and crosses placed in the Mu 61 pattern wherever required. In the first
instance no dots are placed in the Mu61 pattern unless the evidence sup-
porting them is very strong indeed. Efforts are then made to match
the column patterns. Successive assumptions are made as to the number
of dots in Mu61 and inconsistencies used to disprove each wrong assump-
tion. It is best to work with a section of columns under a heavy
band of crosses in MB. This procedure is exactly the same as we have
heretofore used except that it is made more difficult by the breaks
in the columns and by the increase in the number of assumptions to be
disproved. The number of dots in Mu61 now ranges from 10 to 30 whereas
a few months ago the range was only 11 to 19. Of course 24 dots are
never used. The number of dots in Mu37 is invariant throughout the month
(or other key period) in order' to preserve the ab = 1/2 relation.
However, this fact furnishes almost no assistance in recovering the
motor patterns because, by the time the information could be used,
the task will be virtually completed.
4. The number of dots in Mu37 has varied widely in recent months.
For March Bream it is 24. This high a number facilitates solution
from several viewpoints. First, it aids in reovering and setting
the X wheels. 24 dots means that b must equal .74 which is very
good. Second, it helps in recovering and setting the Psi patterns
because the techniques employed depend on finding significant motor
dotsand it therefore helps to have a high proportion.