Alan Turing's Report from Washington, 1942

Alan Turing's Report from Washington DC, November 1942

This is a transcription of pages 21 and 22 of Turing's report from Washington dated 28 November 1942, in the (British) National Archives, box HW 57/10, released from secrecy in October 2004.

See the Alan Turing Home Page for a guide to this website

For an introduction to the report, and its opening pages, continue to this page.

The first section of this extract is notable for the fact that Turing here casually strayed into a matter of immense political sensitivity, viz. how the Americans were going to get their 'cribs' and how independent they would be of the British. Turing's view of the United States as a remote outstation would not have sounded well in American ears. His description of the American Bombe designs as 'a crazy scheme' might also have jarred. The second section contains an interesting reference to 'my' method of attacking the Tunny cipher material, and shows how wide a remit Turing had to advise on Bletchley Park matters.

For more on the context of this report, go to this page of the Alan Turing Internet Scrapbook.

MOST SECRET (iv) Bombe and Enigma Machinery The Bombe machinery is all over in Ohio, so I have not yet seen any of it, but intend to go over there when I make my next visit to Washington. There were however some discussions of the plans for them; these plans left me rather uneasy. They are apparently not adopting our principle of not having the Bombe stop at all. The[y] intend to have them stop automatically, automatically wheel back to the place where the 'stop' was, and then do all kinds of subsidiary testing. This strikes me as a crazy scheme. For one thing the subsidiary testing apparatus will be idle while the Bombe part is working and conversely. All this stopping and starting business just seems all wrong somehow. However Engstrom himself seems to think the same, and indeed he was the first to mention it. I think that now that Eachus is back we need not really have any fears on this question, especially if he keeps in touch with Welchman. The idea that the Americans should do their own cribbing and put what cribs they can get onto their Bombes seems to me a mistake. The only economical method to my mind is for cribs to be produced in England, possible even by American cryptographers, and that Ohio or Washington should be treated as a kind of distant Adstock. I seem to remember that we talked about this once before and that you agreed but said it just could not be helped. You asked me to give my opinion as to the applications of any of the machinery here to problems other than E[nigma] problems. When it comes to the point I find I cannot really give any specific cases, apart from applications of Baby; here I could give a number, but I think that in that case it is only necessary to pass on the relevant part of my report to Freeborn who will be able to think up a few dozen applications. Beyond this I can think of practically nothing outside the E field. I am inclined to think that this is due to my not being so directly in touch with anything else, rather than because there are no applications. As regards all the apparatus connected with the teleprinter tape I am sure that there must be applications in almost all sections at B.P. because in all sections one wants to be able to write one's material out in different forms. I should be inclined to suggest that our orders in this line should at first be made rather more than is required for Hut 8 and Hut 6 requirements, so that other sections could take it up gradually. Teleprinter tape presumably has special applications in the Tunny field, owing to the Tunny working off the same lettercode. One might for instance be able to produce a rather specially simple Tunny deciphering machine which would work off teleprinter tape, and one would then be able to save time by punching up all one's messages onto teleprinter tape as soon as they came in, and then feed them through the deciphering machine when the keys were discovered. The machine which writes out the letters from the tape in teleprinter code should also be useful, e.g. when applying my method of recovering wheel patterns from a stretch of key. These machines are constructed out of units which have to be connected together in much the same way as with our 'Letchworth' machines. In order to be able to give any sort of estimate of the number of things required it is necessary to describe what are the units available and which units are used in which machines. [The report continues with listing of technical data about paper-tape machines].

Continue to The opening of this report.

Continue:

Quick Links: