But even before the first Robinson was finished, Flowers had made a revolutionary proposal which both solved the problem of tape synchronisation, and did away with the laborious production of fresh tapes.
The idea was to store the Fish key-patterns internally, and in electronic form.
If this were done, only one tape would be required. The difficulty lay in the fact that such internal storage would require the more extensive use of electronic valves.
It was a suggestion regarded with deep suspicion by the established experts, Keen and Wynn-Williams. But Newman understood and supported Flowers' initiative.
By any normal standards, this project was a stab in the technological dark. But they were not in normal times, but in the conditions of 1943.
What happened next was a development unthinkable even two years before.
Flowers simply told Radley, director of the Post Office laboratories, that it was necessary for Bletchley work.
Under instructions from Churchill to give Bletchley's work absolute priority,
without questions or delay, Radley had no decision to make, although the development consumed half the resources of his laboratories.
Construction began in February 1943 and the machine that Flowers had envisaged was completed after eleven months of night-and-day working.
No one but Flowers, Broadhurst and Chandler who together had designed the machine had been permitted to see all the parts, let alone to know what the machine was for.
There were no drawings for many of the parts, only the designers' originals; there were no manuals, no accounts, nor questions asked about materials and labour consumed.
In the laboratory the machine was assembled, wired and made to work in separate sections which did not come together until the whole machine was installed and working at Bletchley in December 1943.