THE MANNER OF OCCURRENCE OF MUTATION 465 
body branches). Those portions of the F, cell lines after the 
first division which play any conspicuous role in the production 
of the soma are distinguished with dotted lines; mutations 
occurring here would often be visible as mosaics. 
On examination of this diagram, it is evident that mutations 
in all portions of the continuous line subsequent to its last split- 
ting into two continuous lines and previous to the first cell 
division in the offspring would result in single (entire) mutant 
F, individuals. These sections of the line have been represented 
as double. On the other hand, a mutation in any part of the 
continuous cell line antecedent to the last split into two con- 
tinuous lines would result in twin or multiple mutants among | 
the offspring, the exact number of mutants being determined by 
the number of splits in the line subsequent to the point of muta- 
tion; these parts of the continuous line have been represented as 
single in the diagram. The way, then, to find the a priori 
chance for the appearance of single mutants, as compared with 
that for cases of twin or multiple mutants, is to compare the 
total length (in horizontal projection) of the double lines with 
the total length of the single lines antecedent to the double 
portions. If mutation is entirely indiscriminate, that is, equally 
likely to occur at all stages, then the ratio of the lines will give 
the ratio of single to multiple mutations, but if mutations are 
more likely to occur at one stage than another, then the line- 
ratio will differ from the ratio of mutations by the same propor- 
tion as the average mutability of the ‘double-line’ stages of the 
life-cycle differs from the mutability of the ‘single-line’ stages. 
The projected length of the double line in the diagram as 
compared with the single line, is about 3.5 : 1. Although the 
cell lineage is for the most part only guessed at in this diagram, 
still it should be pointed out that the lengths of those portions 
of the double line lying beyond the multiplication period (the 
four divisions which give rise to the cyst) are approximately 
correct for Drosophila, relatively to the length of the entire 
cycle. For it may be calculated from the data of Plough (’17), 
which gives the number of odcytes in the ovary and the number 
of eggs hatched per day, that the length of the growth period 
