THE MANNER OF OCCURRENCE OF MUTATION 451 
bility been at the same time quite excluded that the new white 
may have arisen from the old through some contamination of 
the food. In the following case, however, the answers to these 
questions admit of no doubt. 
Here, as in the cases of écru and ivory, a stock heterozygous 
for eee mutant factors had been maintained for many 
generations by constantly repeating ie same cross; in this 
/ 
instance the cross was of - males by & rr female (H’ represents 
the dominant factor ‘hairless,’ which is lethal when homozygous; 
tt represents the recessive ‘tilted wings;’ both are in chromosome 
III). No eye color except the normal red had ever appeared in 
any of the flies of this heterozygous stock, until, after several 
months of such crossing, a single hairless male was found which 
had a red left eye and a white right eye. This male was tested 
by mating it to tilted sisters. Half of its offspring were hairless, 
and the rest were tilted, showing that the mutant male had had 
the composition = and must have been derived, without con- 
‘tamination, from the heterozygous stock. All the offspring 
had both eyes red. These were then mated together in a large 
mass-culture and produced in the next generation females all of 
which had both eyes red, and males half of which had both eyes 
red, and the rest of which had both eyes white. The new factor 
for white therefore must be recessive and sex linked. Further- 
more, it was evident that the factor was not one which affected 
only one of the eyes, but it made both the eyes white equally, 
provided it was contained in them. The difference between 
the two eyes of the mutant grandfather must therefore have 
been due to a difference in the factors they contained—the red 
eye must have contained the original normal factor and the 
white eye the mutant factor. Some, at least, of the germ cells— 
all of them, so far as the evidence went®'—contained the mutant 
factor like that in the white eye. 
5 Tt has been found by Morgan and Bridges that in flies whose epidermal 
parts are mosaic (gynandromorphic), the gonad is not mosaic; that is, the germ 
cells are all genetically like one or the other of the portions of the epidermis: 
