18 
Fig. 6. Loxopleurus rotumanus Gressitt, n. sp. 
segments combined; third distinctly longer 
than second or fourth. Prothorax nearly as 
broad as elytron, seven-ninths as long as 
broad, constricted, but not very distinctly 
collared, near apex; disc shiny, sparsely and 
shallowly punctured, obliquely grooved at 
side; median basal process emarginate. Scutel- 
lum suboblong, twice as long, as broad, 
smooth. Elytron with ten rows of fine punc- 
tures, including sutural row, besides an ob- 
lique row between sixth and seventh regular 
rows; outer three rows in distinct grooves, the 
last two complete. Ventral surfaces in large 
part finely and not very closely punctured, 
and with short erect pale hairs; metepisternum 
finely rugulcse; last abdominal sternite densely 
punctured except in the large central cavity. 
Length 3.5 mm.; breadth 2 mm. 
rotuma: Holotype (Bishop 2402) Jolmea, 
Rotuma Island, north of Fiji Islands, Aug. 
13, 1938, H. St.John. 
Differs from L. vitiensis Bryant in being 
stouter, with shorter antenna, more punctate 
pronotum and elytron, and with different 
PACIFIC SCIENCE, Vol. XI, January, 1957 
markings, having the pronotum with a large 
central pitchy area and the elytron pale with 
broad dark borders except along suture. 
17. Cryptocephalus fraterculus Chapuis 
Cryptocephalus fraterculus Chapuis, 1876, Soc. 
Ent. de Belg., Ann. C. R. 18: XCI (Fiji). 
This species was described from Fiji, but 
has not been identified from the material at 
hand. 
Subfamily EUMOLPINAE 
The study of the Fijian members of this 
subfamily has proved most difficult. To adapt 
the Fijian material to most of the previously 
used classifications would have required the 
erection of a number of new genera or tribes. 
On the basis of the material at hand, it would 
appear that many characters used in the past 
as tribal characters do not deserve that value. 
Therefore, a compromise has been made, 
which cannot be considered a final or satis- 
factory treatment, but one that may serve until 
the faunae of the island groups to the west 
are better known in this family. In the main, 
the system proposed by Chen (1940: 486 ff.) 
has been followed, though it does not quite 
satisfactorily solve the problem. It has in- 
volved a reduction from eight or more tribes 
to three. Some of the old tribal names are 
inserted in parentheses in the text. 
KEY TO FIJIAN GENERA OF EUMOLPINAE 
1. Apical tergite of abdomen with a 
median longitudinal groove; gen- 
erally glabrous above; tarsal claws 
appendiculate; middle and hind ti- 
biae not emarginate preapically 
(Colaspoidini) 2 
Apical tergite of abdomen lacking a 
median longitudinal groove; tarsal 
claws generally bifid; middle and 
hind tibiae generally emarginate pre- 
apically 3 
2(1). Elytron regularly seriate-punctate; 
body just over one-half again as long 
as broad Eucolaspis 
