turn compromised by the use of this system - one of the major cryptographic blunders made by the German Navy in the war.
4. Narwhal split off completely from Dolphin in September and further splits took place on the 17th and 23rd of the month, largely because of enemy fear of key compromise. The position was further complicated by the Dolphin September keys being dropped by plane onto American troops at Brest, causing a series of emergency measures to come into force. This capture of the September keys was not intrinsically important as throughout the early part of the month those were being sent to the Channel Islands daily in messages encyphered on August keys so that we knew them in any case; the real difference the capture made was to alarm the enemy so that he stopped using the September keys altogether. On September 23rd the position was this (1) the main Dolphin key (Norwegian and North Sea area) was using October keys with a stichwort (2) Sucker (Channel Area traffic previously on Dolphin) was using August keys in a hatted order (3) a small third area (Tuebingen, Muensingen) was using October keys with a different stecker. Sucker was of course read immediately and as it re-encoded into Dolphin we broke the latter, giving up September and October keys simultaneously. This glorious muddle was typical of the final stages; what we lost through key splits and we recovered through reuse of old keys compromising new ones, keys sent out encyphered in other Enigma keys, capture and so on.
5. Dolphin gave little trouble in October all the W.O's and thirteen days' steckers having been sent to the Channel Islands in September keys. Sucker changed its nature slightly from the "Channel Area" key to the "Channel Fortress" key - as it repeated old keys until almost the end of the month it made little cryptographic difference anyway. The third branch of Dolphin (Tuebingen/Muensingen) which was never christened with an independent name expired and never reappeared