662 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. XXIV. 



b^. Outline not extremely abnormal 

 but the whole body curved, the 

 thoracic portion being bent down 

 from the abdominal part . . . . Arhopalince part, embracing the 



genera Pratapa, Creon Ops, 

 Tajuria, Rathinda, Cheritra 

 and, perhaps, also B.oraga 

 and Zeltus. 

 b. Shape normal . , . . . . . . Arha'palincs part, embracing 



tSurendra, Arhopala, Zezius, 

 Chliaria, Thaduka, Catapm- 

 eilma, Loxiira, Deudorix, 

 Bindahara, Virachola and, 

 probably, Mahathala. The- 

 clince as represented by 

 the genus Aphnceus as far 

 as is known. 

 B. — Last segment not dilated in any way . . Lycmnince 



There is little use in pursuing this pupal classification any further. Each 

 one will be described accurately in the subsequent life-histories of the 

 insects concerned. 



A classification of the larvae will be much ntore interesting but, of course, 

 it can only be based on the mature stage. The earlier stages tare some- 

 what different from the full-grown ones in many cases, chiefly in the nature 

 of the clothing of hairs, &c., though the shape, seemingly, never varies for 

 the same species. Many egg-caterpillars are born with simple hairs 

 which they afterwards may lose and none of them ever seem to possess the 

 star-shaped, flattened, disc-topped, &c., hairs that characterise the mature 

 specimens. The differentiation is best, in the first place, based on the 

 shape, secondly on the covering of the body in the way of hairs, &c. The 

 colour is, as a rule, very variable in the same species, in that the greens are 

 very liable to become brown, rose, or red according to the colour of the ve- 

 getable matter consumed ; green leaves, white flowers, red shoots and so 

 on. There are, however, larvee the colour of which always remains fairly 

 constant as, for example, those of Amblyp)odia anita, Ops, Creon, Bindahara, 

 Virachola, ^c. The pattern hardly ever varies. 



As regards the general shape it is, as before mentioned, more or less that 

 of a wood-louse normally and this characterizes the whole sub-family 

 Lyccenincs without exception. In these larvse the outline is always oval in 

 shape and, when the insect is at rest in an unalarmed state, the greatest 

 height is exactly midway between the two ends, the dorsal outline being a 

 gentle curve, the ventrum being flat. It is necessary always to consider the 

 shape in this state for nearly all these caterpillars can change their form 

 more or less and an excessively stretched specimen, especially when feeding, 

 does not present the same aspect as one that sits with its head drawn in 

 and front segments bunched in a position of alarm ; both, again differing 

 from the appearance it has when at rest. 



A. — Larva normal, longly oval in shape, the dorsal 

 outline more or less gently convex, highest in the 

 middle. 



a. Larva when full-grown never over 15 mm. in 

 length. 



«\ Larva bordered round the margin with a 

 dense fringe of long hairs. These always 

 have also some long subdorsal hairs to each 

 segment. 



