Editing Battle of the Atlantic (section)

===Enigma cipher===
{{See also|Cryptanalysis of the Enigma#German Navy 3-rotor Enigma|l1=German Navy 3-rotor Enigma}}

The way Dönitz conducted the U-boat campaign required relatively large volumes of radio traffic between U-boats and headquarters. This was thought to be safe, as the radio messages were encrypted using the [[Enigma machine|Enigma]] [[Rotor machine|cipher machine]], which the Germans considered unbreakable. In addition, the {{lang|de|Kriegsmarine}} used much more secure operating procedures than the {{lang|de|Heer}} (Army) or {{lang|de|Luftwaffe}} (Air Force). The machine's three rotors were chosen from a set of eight (rather than the other services' five).{{sfn|Erskine|2004}} The rotors were changed every other day using a system of [[Key (cryptography)|key]] sheets and the message settings were different for every message and determined from [[Cryptanalysis of the Enigma#German Naval Enigma|"bigram tables"]] that were issued to operators. In 1939, it was generally believed at the British [[Government Code and Cypher School]] at [[Bletchley Park]] that naval Enigma could not be broken. Only the head of the German Naval Section, [[Francis Birch (cryptographer)|Frank Birch]], and the mathematician [[Alan Turing]] believed otherwise.{{sfn|Copeland|2004|p=257}}

[[File:Enigma rotors and spindle showing contacts rachet and notch.jpg|thumb|upright=1.0|Enigma rotors and spindle]]

The British codebreakers needed to know the wiring of the special naval Enigma rotors. The capture of several Enigma rotors during the sinking of {{GS|U-33|1936|2}} by {{HMS|Gleaner (J83)}} in February 1940 provided this information.{{sfn|Sebag-Montefiore|2004|p=76}} In early 1941, the Royal Navy made a concerted effort to assist the codebreakers, and on 9 May  crew members of the destroyer {{HMS|Bulldog|H91|2}} boarded {{GS|U-110|1940|2}} and recovered cryptologic material, including bigram tables and current Enigma keys. The captured material allowed all U-boat traffic to be read for several weeks, until the keys ran out; the familiarity codebreakers gained with the usual content of messages helped in breaking new keys.

In August 1940, the British began use of their "[[bombe]]" computer which, when presented with an intercepted German Enigma message, suggested possible settings with which the Enigma cipher machine had been programmed. A reverse-engineered Enigma machine in British hands could then be programmed with each set of suggested settings in turn until the message was successfully deciphered.<ref name=Wenger1945>{{cite report |author=Wenger, J. N. |date=12 February 1945 |title="Appendix II: U. S. Army Cryptanalytic Bombe", Solving the Enigma: History of the Cryptanalytic Bombe. National Archives and Records Administration Record Group 457, File 35701. |publisher=United States National Security Agency |url=http://ed-thelen.org/comp-hist/NSA-Enigma.html |url-status=dead |archive-url=https://web.archive.org/web/20141002211707/http://ed-thelen.org/comp-hist/NSA-Enigma.html |archive-date=2 October 2014 |access-date=11 July 2023}}</ref>

Throughout late 1941, Enigma intercepts (combined with HF/DF) enabled the British to plot the positions of U-boat patrol lines and route convoys around them. Merchant ship losses dropped by over two-thirds in July 1941, and the losses remained low until November.

This Allied advantage was offset by the growing numbers of U-boats coming into service. The [[German Type VII submarine|Type VIIC]] began reaching the Atlantic in large numbers in 1941; by the end of 1945, 568 had been [[Ship commissioning|commissioned]].<ref>{{cite web |url=http://www.uboat.net/types/viic.htm |title=Type VIIC |last=Helgason |first=Guðmundur |website=German U-boats of WWII – uboat.net |access-date=13 February 2010 }}</ref> Although the Allies could protect their convoys in late 1941, they were not sinking many U-boats. The Flower-class corvette escorts could detect and defend, but they were not fast enough to attack effectively.