450 H. J. MULLER 
one (or more) egg strings was entirely heterozygous for ivory. 
This latter hypothesis, however, does not agree well with the 
fact that nine lvory sons were produced in the first five days, 
but none in the last seven days. ‘The latter fact suggests that 
only that part of one egg-string that produced the earlier eggs 
was affected. In that case it becomes necessary to know the 
number of egg strings present in order to interpret the 13 per 
cent relation. ‘This information is not available for female 4711, 
though it is known that the number ranges from about ten to 
twenty. If only a part of one egg string was affected, then, | 
since this part included one-eighth of the total number of eggs, 
it would be necessary to conclude either that the female in 
question had only about eight egg strings, even though she did 
produce 280 offspring in a laying period of nine days, or else that 
the egg string containing the mutant gene produced many more 
than its proportionate share of mature eggs during the first 
five days. On the other hand, it may be that a small portion 
of more than one egg string was affected; in that case the egg 
strings would be ‘polygenetic.’ Although the exact period 
where the mutation occurred cannot be determined, it is certain 
that it took place several cell generations after the separation 
of the germ cells from the somatic cells and several cell genera- 
tions before the last oogonial division. 
Ill. WHITE 
A. Manner of origin 
Besides the original mutation of the normal allelomorph W 
to white, at least two unquestionable instances are known (and 
many more doubtful ones) in which white was again produced 
by a ‘reverse mutation’ of the allelomorph eosin (w*), which 
itself had been derived from the original white by mutation. 
A number of cases have been observed in which white has prob- 
ably arisen independently from the normal gene W (Lancefield, 
’18), but in none of these hitherto has the identity of the new 
with the old white been conclusively established and the possi- 
4 We are indebted to Dr. C. B. Bridges for this information. 
