710 Dr. G. D. H. Carpenter on the 



than before, very thick, slightly flattened at the extremity, 

 and are set with minute spines along the edges. They 

 diverge upwards and outwards, and then turn forwards at 

 an angle. The spines on the third segment are only a 

 little larger than those on the middle segments. 



The processes on the tenth segment are a little larger 

 than those on the third, and flattened from side to side. 

 The last pair of processes (on eleventh segment) are very 

 large, and almost leaf-like, owing to the great flattening 

 from side to side : the anterior and posterior edges have 

 a frayed appearance, owing to their being set with small 

 spines close together. These processes curve forwards 

 and upwards. 



The head is very spiny, ash-coloured in front, dark 

 brown at sides. 



Pupation. — On July 29th the larva spent the day curled 

 up on a leaf eating nothing, and on the evening of the 

 80th suspended itself by the last pair of claspers from 

 the tip of a leaf. On the 1st August, very early in the 

 morning, before daybreak, it pupated. After this first ex- 

 perience I have reared numbers of these larvae, and it is 

 very interesting to note that the preparations for pupation 

 always occur in the same way. During one night, after 

 remaining motionless in a curled-up posture on the leaf 

 for some twelve hours, the larva will suspend itself from 

 the leaf tip; and pupation takes place during the next 

 night. Presumably by this adaptation the pupa gains by 

 not being exposed to daylight till the protective green 

 colour is fully developed, which takes some hours. The 

 pupa corresponded exactly with the figure and description 

 of that of Ps. imitator (loc. cit.), save only that the long 

 processes from the head were not twisted but quite 

 straight and parallel to each other. 



The pupa in colour was leaf green, but the lower surface of the 

 dorsum, and head, was slightly tinted with a light bluish grey- 

 bloom, as if to neutralise shadow. From the top of the head project 

 a pair of flattened stalk-like processes, soldered together along 

 adjacent edges except at the extreme tips, which are square. These 

 processes immediately after the shedding of the larval skin are 

 se^jarate from each other, short, and curved dorsally. They appear 

 to be straightened out to their final position and shape by the forcing 

 into them of fluid, and the triangular processes on the abdomen 

 are similarly distended, being very small at first. The cephalic 



