# General Report on Tunny

23D Page 83

C3 R 1259 A 315 Ø 15.4 ST 353
K4 K5
05 06 B 0369
05 13 B 0356
05 22 B 0379       B 64 4.15Ø
17 06 E 0357
18 11 D 0363
26 11 A 0357
26 11 B 0358

SET K4 K5   05 22 GOOD

///   555   UUU
168   212   177

555 A 106 Ø 7.2 ST 115
K3
02 A 0125    .
08 A 0146            B  40 5.6Ø
10 A 0116
12 A 0116
14 A 0118
16 A 0125

K1 K2 K3 K4 K5
40 28  08 05  22    SPAN 01-4500
AG X2 XX .    AG X2 ...
 / 0086 0092 9 0082 0084 H 0063 0067 T 0052 0064 O 0084 0055 M 0086 0078 N 0044 0063 3 0095 0057 R 0058 0059 C 0053 0053 V 0042 0073 G 0075 0082 L 0053 0094 P 0088 0067 i 0072 0055 4 0057 0065 A 0066 0078 U 0111 0067 Q 0067 0066 W 0058 0047 5 0146 0074 8 0066 0060 K 0057 0060 J 0096 0058 D 0049 0064 F 0120 0074 X 0064 0082 B 0054 0067 Z 0055 0071 Y 0087 0068 S 0081 0066 E 0056 0067 ALL CERTAIN SPAN 01-4500E.M. LEWIS.

Instead the operator uses C3, i.e. 4=5=/1=2. R 1259 is the number of places looked at, i.e. of places where 1=2. A, the expected random number of places where 4=5=/1=2 is a quarter of this, because two more conditions are imposed, and .

ST as above.

The best score is 4.15σ. It should be noticed that three scores above the set total have Χ4 at 05; and that the Χ4 settings are 22, 22+7, 22+14, suggesting a wheel-slide [23G] of 7 on Χ4, but fortunately one too weak to make the setting really doubtful.

The most likely letters to set Χ's are /, 5, U: to decide which the operator counts / & 9, 5 & 8, U & K; and because 5 & 8 are numerous (212) chooses 5

A and σ are as before, but ST is now only A+σ, because this is a one-wheel run.

The setting obtained is "certain".
 K1 K2 K3 K4 K5 40 28  08  05 22 is printed by means of P.M.H. so that errors in Colossus
setting will be manifest.

AG X2 XX means against Χ2=x.

The settings for Χ1 Χ2 Χ3 Χ4 X5 were not very strong, so that although the final Χ3 score makes them "certain", it is just as well to inspect the letter count.

This resolves all doubts as to the correctness of the settings, e.g. Χ5 is made very certain by the scores for U>Q, 5>J, F>X. The count is in fact much better than average; the worst contradictions with the pairs in 23 H(e) are CM, DB, ZE.

The dossier is therefore marked all certain.

The letters have been inserted in the letter count for the purpose of this report: Colossus operators knew them too well to need this.

The pen entries at the right of the letter count are the DO's order for a motor run [which appears on the next page but one.]

The part 4500 - end is not yet set: it is believed that there is a message slide so that the settings for all chi's in the part of the message will be increased, or decreased, equally.