11E Page 14
Each QKP number, and each QKP list, should only be used once. However sometimes the same QKP number and settings are used for two (usually consecutive) transmissions. As long as a limitation involving P5 is not being used the key generated will be the same for both transmissions and they will be in DEPTH.
If the Tunny machine is switched out and a new transmission started without resetting, there is said to be a FOLLOW-ON. 11B (k) shows that the decodes of the two parts of a follow-on will be divided by two blanks for which nothing will have been transmitted.
(e) Change of keys.
Once a day (usually between 0600 and 1200), some or all of the wheel patterns are changed. The sender sends out QZZ (usually in clear) and this tells the receiver that he is changing over to the new day's patterns, and that the receiver's incoming Tunny must also be changed.
Before Summer 1944 motor patterns were changed daily but chi patterns were changed monthly and psi patterns monthly or quarterly (see 11E(a)). During the summer changes became more frequent, and after August 1st there was a daily change of all wheel patterns on all links.
Wheel patterns were issued for a month at a time. A day's wheel patterns - as issued - are shown in Fig. 11 (III) where + = Nocke and o = Keine.
11E THE TUNNY NETWORK.
(a) The period of experiment.
The Tunny machine (SZ40 with no limitation) made a first and experimental appearance in June 1941 on the link Berlin - Athens - Saloniki. At first it was used crudely enough.
(i) Wheel patterns were not chosen so that ab = ½, and there was a regular excess of dots over crosses in the ΔΨ' stream.
(ii) The QKP indicating system had not been introduced and wheel settings were chosen by the sender, and sent out in a simple substitution of letters for settings which changed every month and was different for each wheel.
(iii) Motor patterns were changed daily, chi patterns monthly, and psi patterns every three months.
(iv) The machine was not wired to a tone transmitter, but the cipher text was recorded and sent by facsimile (Hellschreiber).
Until October, 1942 there was still only one Tunny link, but the procedure gradually improved with the introduction of ab = ½ and of Tone Transmission before March, 1942.
The replacement of the single link by two links - Codfish from Berlin to Saloniki, and Octopus from Koenigsberg to South Russia - using the QKP system and with monthly change of chi and psi patterns signified the end of the German experimental period and the start of the general expansion of the Tunny system.
(b) The period of expansion.
SZ42A was first introduced on Codfish in February 1943, and gradually replaced SZ40 on all links.
SZ42A was fitted with a P5 attachment which was used experimentally on Herring (Rome-Tunis) in March 1943, but only made a general appearance after December 1943 on Western European links.
At the time of the allied invasion of the continent in 1944, Tunny had reached its most widespread and stable level of organisation. There were 26 links and two main central exchanges.
STRAUSSBERG near Berlin - the terminus for the 9 Western links and
KOENIGSBERG the terminus for the 10 Eastern links.