GENETICS: METZ AND BRIDGES 
677 
This scheme involves an obvious difficulty, in that it assumes the 
ehmination of what would probably be the most numerous and wide- 
spread of the three types — the normal or parent type — while the proba- 
bilities greatly favor the elimination of one or both mutant types. If we 
slightly alter the conditions of the case, however, the difficulty is re- 
duced. Supposing that instead of the two mutant races arising from the 
normal stock, only one (a) has this origin, while the second (b) is derived 
in turn from the first one, and that, instead of the two mutants being 
incompatible with each other, the second is incompatible with the normal. 
In this case the first mutant (a) acts simply as an intermediate step be- 
tween the two incompatible forms, normal and mutant (b), and if it 
is eliminated the other two stand as distinct species. Here we have a 
case which differs in no essential respect from the former, but which is 
less improbable in that it involves the elimination of one of the mutant 
forms instead of the more widespread parent stock. 
The same sort of scheme applies equally well to the case in D. melano- 
gaster, so far as illustrative purposes are concerned. In this particular 
instance it happens that one of the mutant characters (notch) is a domi- 
nant that cannot be obtained in a homozygous condition because of its 
lethal effect, and hence the mutant could never make a pure race; but 
there is no necessary connection between lethal action and incompati- 
bihty. 
In connection with these experimental data it is interesting to note 
two other lines of evidence, taken directly from wild ffies. First, with 
respect to the question of whether or not mutant races can survive in 
nature, it may be recalled that the possibility of such survival is prac- 
tically demonstrated by such cases as that described by Sturtevant in 
Drosophila repleta.^ Two forms of this fly are found existing side by 
side in the wild state, and one of them is a typical sex-linked recessive to 
the other. There is every reason to believe that one of these has arisen 
from the other by mutation, and bears the same relation to it that any 
one of the above mutants does to the normal of its species. If we sup- 
pose that either of these forms in D. repleta gives rise to a third form that 
is incompatible with the other, and then becomes eliminated itself, we 
have all of the necessary steps in the formation of a new species; and 
the case differs in no essential respect from the hypothetical ones outlined 
above. 
Another line of evidence (also from D. repleta) bears more especially 
upon the question of incompatibility in wild races. Here, instead of 
two wild forms that interbreed freely but are unlike in appearance, we 
have two varieties that refuse to interbreed, but are extremely similar in 
