470 H. J. MULLER 
duces the mutation must be extremely localized in its site of 
application. The previously known fact that usually only one 
locus mutates, and not the neighboring loci of the same chro- 
mosome, might have been explained on the supposition that 
the influence at work was chemically specific, only affecting a 
gene of a given composition, but in that case it might be ex- 
pected that the two similar genes in homologous chromosomes 
would both be affected. The fact that they are not shows that 
the immediate cause of the mutation is not a diffuse influence 
existing throughout the body, the cell, or even the nucleus; the 
mutation is due to an event of such minute proportions, so 
circumscribed, that it strikes only a single one of two near-by, 
similar loci in the same nucleus. 
Emerson (711) arrived at a similar result in corn, in the case of 
his mutable locus for variegation. In portions of the plant in 
which the dominant mutation to red was somatically visible, 
only half of the germ cells carried the mutated factor, indicating 
that the mutation had occurred in only one of the two homol- 
ologous chromosomes.!! 
In view of this conclusion, the oft-suggested possibility 
of artificially influencing the kind of mutation that occurs 
(ef. Stockard, Tower, MacDougal, Kammerer, Guyer) would seem 
to recede indefinitely, unless some unique method is found which 
does not merely consist in an acceleration or intensification of 
the ordinary process of mutation. 
SUMMARY 
1. Three new mutations of the gene W in the X chromosome 
of Drosophila melanogaster have been described. They have 
given rise respectively to: 
a. Eeru, causing the eyes to be of a light straw color, not 
much darker than ‘white.’ 
11 Baur, too, reached this conclusion, through his finding that mosaics show- 
ing a recessive character cannot result from plants originally containing the 
dominant gene unless the latter was heterozygous, but, as previously pointed 
out, he may not have been dealing with real mutations. 
