THE MANNER OF OCCURRENCE OF MUTATION 459 
VI. THE STAGE IN THE LIFE-CYCLE AT WHICH MUTATION OCCURS 
The four cases described in sections I to IV illustrate very 
well the principle that mutation is not a phenomenon connected 
with the segregation division or peculiar to gametes, but may 
occur at any time and place during the life-cycle. Thus, the 
new white arose during that stage of the insect egg which cor- 
responds to early cleavage, in a nucleus that belonged to the 
common germinal and somatic stocks, before the latter had 
separated off from each other. The mutation that produced 
the orange-colored eye also occurred in the very early embryo, 
in a cell that was purely somatic. Ivory arose at a considerably 
later stage, after the germ cells had separated from the somatic, 
and had themselves divided several times to form a number of 
gonia. Ecru, finally, probably occurred much later in the 
life-cycle, either in a very late gonial cell or in an oocyte. 
Cases apparently resembling, in principle, the first two listed 
above are those described by Baur (?) in Antirrhinum. Baur 
found that cuttings taken from this plant at various times occa- 
sionally failed to contain a dominant factor for which the orig- 
inal stem had been heterozygous, and he interprets this as 
meaning that the dominant factor had mutated at some time 
during the development of the cut portion. To the present 
author, however, it would also seem possible that the disappear- 
ance of the heterozygous factor may have been caused by an 
abnormal distribution of chromosomes occurring at some cell 
division—either a ‘somatic segregation’ or a ‘mitotic dislocation’ 
such as gives rise to gynandromorphs in Drosophila. <A test 
between these alternatives would be furnished if the behavior of 
a second heterozygous linked factor could be observed simul- 
taneously with the first. The same alternative explanation of 
mitotic irregularity may also be applied to many of the other 
eases of clonal or asexually produced ‘mutations’ which have 
been reported in the literature. 
Nevertheless, whatever the explanation may be for these more 
doubtful cases, the cases which have been worked out in Dro- 
sophila are by themselves sufficient to refute the doctrine that 
mutations are essentially phenomena of maturation and fertiliza- 
