288 JOURNAL, BOMBAY NATURAL HIST. SOCIETY, Vol. XXL 



bosticics and himself gives a fourth. The larva as met with in the Hyder- 

 abad district of Sind differs from all four and is as follows : Head shining 

 black ; spiracles dark brown; ground colour a dirty flesh colour or crimson 

 with the following darker crimson markings : — a mid-dorsal line, below 

 this on either side a series of oblique streaks, one to each segment, running 

 backwards and outwards to meet a sub-dorsal line, a third line just below 

 the spiracles. The whole body is covered thickly with short, stiff black 

 bristles but on the eleventh segment, in the usual situation for the two 

 pillar-like processes from which the tentacles are thrust forth, are two 

 bald patches. In some specimens, I examined, these two areas were cov- 

 ered with a scanty covering of bristles shorter than those clothing the 

 rest of the body and what was still more interesting, in one particular 

 specimen, whereas one area (the right) presented immature bristles, the 

 other was entirely bare. The pillar-like processes were absent in all spe- 

 cimens examined but the mouth like opening situated in the middle of the 

 posterior part of the segment was always present. The larvge as one 

 might expect from this description, were unattended by ants. I attempt- 

 ed to persuade the larva to secrete the sweet liquid so much appreciated 

 by ants, by means of tickling them with a camel-hair brush but no 

 amount of irritation had any effect, and I conjecture the mouth-like organ 

 is functionless. The larva met with in other parts of India has the pillar- 

 like processes well-developed and is always attended by ants. Referring 

 again to De Niceville I see in his preface to the Lycaenidae that he mentions 

 that these organs were first described by M. Guenee in 1867 from a speci- 

 men of P. boeticus and in his description of the insect observes that no less 

 than three species attend it. The habits of the Sindh form fully explain 

 why these organs should have degenerated, for degenerated they 

 clearly have, the two bald areas representing the sole vestiges of the 

 organs. 



The larva3 I examined are winter forms, a period of the year when ants 

 are more or less dormant in Sindh. The young larviB as soon as they 

 hatch out, eat their way into the bud on which the ovum has been depo- 

 sited and then close the hole of entrance with fine web, and also prevent the 

 bud from opening, with other strands which bind the leaves together. The 

 larvae are thus fairly easy to find by noticing in any panicle of bloom 

 whether any bud has remained unopened. Examination of such a bud 

 usually results in finding the owner at home. Thus not only are possible 

 foes excluded but also the friendly ant, who would find it quite impossible 

 to enter the bud in order to milk the larva. It is significant that two 

 other larvae, those of Aphnaeus hypogyrus and Tarucus theophrastus belonging 

 to the same family and which are invariablj^ attended by ants do not breed 

 in this same district, during the winter months, probably because there are 

 no ants in the cold season to care for them. It is to be conjectured that 



