To have sufficient material to break a day on the bombe a crib of 30 letters or more was normally needed; when cribs are referred to, the phrase should be taken to mean a guess at the plain text for not less than 30 letters.
The bombe was rather like the traditional German soldier, highly efficient but totally unintelligent; it could spot the perfectly correct answer but would ignore an immensely promising position involving one contradiction. The effect of this was that if one letter of the cypher text had been incorrectly intercepted, the menu would fail although both the crib and the text were elsewhere absolutely correct... This was the cause of the extensive double banking programme, which has already been described.
Cribs were sent to the bombe Hut in the form of menus with directions as to the wheel orders on which they were to be run. No more was heard of them until the possible positions, known as 'stops', started to come from the bombes as they worked through wheel orders. The strength of menus was calculated with a view to the bombe giving one stop on each wheel order, thus supplying a check that the machine was working correctly. The identification of the right stop, the stop giving the correct Stecker, and the rejection of the wrong ones was done in Hut 8.
In order to get the maximum use from the bombes they had of course to be kept fed with menus for 24 hours a day and the art of bombe management required a certain amount of skill and experience. The plugging up of a new menu was a comparatively complicated and lengthy process so that it was desirable to give a bombe as long a run on a menu as possible: on the other hand an urgent job would justify plugging up a large number of bombes for only a few runs because of the importance of saving time. Efficient bombe management was largely a matter of striking the happy medium between speed and economy, of making sure that, with a limited amount of bombe time, everything of importance got run and that as far as possible the urgent jobs were run first. Bombe management was interesting because the situation a few hours ahead was to a large extent incalculable: allowance had to be made on the one hand for jobs which one expected to have to run in 12 hours time (and which when the time came did not always materialise) and for the fact that sometimes 2 or 3 jobs would come out in quick succession on one of the first wheel orders to be run, thus releasing large numbers of bombes: because of this possibility it was necessary to keep a reserve of fairly unimportant jobs to fill the gaps.
Pressure on the bombes varied greatly with the immediate cryptographic situation and the period of the month, it being generally true to say that towards the end of the month the