GUBETIN^. 241 



Larva, on emergence, eats a hole through the top of the egg about equalling 

 one-third of its upper surface, and crawls out. The empty shell has a close superficial 

 resemblance to an echinous shell. The larva is pale ochreous in colour, and is furnished 

 with long stout hairs, of which a sub-dorsal series is on each side, with one long hair 

 springing from the apex of each tubercle ; there are besides other lateral series, and 

 numerous hairs projecting forwards in front of the head and backwards over the anal 

 segment. The full-grown larva is the most beautiful known to me among the 

 Lycsenidae. It is about \^ of an inch in length, of the exact shade of the green of the 

 leaves on which it feeds, the head pale ochreous and more completely hidden than in 

 any larva I have ever seen, and even when the animal eats it is not visible from above, 

 the second segment entirely enclosing it. The second segment is half as wide as the 

 following, the third, fourth and fifth rapidly and progressively enlarging, the next 

 three segments as rapidly decreasing in size, the remainder sub-equal. The second 

 segment is quite unmarked, the third to the thirteenth have a sub-dorsal series of 

 short oblique pale yellowish-green lines, between which the ground colour is paler than 

 the rest of the body ; there is a dark green dorsal line ; on each side of the ninth 

 segment there is a prominent pure dead white somewhat diamond-shaped mark. The 

 twelfth segment bears two most extraordinary structures, which consist of two 

 diverging cylindrical rigid pillars arising from the sub-dorsal region and of a pale green 

 colour. When the insect is touched or alarmed, from each pillar is everted a deep 

 maroon tentacle as long as the rigid pillar, bearing at its end two long parti-coloured 

 hairs, the basal third of each hair being black, the upper two-thirds white. I made 

 the following notes when breeding this species : — " June 24th, eggs laid. June 27th, 

 larvae hatched ; placed on young leaves of Derris scandens, Benth. (they also eat Heynea 

 trijuga, Roxb., in Calcutta). June 28th, larvae grown enormously. June 29th, changed 

 skins ; when first hatched they are pale ochreous and hairy, just before changing their 

 skins they turn green, and are without any protuberances ; after their first moult they 

 are quite naked, green, with two black horns on the twelfth segment. Before 

 changing they fasten themselves with silk to a leaf on the underside near the midrib. 

 July 1st, changed skins again, appearance much as before, only that they have a kind 

 of white saddle-mark on the ninth segment, horns the same colour as the rest of the 

 body. They devour their old skins. July 3rd, changed their skins again, their colour 

 now chocolate-brown marked with green and white. July 5th, probably changed their 

 skins again to-day, colour green. July 7th, changed lo pupae. Before changing they 

 lose all their bright colouring and markings, becoming a dull uniform green of the 

 same shade as the leaves of their food plant. Butterflies emerged in three or four 

 days." 



Pupa almost as curious an object as the larva, reminding one of a lump of green 

 jelly. Its colour is pale transparent light green, the wing cases being bluish-green and 



VOL. VIII. 2 I 



