpretation of the espionage materials. In the second, dated 22 March 1943, he emphasised the importance of element 94: ‘The prospects of this direction are unusually captivating’, he wrote. When asked for his opinion on the materials by Molotov, he declared that they were wonderful: ‘[T]hey fill in just what we are lacking.’ Molotov introduced him to Stalin, who promised every kind of support.

It seemed reasonable to suppose that further work on uranium isotope separation and element 94 had been done in America, and Kurchatov closed his second memorandum with an appeal: ‘In this connection I am asking you to instruct Intelligence Bodies to find out about what has been done in America …’ He listed four questions concerning element 94 that he wanted answers to, and requested an update on the work being carried out with cyclotrons. He also provided a list of American laboratories that should be targeted.

Berkeley’s Rad Lab headed the list.

The Russian diplomatic problem

As the battle of Stalingrad wound to its bloody conclusion, Colonel Carter Clarke, chief of the Special Branch of the US Army’s Military Intelligence Division, grew concerned that Moscow and Berlin might seek to negotiate a separate peace. On 1 February 1943 he ordered the army’s Signals Intelligence Service to begin a small programme to study the encrypted Soviet message traffic passing between the US and Moscow. He had hoped
to find clues to any such negotiation. The effort was referred to simply as the ‘Russian diplomatic problem’.

The Soviets operated a ‘belt and braces’ approach to coding their messages. A message from a Soviet diplomat – or a spy – resident in America would first be coded into sequences of four-digit groups, using a code book. These were then regrouped into five-digit groups. The five-digit groups were then enciphered using a so-called one-time pad. Each page of the pad consisted of 60 five-digit random number groups which constituted the encryption key. By adding the original five-digit groups to the appropriate sequence of groups on the key page, a cipher clerk would produce a new series of five-digit groups. These were then translated into five-letter groups. The end result was a message consisting of seemingly random five-letter groups, which was then cabled to Moscow.

The cipher clerk at the receiving station would identify the specific page of the one-time pad that had been used, and decipher the message using the key page and the code book. It was a virtually unbreakable cipher system.

However, the whole point of a key page in a one-time pad is that it should be used only once. But the Soviet message traffic had grown so vast during the second half of 1941 that the cryptographers were unable to produce new key pages quickly enough to meet demand. In early 1942 the cryptographers started to produce duplicate key pages, doubling their output for no extra effort. About 35,000 duplicate key pages were shuffled and bound into different one-time pads during 1942, effectively making them two-time pads.

It also made the messages vulnerable to cryptanalysis.

¹
Otto Frisch made the same discovery (see Chapter 3), but did not publish his results until after the war.

²
On 28 October Groves drafted a letter in Secretary of War Henry Stimson’s name requesting that Szilard be interned as an enemy alien. Stimson refused to sign it. See Lanouette, pp. 238–41.

³
Oppenheimer later purchased the freehold.

⁴
Falkenhorst was tried for war crimes in July–August 1945. He was found guilty of seven (of nine) charges, including a charge relating to the murder of nine commandos of Operation Freshman. His initial death sentence was commuted to twenty years’ imprisonment. He was released on the grounds of ill-health in 1953, and died in 1968. Rediess committed suicide in May 1945.

Chapter 9

January–August 1943

B
y January 1943, the collaboration between British and American nuclear scientists had ground virtually to a halt. The Americans were stalling, and the British were deeply suspicious of the reasons.

Akers had met with Groves in Washington in early November 1942 and had been advised that concern for security was behind the Americans’ growing tardiness. Groves had been alarmed by the extent to which British physicists had been in contact with their American counterparts across very different parts of the programme. Why, Groves enquired, had Peierls, supposedly busy with the physics of gaseous diffusion as part of Tube Alloys, been in contact with American physicists involved with the study of fast-neutron chain reactions?

Groves favoured the preservation of security through compartmentalisation, breaking the various parts of the programme down into self-contained projects in which the scientists and engineers involved knew absolutely nothing of the rest. The behaviour of British physicists, who were in any case too few in number to be compartmentalised this way, was therefore at odds with Groves’ requirements.

Akers feared that there was more behind the Americans’ unwillingness to co-operate than concerns for security. His fears were justified. Groves – a committed Anglophobe – Conant and Bush were all firmly agreed that full exchange of information with the British meant providing Tube Alloys with details of processes and plants to which the British had made no contribution. These were processes and plants being developed and built by American science funded by the American taxpayer. To gift information to the British, they concluded, would do little to aid the American (or, indeed, the Allied) war effort and could only be of value in supporting Britain’s post-war ambitions.

Matters finally came to head in January 1943 when Conant shared a draft memorandum listing proposed general rules and regulations governing Anglo-American collaboration on the atomic bomb. The British were horrified.

The Conant memorandum ‘derives from the basic principle that interchange on design and construction of new weapons and equipment is to be carried out only to the extent that the recipient of the information is in a position to take advantage of this information
in this war’.
This meant no further information was to be supplied by the Americans on electromagnetic separation, the production of heavy water, fast-neutron chain reactions and the manufacture of uranium metals and compounds such as uranium hexafluoride. Information transfer on the subject of gaseous diffusion was to be controlled by Groves.

The memorandum extended the proposed rules to information exchange with the group that had in the meantime been established in Montreal. It stipulated that while information on chain reactions could be exchanged with Met Lab physicists in Chicago, there was to be no further information transmitted on the properties or production of plutonium.

When Anderson saw the memorandum he was furious. He urged Churchill to discuss it with Roosevelt. But, unknown to the British, Roosevelt had already endorsed the stance the Americans were now taking. Churchill’s telegrams to Roosevelt’s aide, Harry Hopkins, went unanswered. A review of the cost to Britain of going it alone on the atomic bomb only served to confirm earlier conclusions. Without the Americans
the bomb was out of Britain’s reach, certainly within the likely timeframe of the war, and probably for many years thereafter.

In May 1943 Anderson discovered that the Americans had been quietly buying up refined uranium oxide from Ontario’s Eldorado mine, and had secured rights to production for years to come. This had happened despite commitments from the Canadian government that the British would have joint control over the supply of uranium. The Americans had also secured a supply commitment from the heavy water plant which was to be built by the Consolidated Mining and Smelting Company in Trail, British Columbia.
¹
Without access to uranium or heavy water, the Montreal project could not continue, and work ground to a standstill in June.

Churchill raised the matter with Roosevelt directly, following the conclusion of the two-week Trident conference in Washington in May. Roosevelt gave his personal assurance that the exchange of information with Tube Alloys would resume. But any hopes that this would happen were scuppered by Cherwell who, at a meeting held simultaneously in Washington on 25 May with Bush and Stimson, explained the real reason for Britain’s desire to resume the exchange of information. This was necessary, he argued, so that Britain could pursue an independent atomic bomb programme after the war. The post-war vision for Britain was one of empire and atomic power. Cherwell’s gaffe merely confirmed what Bush had always suspected. Bush informed Roosevelt of the discussion.

Having given his personal assurances to Churchill, Roosevelt was now perturbed by this evidence of British intent to hold on to its position on the world stage. Roosevelt was secretly committed to dismantling the British empire, turning former British colonies into sovereign nationstates with an American-style constitution. He made no final decision and the impasse dragged on.

Espionage assault

Stalin had realised that the defeat of Germany and Japan would leave three major world powers – the Soviet Union, the United States and Great Britain. Of these, Britain was the weakest, the sun surely setting on Churchill’s dreams of empire. The Soviet Union already had a strong intelligence network operating in Britain, reliant on prominently placed, ideologically motivated individuals across all walks of political, military and scientific life. However, until the German invasion of the Soviet Union in 1941, America had not been regarded as an important target for Soviet espionage.

The German invasion forced the Soviet Union into an uneasy alliance with Britain and America, and the Soviets began to benefit from American ‘Lend-Lease’ military aid.
²
To manage the procurement and transport of Lend-Lease
matériel
– weapons, aircraft, vehicles and machinery – the Soviet Union greatly expanded its diplomatic presence in America. Among the diplomats arriving on American soil were many NKVD and GRU spies, as the Soviet Union launched an unprecedented espionage assault on its ally.

Pavel Fitin, head of the NKVD’s First Directorate, established dedicated
rezidenturas
– bases for espionage operations – at the Soviet embassy in Washington, the Soviet consulate in New York, and the consulate in San Francisco. The Washington
rezident
, or station chief, Vasily Zarubin (cover name Vasily Zubilin) had arrived in America in October 1941. Supported by his wife Elizabeth, who was also an NKVD agent, he was charged personally by Stalin to gather intelligence on American intentions towards Germany.

In late 1941/early 1942, Soviet vice consul and NKVD agent Anatoly Yatskov (cover name Anatoly Yakovlev) and trainee Alexander Feklisov (cover name Alexander Fomin) had been installed at the Soviet consulate at 7 East 61st Street in New York, between Madison and Fifth.
³
The New York NKVD
rezident
, Gaik Ovakimyan, had been exposed as a spy by the FBI in April 1941 and had been obliged to leave the country. His role was filled temporarily by Pavel Pastelnyak, before Zarubin was assigned responsibility for the New York station.