106 JOURNAL BOMBAY, NATURAL HISTORY SOCIETY, 1892. 



as an " unpleasant body, " we therefore failed to raise a motli from 

 it, and so were unable to determine its species. 



In tbe otber specimen the likeness to a cobra was even stronger. 

 This also was a Geonietra caterpillar, of about the same size as that 

 just described, which we found in August, 1890, at Nasik, feeding on 

 a species of Evolvidus, that small creeping herb with bright blue 

 flowers, like a tiny Convolvulus, which grows commonly in spreading 

 tufts on rocky ground during the rains in Bombay. Its markings 

 were at the head end, and gave it when "looped'" exactly the 

 appearance, from the front, of a cobra reared in act to strike. Not 

 only so, but if the annoyance which caused it to assume the " terri- 

 fying attitude" was continued, it actually did strike, though of 

 course quite innocuously, exactly like a cobra, in the direction of 

 its assailant, turning for the purpose to the side or rear and with 

 such hearty good will as sometimes to over-reach itself and fall 

 prone. I frequently tested it, sometimes with so uninviting a 

 subject for attack as the toe of my boot, and never failed to " get a 

 rise out of it." 



Unfortunately this specimen was lost on its way down to Bomba}^ 

 before it had turned into a chrysalis, so in this instance also we 

 failed to determine the species. 



Now these are to my mind two very interesting cases, well worthy 

 of further consideration. In the first place, it will be noticed that 

 the one iirst described uses its likeness to a venomous snake for the 

 purpose of making an opportvmity to avoid its assailant ; but the one 

 last described, though evidently belonging to a closely allied species, 

 uses a very similar likeness for the purpose oi making its assailant 

 avoid it. 



In the next place, it will be observed that in the latter case the 

 likeness is more complete, not only in aj)pearance, but in conduct. 



Then comes the question, Whence the likeness ? Is it because it 

 has proved of use to the caterpillar to be like a cobra ; or is it because 

 it has proved of use both to cobra and caterpillar that a " creeping 

 thing " should be able to suddenly assume an erect arid minatory 

 attitude with expanded crest and spectacled head? 



If the former is the true answer, we are met with the difficulty 

 that, for the likeness to be of any general use to the caterpillar, the 



