36. In April the first 4 wheel bombe arrived but we were to get no real value from these machines until June. The whole 4 wheel bombe problem is discussed in Chapter so I will not interrupt the narrative here to consider it.
37. May saw the beginning of regular meetings between Hut 8 and those parts of Naval Section that had contact with us. These meetings continued fortnightly until the end of the war and performed a valuable function as a general lubricant and in keeping Naval Section fully informed of the cryptographic position; they arose out of a series of meetings on the four wheel bombe position but later covered the whole field of Hut 8 - N.S. relations. At this time also we started a drive to reduce delays in handling traffic at all stages. Time clocks were introduced into the Hut and figures kept of the average time for each stage of the work. The results of these checks showed quite conclusively that the only satisfactory way of improving times was to take actual figures and show the results to those concerned - vague statements that things were slow here and should be quicker there were completely valueless. Through these tests times at all stages from Scarborough to the Z watch were very markedly improved. In June 1944 when we temporarily had an intercept station on the spot current traffic reached Admiralty on an average 30 minutes after the time of intercept. The record was 19 minutes in which time the message was sent from the intercept set to us, registered, decyphered, sent to the Z watch, translated by them and teleprinted to Admiralty.
38. June was an eventful month and marked the change from breaking Shark on short signals and Dolphin on Banburismus to a period when R.E's and cribs dominated all keys. There were two notable individual feats in the early part of the month. Pendered broke May 27th Shark by hand on a 400 letter R.E. using a method known as the "Stecker Knock Out". Although the theoretical possibility and general method were well known this was the first success ever obtained on a 10 stecker key and was of considerable value. Pendered had refined existing methods in various ways and his