224 
PACIFIC SCIENCE, VoL IX, April, 1955 
center of this group is undoubtedly in New 
Guinea where argenteiventris (Theobald, 1905), 
atra (Taylor, 1914), and microlepis (Edwards, 
1927) are found. Still more remotely allied is 
the atripes section of Mimeteomyia , which is 
represented in Australia (N. S. Wales, Queens- 
land and Northern Territory) by atripes (Skuse, 
1899) and punctolateralis (Theobald, 1903), 
and is apparently absent from New Guinea or 
at least not represented by the atripes complex. 
This particular complex is of interest here for 
it has a member, T. solomonis (Edwards, 1924) 
in the Solomon Islands. Both the atripes com- 
plex and the caledonica complex utilize arti- 
ficial containers to some extent for breeding 
and it is possible that they arrived or spread 
in Melanesia in recent times through acci- 
dental human transport. On the other hand 
their mutual exclusion in Melanesia is against 
such a simple interpretation and favors an 
earlier origin from continental Australia, at 
least for the caledonica complex. This does not 
imply a continuous land connection, for it is 
quite evident that mosquitoes are capable of 
crossing considerable expanses of water 
through natural means of dispersal. The pres- 
ent distribution of mosquitoes of the genus 
Tripteroides in southern and eastern Melanesia 
is peculiar and analogous to the distribution 
of the human racial stocks in this area, al- 
though it does not follow the same plan. 
The caledonica complex has apparently come 
from Australia and has more recently ex- 
tended to the northeast to Rotuma Island. 
T. purpurata , the only other eastern outpost 
of the genus, has undoubtedly come by way 
of the Solomon Islands to its present position 
to the southeast, in Fiji. This has resulted in 
a complete crossing of the paths of dispersal 
of these two different lines. Unfortunately we 
have no records of mosquitoes of this genus 
from the Santa Cruz group which may have 
served as a stepping stone for both dispersals. 
It is not beyond the realm of possibility that 
the extensions to Rotuma and Fiji have been 
made in recent times through human agency 
by transport in canoes. Considering the nu- 
merous movements of the Melanesians and 
Polynesians in this area, it is surprising that 
more species of mosquitoes capable of being 
transported in such a way have not spread 
more widely. 
Within the caledonica complex itself the 
geographic and ecological relations are also 
of interest. In New Caledonia, adjacent is- 
lands, and the Loyalties two species of the 
complex are represented, caledonica in Nepen- 
thes pitchers and melanesiensis in tree holes, 
bamboo, artificial containers and "palm- 
bracts.” All the larval ecological types of 
melanesiensis in this area are remarkably sim- 
ilar when compared with the parallel forms 
in the New Hebrides, but one of them, the 
"palm-bract” race, may prove to be a distinct 
species. In the New Hebrides two distinct 
species are also recognized, folicola restricted 
to leaf axils of living plants on Espiritu Santo, 
and melanesiensis , breeding in a wide variety 
of habitats throughout the New Hebrides. 
The ecological and geographic forms of me- 
lanesiensis in this area are much more numer- 
ous and exhibit much greater divergence. A 
single island, as Espiritu Santo, may have as 
many as four distinct larval types, each re- 
stricted to a distinct habitat. Furthermore 
parallel ecological types on different islands I 
often exhibit striking differences which be- 
come more extreme toward the northern por- 
tion of the range. Finally, some 600 miles to j 
the northeast of the New Hebrides, on the 
small island of Rotuma, we find rotumana , a j 
species strongly differentiated from the rest of 
the complex in the adult stage, but very \ 
similar to melanesiensis in the larva. 
To summarize: On the basis of present 
knowledge of the caledonica complex, it ap- 
pears that it was derived from continental 
Australia and first reached New Caledonia, 
probably by means of intermediate islands no 5 
longer in existence in the region of the Ches- 
terfield group (Routhier, 1953: 244-246). In 
New Caledonia, the uniform southern race of 
melanesiensis represents the original stock from 
which were derived at an early date the aber- 
