98 VARIATION, DISTRIBUTION, AND EVOLUTION OF THE GENUS PARTULA. 
of these offspring would repeat the parental mode of coloration; apparently the 
only conclusion to be drawn from the data is that the distinctive colors develop 
very much later. 
For the case of vespera the Barrigarda statistics serve best (table 49): 79 out 
of 80 young taken from such adults belong to type a; the remaining individual is 
brown-corneous, and it suggests the cross-breeding of vespera with a dark-colored 
kind, probably phea. From the light colors of very young vespera adolescents 
we would expect the embryonic shells to belong to type a, as they do. Now, in the 
Barrigarda adult series, this class amounts to about one-quarter of all, while 
type a young from vespera progenitors number more than half of the total embry- 
onic series, 79 out of 142; presumably the excess of light young beyond the number 
required to recruit the parent class were destined to become adults of other unde- 
terminable kinds. Thus vespera seems to intercross with contrasted classes. 
Finally, it transpires that all of the 7 young taken from the marginata of the 
Dungcas locality display the type c color. As the peculiar adults constitute about 
one-sixth of the adults collected, it is interesting that their brownish young form 
very nearly the same proportion of the total embryonic series of this locality. The 
data are unfortunately too scanty to deserve further discussion. 
Despite the difficulties imposed by the late development of the distinctive 
color characters of the gibba classes, and despite the errors which have undoubtedly 
been made in some instances in assigning the embryonic young to their true classes, 
there remains a substantial body of proof that the several color-classes of sh present 
species can interbreed mutually. 
