INTRASPECIFIC  VARIABILITY.                               41

VIII. INTRASPECIFIC VARIABILITY.

The species of Drosophilinse, as they occur in the wild state, are
as a rule liot strikingly variable. They do vary in size or in intensity
of color as a result of the amount of food obtained in the larval stage.
Especially striking examples of this sort of modification are furnished
by D. cardini and D. funebris, in both of which breeding experiments
have shown that the variations are not inherited. But real inherited
variations seem not to be common. Slight variations in the abdominal
pattern of such forms as D. busckii, D. transversa, and Mycodrosophila
dimidiata are probably of a genetic nature, and the variations in the
intensity of the slight "trident" mark that is often present on the
mesonotum of D. melanogaster have been shown to be inherited. I
have studied an inherited variation in mesonotal pattern in D. repleta
that occurs in wild flies (Sturtevant 1915). A number of investigators
have shown that the occasional extra dorsocentral bristles found in
wild stocks are in some cases heritable. Miss Hoge (1915) found
inherited differences in the number of teeth in the tarsal combs of wild
D. melanogaster males. Although it is not uncommon to find indi¬
vidual wild specimens that contain mutated genes, I know of no
cases in the group other than those just mentioned in which a variation
is established in nature and can be found persisting side by side with
the parent stock.

There are certain measurable characters in which the various species
show variability, which may be heritable or not. Since most of these
characters are used taxonomically, it becomes of interest to know just
how variable they are, and what are the limits for different species.
Unless this information is at hand it is not possible to judge as to their
value as diagnostic characters.    The following data bear on this point:

The possession of two pairs of dorsocentral bristles is characteristic
of the Drosophilinse. Four pairs are present in Blcesochcetophora and
Dettopsomyia; one in Acletoxenus and Drosophila superba, and only
one large pair in Mycodrosophila; aside from these forms two pairs
occur in all the forms known to me. The same number is to be found
in many other Acalypterse, scattered through most of the subfamilies.
It is especially common in the Sciomyzinse, Geomyzinse, and Mili¬
chiinse, but in none of the other subfamilies is it as usual as in the
Drosophilinse. In species where two pairs are the rule, however,
exceptional individuals can often be found. I have examined a large
number of specimens of several species for this character, with the
result shown in table 3. The numbers given refer to the total number
of dorsocentrals present. Two pairs, for example, equal "4." The
flies recorded in these tables were all bred on banana agar under
laboratory conditions. The numbers from each stock are based on
examinations of individuals from several cultures made at different