THE COMMON B UTTERFLIES OF THE PLAINS OF INDIA . 661 



smooth, in a few cases there are long fleshy teeth along the dorsal line and 

 occasionally, on the dorsoventral margin also. The shape is liable to 

 variation, in that the oval form is sometimes constricted or narrowed 

 into a waist about segments 7-10 {Amblypodia, Tajuria cippus) and, 

 then, also, the body is lowered in the same region. In most cases 

 there are secondary organs on segments 11 and 13 ; those on the 

 latter being little, white, cylindrical tubes that can be protruded and 

 withdrawn at pleasure and seem to serve as signal-towers to attract the 

 attention of the ants that generally are found with the larvse. These 

 organs are situated near the spiracles. In Curetis they are permanently 

 exserted and coloured like the rest of the skin^ — but this is the only species 

 in which they are very long ; Aphnceus has them also exserted but short. 

 The organ on segment 11 — there is only one — is dorsal and central, consisting 

 of a transverse, mouth-shaped, generally slightly thick-margined gland 

 which exudes a sweet fluid of which the ants are evidently fond as they 

 stroke it with their antennae to excite production of the honey and then lap 

 it up with avidity. Some of the caterpillars, that is those of some species, 

 do not seem to have any gland at all, as, for example, Curetis where, if it 

 does exist, it is certainly not active and ants do not seem ever to pay it 

 any attention in consequence. In this last case the towers are supposed 

 to be organs of defence or protection, for the larva can throw out from 

 them a brush of long hairs which it flourishes rovind and round quite a 

 number of times when disturbed, and then suddenly withdraws. All the 

 hairs in the brush separate in the process. 



As regards the pupse or chrysalides, there are three very distinct types. 

 In two of them, characterising the sub-familes Lyccenince, part of the Arho- 

 palince and Theclince, the shape is normal as shown in figures 21«, 22a, 

 23a and 24a of Plate II ; although, here, too, there are difierences in in- 

 dividual facies, some being much stouter than others, some rather narrow 

 for their length and so on. The genera in the Arhopalince that have it 

 normal are Surendra, Arhopala, Zezius, Chliaria, Thaduka, Catapacilma, 

 Loxura, DeudorLv, Bindhara, Virachola and, probably, MaJiathala. The 

 normal pupa is always attached by the tail and has a tight band over 

 segment 5 attached to the surface on which pupation takes place on either 

 side. The second type belongs to Curetis and is depicted also on Plate II, 

 figure 28a. Here the shape is nearly quite hemis])herical with the whole 

 ventral surface quite flat; and the pupa is attached only by the tail 

 although, in some cases, there are signs of a body -band; The third kind 

 is confined to the arhopaline genera Creon, Pratapa, Tajuria, Bathinda, 

 Cheritra and, perhaps, Horaga and Zeltus, In this case the shape is 

 that of figures 25a, 26a and 27a and there is no vestige of a body-band, the 

 chrysalis standing free on its tail. 



There is one character common to all the arhopaline pupse which can be 

 used to distinguish them from the lycsenine ones and that is the shape of 

 the last segments of the abdomen. In the former the last segment is 

 dilated in the form of a horse's hoof round the under circumference of 

 which are fixed the suspensory booklets ; in the latter there is no such 

 widening. 



A key might therefore be made as under : — 



A, — Last segment of abdomen more or 

 less dilated and resembling somewhat a 

 horse's hoof. 

 a. Shape abnormal. 



a^ Outline more or less circular with 

 the ventral surface absolutely and 

 completely flattened . . . . Curetincs, 



