Uramtaenia in Solomon Islands — Belkin 
is needed to characterize the populations de- 
scribed and discourage the publication of 
much valuable material which could easily 
be incorporated in the original composite 
description. Furthermore, this practice tends 
to the selection of extremes for description 
and encourages hasty decisions. In mosqui- 
toes it is practically impossible to find a spe- 
cimen in a condition to show all the important 
features, particularly in reared material. When 
males are selected for holotypes, microscopic 
preparations of genitalia and appendages de- 
stroy some diagnostic characters. 
To alleviate in part these disadvantages 
without undue repetition in the text, I have 
resorted to comparative tables and composite 
figures. In the case of the latter the same 
arguments may be used in favor of figures 
of types as in the case of descriptions. Com- 
posite figures based on a study of variation 
in the new form are of considerably greater 
importance in the identification of other spe- 
cimens than a graphic representation of char- 
acters already described in detail for the type 
specimen, whose departure from the norm 
can be determined by comparison with the 
composite figure. 
I have made the descriptions, tables, and 
illustrations as complete as time and the mate- 
rial allowed. They may appear unnecessarily 
detailed and repetitious. Elsewhere it has been 
pointed out (Belkin, 1951: 678) that, in the 
immature stages of mosquitoes, culicidolo- 
gists have largely restricted themselves to a 
few diagnostic characters, particularly in non- 
anophelines, and have neglected a multitude 
of other morphological features which appear 
to be at least as useful. In a revision such as 
this, limited both geographically and taxo- 
nomically, it is not possible to determine the 
important characters and leave out the non- 
essential features since it is essentially a pre- 
liminary attempt and no comparable pub- 
lished study is available at the present. It is 
my intent to continue such studies as it is 
becoming evident, even in such a limited 
field, that in addition to those already in use 
313 
excellent group characters are to be found in 
the thoracic and abdominal chaetotaxy of the 
larvae and the cephalothoracic chaetotaxy of 
the pupae. As every species treated here be- 
longs to a distinct species group and adequate 
material of related forms from adjacent areas 
was not available for study, little can be said, 
as far as specific diagnostic characters are con- 
cerned, except that it appears that the larval 
and pupal chaetotaxy may be of considerable 
value if it is studied quantitatively as well as 
qualitatively. For the present I have had to 
use a very few disjunctive characters to dis- 
tinguish the new species described here be- 
cause of the paucity of material and inade- 
quate descriptions. Whether the forms de- 
scribed here are species or subspecies cannot 
be decided without further comparative stud- 
ies. I prefer to consider them for the present 
as distinct species. In any event, I hope that 
it will be possible for other workers to recog- 
nize the forms described here and also to test 
the usefulness of some of the characters in 
the diagnosis of related species. 
In the descriptions, strictly generic char- 
acters are not repeated since they are indicated 
under the discussion of the genus at the be- 
ginning of the paper. In the larva the mouth- 
parts were not studied at all, and in the female 
the genitalia were also neglected. Otherwise 
all characteristics previously described were 
investigated. 
In the tables on the setal branching, the 
mean and other statistics were not determined 
as the samples were not adequate to provide 
significant figures except in cases where the 
desired information was evident without com- 
putations. It appears that mathematical treat- 
ment of multiple characters of chaetotaxy may 
prove a valuable tool in diagnosis of related 
species. For the present I have confined the 
data to an indication of the extremes of varia- 
tion and of the "usual number” of branches 
in the order of frequency. The latter is the 
mode alone or the mode followed by the next 
class (or classes) if it is one half or more of 
the mode. In general the "usual number(s)” 
