28 



SPOLIA ZEYLANICA. 



under the shade of a tuft of grass. The plant had evidently been 

 badly eaten by the larva, and there was no flower stalk. The pupa 

 was attached by its cremastral hooks to a silken pad spun on the 

 base of a leaf just below the central bud and was lying, dorsal 

 surface uppermost, across some leaves whose gummy petioles had 

 been eaten away by the larva. This pupa was of a greenish-yellow- 

 brown colour, just the tint of the faded sundew leaves, and it looked 

 rather like a grass seed which had fallen on to the plant and stuck 

 to the gum; it may be added that ripe grass seeds are often so found. 

 In confinement the larva exhibits a certain preference for suspen- 

 sion from the flower stalk of its food plant, whose colour is of a reddish 

 green. Even when the stem is growing at an angle, its double set 



Fig. 6. — (Upper figvire) Pupa from the side. 



Fig. 7. — (Lower figure) Full-grown larva eating into a seed-capsule. 



(From drawings by E. E. Green.) 



of cremastral hooks enables the pupa to keep its ventral surface 

 closely appressed to the lower side of the stem, so that it is not sus- 

 pended freely. It seems possible that this pupa possesses a certain 

 amount of colour adaptability, those pupae attached to the reddish 

 flower stems having usually an increased red suffusion in comparison 

 with those attached to glass or white paper. 



When on an approximately horizontal surface, the pupa is usually 

 found dorsum uppermost ; otherwise it invariably suspends itself head 

 downwards and with the ventral surface appressed to its support. 



In the case of a pupa in a horizontal position the cast larval skin 

 is sometimes seen lying nea r it , but quite free 9-nd shrivelled up. The 



