128 
SPOLIA ZEYLANICA. 
angle of the roof was in contact with the opposed blade of the leaf, 
and the larva then cut out a notch at the front end (Fig. 4). This is 
the commencement of the fringing of the openings of the tube which 
usually takes place. Three hours later the folding was complete, 
the edge of the roof was sealed down upon the blade, a notch had 
been cut at the hinder border, and the larva had begun to 
feed from the front opening (Fig. 5). This was the third house 
in the life-cycle of this individual. On the same day another larva 
built a house also 1 inch long and 1J inch from the cut end of the 
leaf. After bending over the cover, it began feeding from the 
opposite edge of the leaf before fringing the openings. After having- 
fed, it resumed the work of shortening the trabeculae to close down 
the edge of the roof. There is thus both method and variety in their 
manner of working. 
The larva, to which figures 2-5 refer, remained in that shelter, 
feeding at intervals from both openings, for four days. Then, 
having exhausted the feeding-area within reach of its domicile, it 
left the latter and cut it off as described above. It then moved to 
another spray and commenced its fourth house, in which it remained 
for another period of four days ; at the end of this time the tube- 
shelter, about 1J inch long, was suspended by a midrib-stalk 2 inches 
in length. The stalk had been partly broken, the two parts held 
together b}^ webbing only. The caterpillar had to pass over this 
portion in order to get to the base of the stalk for the purpose of 
cutting it off. It then moved to another spray, where it made the 
usual floor of webbing, and began to draw the leaf-halves together 
without cutting, by weaving two strands, 1J inch apart, from one 
side to the other. The larva at the time of making its fifth house 
is nine-tenths of an inch long, colour a milky green with dark green 
dorsal line, dull whitish head and white-rimmed pygidium. A 
larger feeding-area is provided for, the front strand or silken beam 
of the house being laid down at a distance of 5J inches from the leaf- 
end. It occupied the fifth house for twenty-four hours only, when it 
quitted the shelter and cut it off as before. It then made the sixth 
and last house, this time binding two sprays together. Two days 
later it had disappeared, but was subsequently found on the ground 
inside its tube ; it was placed in a chatty with moist earth and 
leaves, and the butterfly emerged a fortnight later. 
When the caterpillar has finished feeding and has achieved its 
full growth and is ready to pupate, it remains in the tube which is 
hanging suspended from the leaf by the bare midrib (Fig. 6). Eventu¬ 
ally it bites through the fibre more or less flush with the basal end 
of the tube, the latter, with the larva inside, falling to the ground, 
leaving the suspending fibre attached to the leaf. The distal or 
hinder end of the last nest is fringed as usual, and the tassels are held 
together by webbing. The proximal or anterior end is plain, and it 
is from this end that the larva projects the forepart of its body, 
