1929] 
Stylopized Vespidx 
279 
isolated insular regions must be very remote, much more 
so than that of their host. Yet the first record of an Ody- 
nerus from Bermuda, the holotype and only known speci- 
men of Odynerus bermudensis Bequaert, bears a female 
Strepsipteron. Polistes fuscatus var. aurifer, which was 
introduced into the Hawaiian Islands fairly recently, 
through the agency of man, carried its parasite with it. We 
list above several stylopized specimens from various islands, 
more or less remote, of the West Indies, the East Indies, 
and the Canary Islands. 
Since the powers of flight of the host are impaired by 
stylopization, and female hosts are frequently rendered 
sterile, it is probable that the introduction of Strepsiptera 
into distant regions is brought about in the triungulinid 
stage. Several of these minute larvae could cling to the 
body of a perfectly healthy host without seriously burden- 
ing it. Brues, indeed, has already reported (1924) the oc- 
currence of triungulins, presumably of the meloid Horia 
maculata, on a bee (Xylocopa transitoria) from the Galapa- 
gos Islands. 
The Problem of Host Specificity 
It seems to have been generally assumed that Strepsip- 
tera obtained from different species of hosts are themselves 
specifically distinct. Pierce (1909, p. 70) stated that as a 
rule species of Strepsiptera are confined to single species of 
hosts; and in his classification used tribes as divisions of 
convenience to unite genera of parasites infesting different 
families of hosts. Following a decision to unite the vespid 
groups Leionotus, Ancistrocerus, etc., with Odynerus, the 
same author (1911, p. 498) combined the genera Leiono- 
toxenos and Pseudoxenos parasitic on those groups. Many 
species and several genera in the Strepsiptera are based 
solely on the unsatisfactory characters of the female to- 
gether with the name of the host, which latter one some- 
times feels is the more valuable part of the description. 
It is not our intention to discuss here the validity of 
this tacit assumption of host specificity in the Strepsiptera, 
but some points arising from our redetermination of several 
