THE INSECTS OF MACQUARIE ISLAND—TILLYARD—BRUES—LEA. 21 
For this purpose, careful drawings were made of the position of the ocelli on the 
head, the arrangements of the crochets on the abdominal prolegs, and the chetotaxy 
of the prothorax, mesothorax, and fifth and ninth abdominal segments. The result 
is that the caterpillar is determined, without any doubt whatever, to be a Pyralid 
belonging to the sub-family Orambine, and therefore almost certainly a feeder on the 
tussock-grass that is so abundant on the Island. 
In this connection, the Collector’s note attached to tube No. Co. 5, in which 
the caterpillar was placed, runs as follows :—‘‘ Found on grassy tuft, near West Point, 
Macquarie Island. Feb. 9th, 1913. In 70 per cent. alcohol.” 
As there is a possibility that species of Crambine may be bred from caterpillars 
found on this or other sub-antarctic islands in the future, I think it best to give a full 
description of the insect, with special reference to those structures which determine 
its systematic position :— 
Description of Caterpillar belonging to the Sub-family CRAMBINZ. 
(Text-figs. 13-17.) 
Total length 13-5 mm. Shape, sub-cylindrical, the head of medium size, the last 
three abdominal segments tapering slightly. General colour of the preserved specimen, 
_ yellowish brown, with the head and prothoracic shield very dark brown, the spiracles 
blackish, and the pinacula brown, more or less conspicuously darker than the general 
body colouration. Sete dark brown or black, mostly very distinct. This colouration 
almost certainly indicates a fairly dark-green caterpillar when alive, with darker head 
and prothoracic shield, and fairly conspicuous markings, perhaps of olive green colour. 
(Text-fig. 13.) 
Fig. 13.—Caterpillar of a Crambine moth, found on tussock grass. (x 74.) 
Heud with the ocelli arranged as shown in Text-fig. 14. In the dorsal group, 
the first, second, and third stand well apart, forming an obtuse-angled isosceles triangle 
with the second ocellus at its vertex. The fourth dorsal ocellus is close up to the 
third, both being very close to the antenna. In the ventral group, the more anterior, 
