Uranotaenia in Solomon Islands — Belkin 
scription of U. atra (1905: 114; 1907: 563- 
564), but the remaining nominal species are 
evidently applicable to the same species or 
species complex in different portions of its 
range. It is possible that the poor original 
condition of the type female of U. atra is 
responsible for Theobald’s failure to mention 
the characteristic ornamentation of the head, 
pleura, and abdomen and the absence of 
white scales on the wing. On the other hand 
it is also possible that U. atra Theobald rep- 
resents a species distinct from Edwards’ con- 
cept and one which has not been recognized 
since the original description. Theobald’s de- 
scription fits U. annandalei Barraud, 1926, 
much better than any other species, but U. 
annandalei has not been reported to date from 
New Guinea although it is known from the 
Philippines (Baisas, 1935: 64). For the present 
I prefer to follow Edwards’ synonymy and to 
retain the name U. atra Theobald, 1905, for 
this species. 
The reportedly wide geographical distribu- 
tion of U, atra in the Oriental and Australasian 
regions leads one to suspect that it may break 
up into a number of geographical subspecies 
or that it may be a complex of related species. 
The more or less detailed descriptions of the 
adults from India, Ceylon, Malaya, Philip- 
pines, and Australia agree quite well in gen- 
eral features with the specimens from Guadal- 
canal. I find small differences, apparently con- 
stant, in the foreleg of the males from the 
Solomons when compared with the descrip- 
tions and figures of Philippine and Indian 
material. The larvae from Guadalcanal are 
quite distinct from those described from India 
(Barraud, 1934: 74) and Netherlands East 
Indies (Bonne-Wepster and Brug, 1939: 1231; 
Brug, 1924: 441-442) in the length of the 
siphon (Solomons 4:1, others 3:1 or less), 
and in addition there appear to be constant 
differences in chaetotaxy which cannot be 
determined positively without a study of large 
series. Baisas (1935: 65) states that the Philip- 
pine U. atra breeds in forest streams, an eco- 
logical niche different from that occupied by 
331 
this species in other portions of its range 
where apparently it is confined to coastal 
swamps and to crab holes. The pupal stage 
has not been previously described. 
Although it appears likely that U. atra 
splits at least into several geographical sub- 
species, I consider that our present knowledge 
is insufficient to recognize these forms. As 
pointed out above, the name U. atra is asso- 
ciated with the complex under consideration 
on rather inconclusive evidence. As far as I 
have been able to determine, neither U. atra 
sensu Edwards (other than the single female 
of Theobald) nor U. annandalei have been 
collected on New Guinea. Until such collec- 
tions are made and studied and the types 
compared again, it would be unwise to in- 
troduce new names in a species complex 
which has more names already associated with 
it than any other Uranotaenia. 
It appears that U. atra is most closely re- 
lated to the tibialis- gtou.-^ and to other species 
with modified fore tibia and tarsus, as well 
as to U. annandalei which shares with this 
group the peculiar leaf-like development of 
the antennal hairs of the larva. The adults of 
these species have very broad scales on the 
pleura arranged in broad patches and usually 
showing very strong azure-blue iridescence. 
In the Oriental and Australasian regions the 
known pupae of all these species have ex- 
tremely elongate slender respiratory trumpets, 
greatly elongated hairs 4-IV, V, and a peculiar 
development of hair 8-C. On the other hand 
U. atra is quite distinct from all its nearest 
relatives in having a larva with a wide head, 
poorly developed spike hairs, very short hairs 
1-C, unmodified antenna, and very short anal 
gills. 
Biology 
The three larval collections on Guadalcanal 
were all made in slightly brackish-water pools 
immediately back of the beach in open sunlit 
areas. Thorough collections in other Urano- 
taenia breeding areas never revealed this spe- 
cies. It appears to be restricted on Guadalcanal 
