Cryptographic History of Work on the German Naval Enigma

7. The only other event in July was the arrival in Hut 8 of the party working on the J.N.A 20 machine. They came to Hut 8 partly because we had some spare room and Naval Section were very short [of] space and partly because it was thought that an exchange of ideas might benefit them and us. We welcomed this move very much because the policy of running Dolphin on all W.O. cribs had caused a good deal of unemployment amongst both cryptographers and the girls in the Big Room; we therefore did quite a large amount of work with the J.N.A. 20 party, this starting a profitable association which lasted until well on in 1944.


8. The chief event of interest in August was a break into Sunfish. This was a key carrying information about blockade runners trying to return home from Japan and, to a lesser extent, about outgoing ships. It used the same indicating system - "throw-on" - as Porpoise and but for this we should quite certainly never have broken it. The indicators provided in themselves enough information for a bombe menu and so, although we did not think that the traffic was in the Naval Enigma (some form of the unsteckered machine was thought more likely), we ran the menu to make sure and the day came out. Normally the bulk of the traffic on Sunfish was what we called "Love and Kisses" i.e. family messages but when boats were returning home it carried traffic of more interest; as it was rather an expensive key in bombe time - throw-on menus derived from the indicators only were troublesome on the machine - our policy was not to bother much about it as a rule but in times of "flap" (i.e. returning blockade runners) to break as much as we possibly could.


9. In September another new key Seahorse [Kriegsmarine] of rather similar type to Sunfish, was broken. This was the Naval Attaché cypher used between Berlin and Tokio and also had a throw-on system of indication, "throwing-on" on four letters and was in fact a 4 wheel key. Since it had used a four letter indicator since 1940 it is highly probable - we never tried to break the very early days - that it was four


< previous

next >