238 
PACIFIC SCIENCE, Vol. IX, April, 1955 
Discussion 
T. melanesiensis exhibits more striking varia- 
tion in the larval stage than any other species 
of mosquito that I have seen. Each type of 
breeding place appears to have a peculiar form 
and every island has morphologically dis- 
tinguishable populations. It would appear 
therefore that we are dealing with numerous 
ecophenotypes or possibly ecotypes or eco- 
species as well as geographical races. The 
material at hand does not permit full analysis 
of this complex but indicates certain distinct 
trends. 
The most clearly marked of all the forms of 
melanesiensis is a tree hole breeding form found 
in the northern and central New Hebrides. 
For this reason and because it has been de- 
scribed and figured by Buxton and Hopkins 
(1927: 74-78) I am selecting it as the typical 
race. It is characterized chiefly by the follow- 
ing characters: hair 4-X about as long as the 
saddle and usually with three or more branches, 
hair 1-X with three or more branches, siphon 
index 3.5 to 6. Outside of the type locality 
of Espiritu Santo (Fig. 4,5^) I have- seen 
specimens from Aore Island (Fig. 5c) and 
Efate (Fig. 5 b). In all probability the speci- 
mens collected by Buxton on Malekula and 
Pentecost also belong to this race. While there 
appears to be relatively little variation within 
populations of this race on the same island, 
there are striking differences between those 
from different islands, particularly in regard 
to the length and shape of the siphon. The 
typical race of melanesiensis does not appear 
to be entirely restricted to tree holes for spec- 
imens collected in foul water in a cold storage 
house on Espiritu Santo exhibit all the char- 
acters of this race. 
All the remaining forms of melanesiensis have 
hair 4-X longer than the saddle and usually 
double or single, hair 1-X usually double, and 
the siphon shorter, index 3.5 or less. For the 
present all these forms are considered as 
atypical races of melanesiensis (Fig. 5a, e-h). 
With additional material it may be possible 
to characterize several races and possibly dis- 
tinct species in this complex. 
In the northern New Hebrides (Espiritu 
Santo, Aessi and Tutuba) atypical melanesiensis 
have been collected in coconut shells (Fig. 
5/), bamboo (Fig. 5a), and cacao pods. In 
addition, at least one collection from tree 
holes on Espiritu Santo contains both typical 
and atypical melanesiensis without any inter- 
mediate forms. The atypical forms in this 
collection resemble closely the bamboo and 
coconut types and have a much shorter siphon 
as well as fewer and shorter branches in the 
stellate hairs than the typical race. Unfortu- 
nately it is not known whether or not the 
two types of larvae came from the same tree 
hole. The larvae from the other three types 
of habitats are generally similar but each has 
its peculiar morphological features. Since the 
number of collections is small, it is impossible 
to determine how constant these differences 
are. I have seen a number of larvae from 
Espiritu Santo which have the siphon longer 
than the other atypical melanesiensis but un- 
fortunately no information is available as to 
their breeding place. These larvae have none 
of the diagnostic features of the typical race. 
Although melanesiensis has been reported from 
artificial containers on these islands (Perry, 
1946: 14) none of the specimens I have ex- 
amined are recorded as being collected in 
such habitats. It is possible that the above 
mentioned larvae without habitat data are 
from artificial containers. It should be noted 
that Buxton (Buxton and Hopkins, 1927: 76) 
collected only typical melanesiensis in the 
northern New Hebrides and only in tree holes. 
On the other hand during World War II 
atypical melanesiensis were collected more fre- 
quently than the typical. It is not beyond 
the realm of possibility that the atypical mela- 
nesiensis are not endemic to these islands but 
were introduced at that time. 
In the central New Hebrides collections 
were seen only from the island of Efate. In 
addition to the typical melanesiensis from tree 
holes, larvae have been collected only in a 
