30 3POLIA ZEYLANICA. 



thread. It feeds on the unripe seeds of Boerhavia repens, com- 

 mencing by eating the viscid exudation on the outside of the 

 perianth tube, through which it then gnaws a hole and excavates 

 the contents. Small insects, especially ants, are often seen to be 

 caught by this gummy secretion, but the gum does not seem to 

 incommode the larvse at all ; probably their extremely long prolegs 

 are specially modified to carry them over it without touching it as 

 they walk, and the long larval hairs prevent contact of the body 

 with neighbouring drops of gum. 



Pupation. — The larva seems to pupate almost invariably on the 

 slender stem just below a seed-head, although I have once found an 

 empty pupa-case attached to the midrib on the under- surface of a 

 small leaf. The pupa hangs freely suspended, the discarded larval 

 skin not being shrivelled up but stretched out at full length along 

 the stem just above it. The rain soon destroys the empty pupa 

 cases and one finds only the anal portion with the discarded larval 

 skin. The colour of the pupa is very variable ; sometimes it is a 

 light apple-green, sometimes a brownish-gray. (Plate F., figure 3.) 



Trichoptilus xerodes, Meyr. 

 (Plate A., figure 9.) 



Meyr., T. E. S., 1886, 14; 1. c, 1885, 422; 1. c, 1887, 267; 

 B. J., XVII., 134. 



Distribution. — Colombo, Trincomalee, Peradeniya, Madulsima. 



Widely distributed, but by no means a common species in Ceylon. 



This species is widely distributed throughout Australia, but does 

 not seem to have been recorded from elsewhere outside of Ceylon. 



Early Stages . — Unknown . * 



■ id 

 .■yC * A larva (supposed to be that of T. xerodes) was found at Peradeniya on 



P k«/> Gynandropsis sp. (Capparid) on December 26, 1907. A description of this 



^ A^fC^ larva reads as follows : " About 12 mm. long, cylindrical, moderately stout. 



< S^ uL > Head yellowish with an orange tinge. Colour of other segments a uniform 



W^- 



-twU" ** * /J pale yellow. A large brown latero-dorsal wart emits a long white hair and 



iSAP^- about five short ones. Below this is a small black supra-spiracular tubercle 



vAy-iS- ^"emitting a single short white hair and bearing a short secondary hair. Spiracle 



e. .small, black. A small black subspiracular tubercle emits (i.) a short white 



IW}^-- h a i r directed forwards and downwards, (ii.) a longer white hair directed 



backwards and downwards. Below this and a little behind it is a small 



black wart emitting a single hair. Towards the ventral surface are two 



(? three) small black-warts emitting white hairs. There are numerous small 



knobbed white secondary hairs. All warts are well raised above the surface 



of the skin, and the divisions of the segments are well marked." (Plate F., 



figure 4.) 



Although the exact identity of this larva is doubtful, its general appearance 

 led me to consider it that of a Trichoptilus, and in this connection it is note- 

 worthy that the Gynandropsis has well-developed sticky glands, as in the case 

 of the other known food plants (Drosera, Boerhavia) of the plumes of this 

 genus. It is possible that the larva may have been that of T. wahlbergi, or 

 even that of an Oxyptilus, but an examination of plants of Gynandropsis and 

 £ of (Jleomh viscosa will probably settle this question. 



