i6o 
Psyche 
[September 
respect to the head, although their wings are only about .1 mm. long, 
about half as long as the hind tibiae (in fully winged individuals the 
wings measure .8-1.0 mm.). In the subapterous females the wings are 
further reduced only slightly, being .05-.07 mm. long, roughly one 
third the length of the hind tibiae. Yet this slight reduction seems to 
influence head structure profoundly, as though wing reduction had 
crossed a narrow threshold below which reduction in the organs of 
vision was essential. 
What is still more surprising is that the four types of females 
apparently never occur together (Table I). Although macropterous 
females occur throughout the range (except, so far as known, in zone 
A) , micropterous females are known only from zone A, subapterous 
females only from zones B and C, apterous females only from zones 
C, D, and D 1 . These zones are arranged somewhat concentrically, 
zone A being more or less in the center of the range, closed in by an 
elongate zone B, which is followed by zone D, comprising the more 
mountainous parts of the northern half of eastern California. Zone C 
(Arizona) is interposed between B and D 1 toward the south and is 
the only zone in which three types of females are known to occur; in 
the North, zones B and D are contiguous. Zone D 1 (Baja California 
and Nayarit) contains, like D, no short-winged females; nearly all 
females from this zone are apterous (7 alate females are known as 
compared to 62 apterous females). It looks very much as though the 
center of the range was the area of “greatest wingedness”, and as one 
passes to the periphery in any direction the incidence of wing reduc- 
tion increases. 
It is tempting to erect hypotheses in an effort to explain this inter- 
esting situation. As I have noted, the closest relative of perpusilla 
appears to be the Palaearctic formiciformis , a species in which the 
males are always macropterous, the females macropterous or brachyp- 
terous in roughly a 1 :i ratio. Actually these short-winged females 
are more properly called micropterous, since the wings are scale-like 
and reach only to the anterior end of the propodeum ; they are also like 
the micropterous females of perpusilla in lacking any substantial reduc- 
tion in head width, eyes, and ocelli. Presumably the ancestor of 
perpusilla entered North America via the Bering land bridge and 
spread down the west coast, where selection favored smaller size and 
greater winglessness. The San Francisco Bay area may have served 
as a refugium and point of radiation for the species. As populations 
spread out from the center, selection favored still further reduction in 
the wings. Through changes in the genetics of wing inheritance, the 
