330 
PACIFIC SCIENCE, VoL VII, July, 1953 
merit VI: hair 0(1), l(7b), 2(1), 3(2, 3f), 
4(3b), 5(1, 2b), 6(7b), 7(4b), 8(2, 3f), 9(2b), 
10(2f), ll(2b), 12(2f), 13(16, 17d), 14(1, 2f). 
Segment VII: hair 0(1), 1(6, 7b), 2(1), 3(6, 
9b), 4(2f), 5(2, 3b), 6(5b), 7(2f), 8(5, 6b), 
9(1, 2f), 10(2b), ll(3b), 12(1), 13(5b), 14(1). 
Segment VIII: comb plate strongly sclero- 
tized, moderately pigmented, ornamented 
with numerous imbrications but lacking spi- 
cules except apically; comb scales 7-8, ven- 
tral ones smaller, sharply pointed, with short 
lateral fringe extending to about 0.75 as fig- 
ured; hair 0(1), 1(4, 5b, outer branches short- 
er), 2(2, 3f), 3 (4b), 4(2f), 5 (7b, outer branches 
shorter), 14(1, unusually strong). Siphon: as 
figured; length 4.0 median width; moderately 
pigmented, imbrications as on comb plate 
but all without spicules; pecten extending to 
0.5, teeth 11-12, with lateral and apical fringe 
as figured; valves as figured, very darkly pig- 
mented, ventral valve short; hair 1(9, lOb, 
outer branches shorter, basal expansion short 
and asymmetrical, barbs absent), 2(1, unusu- 
ally long), 3-5 (only two setal rings visible), 
6(1, strong, longer than valve), 7(1, short, 
slender), 8(1, moderate, shorter than valve), 
9(1, long but very slender and attenuate), 
10(hairless setal ring), 11(1, well developed), 
12(1, minute), 13(1, strong, twisted at base, 
about length of valve). Segment X: saddle 
lightly pigmented, darker on the apex dorsal- 
ly; median width 0.8 of length; imbrications 
much fainter than on comb plate except cau- 
dally; caudal margin without long teeth, but 
with several rows of small spicules arising 
from imbrications; gills short, stubby, round- 
ed apically, about 0.3 of saddle length; hair 
1(7, lOb, about 0.45 of saddle length), 2 (2b, 
about 3.2 of saddle length), 3(2b, about 3.8 
of saddle length), 4a(2b, about 1.3 of saddle 
length), 4b (2b, about 2.0 of saddle length), 
4c(l, about 2.4 of saddle length), 4d(l, about 
1.6 of saddle length), 4e(l, about 0.25 of 
saddle length). 
Variation 
. The adults of JJ. atra show considerable 
variation in the amount and intensity of blue 
scaling on the head and the thorax as do the 
other ornamented species, but in all the spec- 
imens the azure-blue iridescence is very dis- 
tinct. The white scaling of the abdominal 
tergites is also subject to considerable indi- 
vidual variation but is never absent. 
The variation in the chaetotaxy of the im- 
mature stages is summarized in Tables 1 and 
2. The range of variation of branching in this 
species appears to be less than in any other 
species known from the Solomons, but this 
may be due to the fact that only three collec- 
tions were made and from only two localities 
on Guadalcanal. One larval specimen shows 
a duplication of hair 1-C on the left side, 
another has 2-A placed at about the middle 
of the antenna instead of the apex. 
No information is available on geographical 
variation in the Solomon Islands. A single 
male from Munda, New Georgia, agrees in 
all respects with the material from Guadal- 
canal. 
Specimens examined: 27M; 28F; 81P; 27L. 
Individual rearings: 16 larval, 1 pupal. 
Taxonomic Discussion 
The application of the name Uranotaenia 
atra to the ornate species with the characteristi- 
cally modified male front tarsus, entirely dark 
legs, and largely dark abdomen rests on Ed- 
wards’ identification (1913: 238) of Theo- 
bald’s ''mouldy and rubbed and scarcely rec- 
ognizable” single female type specimen of 
f/. atra as conspecific with IJ. ceylonica Theo- 
bald, 1910, another species described from a 
single female specimen. Later Edwards (1922: 
460) synonymized U. coeruleocephala lateralis 
Ludlow, 1905 , based on four females from 
the Philippines, and JJ. cancer Leicester, 1908, 
based on a male and a female from the Malay 
Peninsula. Still later (1924: 357-358) Edwards 
synonymized JJ. propria Taylor, 1914, based 
on a male from Queensland, and U. cairnsensis 
Taylor, 1919, based on two females also from 
Queensland. 
The species as understood by Edwards could 
hardly be recognized from Theobald’s de- 
