General Report on Tunny
25D Page 165
It may well happen for example, that in one part of a message 8's are more numerous than 5's, whilst in the rest 5's are more numerous than 8's. Evidently more evidence is obtained by spanning the parts separately. This however take twice the time and is not done unless there is difficulty in completing the wheels, or in making them certain. (R5 p 11.)
(e) Wheel characteristics.
Restrictions on permissible chi wheels have been mentioned (11C). Wheels which conform are said to be legal. Any deltaed wheels must have an even number of crosses, and for legality both Δ'd and un-Δ'd wheels must have, as nearly as possible, equal numbers of dots and crosses. For the five wheels the number of crosses is:
| Δ | 20 | 16 | 14 | 12 or 14 | 12 | |
| un-Δd | 20 or 21 | 15 or 16 | 14 or 16 | 13 | 11 or 12 |
The final form of legality forbade more than four consecutive like characters in the un-Δ wheel, i.e. more than three consecutive dots in the Δ'd wheel.
If most characters are certain on the evidence from wheel-breaking runs, the requirement of legality may suffice to complete the wheels: various tricks have been devised for doing this easily.
 |
The underlining of three characters indicates, conventionally, that they are doubtful; the first and second of these cannot be interchanged, for if they were the characters of chi 5 between them would be changed in sign, increasing the number of crosses to 13. Similarly the first and third of these can be interchanged.
At an early stage in wheel-breaking it may be justifiable to accept rather weak scoring characters which interrupt what would otherwise be long strings of dots. Of course until it is known which may round (25D(F)) the
Back to General Report on Tunny. Contents.