310 NOVITATES ZoOLOaiOAE XXVI. 1919. 



If the caterpUlar of ihcas is thus successful in warding off the foe by one or 

 other of these diverse methods, the pupa is hardly less successful in its ability 

 to look wooden and unattractive ; for when formed on the trunk or branch of 

 the orange tree where the larva has been feeding, it is lost in obscurity, not by 

 being hidden with leaves, but by its perfect reproduction of the stump of a 

 broken and half-decayed branch. It is not, however, immune from parasitic 

 attack, and is often found as a discoloured shell full of holes, from which hymenop- 

 tera have emerged. These small yellow wasps may be identical with the species 

 bred from P. lycidas and from the pupae of at least three other local Papilios. 

 In shape the pupa of thoas is moderately rotund, being swollen in the central 

 abdominal segments and tapering considerably towards the anal extremity. 

 By comparison with others it seems small for a butterfly with such an expanse 

 of wing as thoas possesses, not to mention its tails. The " ears " are stout and 

 are well projected forwards, and the thorax is surmounted by a short hump 

 also pointing forwards. 



The deep tone of Indian yellow which the butterfly is sometimes seen to 

 possess, even when on the wing in Para, is a feature worth noting ; and I presume 

 it is due to atmospheric humidity and sunlight. Experiment shows that the 

 same deep tone may be produced in a light yellow specimen by kilUng it in an 

 old and wet cyanide bottle. If left in the fumes of ammonia still longer, a much 

 heavier tone approaching brown results. As this is not the case with other 

 yellow butterflies, it is obvious in the present instance that its scales are particu- 

 larly sensitive to colour change by chemical action, and possibly even during 

 life by sunlight and moisture in combination. 



Twice have I secured good varieties of thoas, one being so heavily blotched 

 with black that it resembled a distinct species. 



P. androgeus androgeus (pi. ii. fig. lf>, pl- iv- fig- 4). 



This is a very uncertain species in Para, and at times appears to be entirely 

 absent. Indeed, for a couple of years I had taken nothing but a single empty 

 pupa-case on the trunk of a lemon tree at Marco da Legua. In the early part 

 of 1914, however, the species turned up in sufficient force to reveal its life-history- 

 and furnish my collection with a perfect series of bred specimens of both sexes. 

 Where it came from and whither it has since vanished is a mystery. I have, 

 however, traced it along the railway lines leading both to Pinheiro and Bragan9a, 

 and taken its larva on the isle of Cafezal and on another island beyond the Eio 



(juama. 



The egg, larva, and pupa of androgeus are constructed on lines very similar 

 to thoas, but with certain well-marked specific difierences. Indeed, almost 

 all that I have said in description of that species and its habits, including 

 its positions in nature, the measures resorted to for protection, and even 

 its liability to the attack of the small yellow wasp, applies with equal force to 

 androgeus. 



Noteworthy characteristics are as follows : 



Food-plants: Citrus, e.g. lima, lime, ■'limao galego," and tangerine, with 

 special preference for the last and for young trees. Not found on Piper. Ova Ught 

 green when first laid, quickly turning deep yeUow ; easily found on the freshest 

 and tenderest leaves, sometimes five or more on a single bush. Young larva 



