21 G Dr. T. A. Chapman on the 



moulting' the cremaster is thrust back over the skin to 

 reach the pad. The loss of the girth, however, deprives 

 the pupa of the power of preserving a correct alignment, 

 and so the fulcrum provided by the skin held in place by 

 the claspers cannot be efficiently utilized. 



The whole process of spinning the girth seems therefore 

 to be identical in Papilio and Thais. 



P.S. — The abortive journeys across the loop, between 

 the spinnings of its separate threads and the final journey 

 iu which the loop is thrown back, may be speculated on as 

 showing that the movement for spinning and that for 

 throwing back are variations of one and the same move- 

 ment, and that the definite distinction between them is 

 not so fixed, but that an intermediate movement may 

 occur, either as not being yet entirely eliminated or by 

 reversion. 



The larva of rapse makes the girth in a way that is 

 essentially the same as in Papilio, but yet with an amount 

 of variation that renders it actually very different. 

 Essentially the girth is made in front of the larva and 

 between the head and first pair of legs (not between the 

 1st and 2nd pair of legs, as in Papilio), but the raising of 

 the front segments of the larva, which in Papilio may be 

 likened to the "Sphinx" attitude, is in rapas carried to an 

 extreme, so that when the larva is adding to the middle 

 point of the girth the head is bent back so that the back 

 of the head touches the dorsum of the abdomen, about 

 the incision between 2nd and 3rd abdominal segments, 

 the ventral face of head and first thoracic segment being 

 directed exactly dorsal, the legs of 2nd and 3rd thoracic, 

 forwards. As the head is carried to either side, these for- 

 ward segments so rotate that the venter becomes ventral 

 over all segments, but the forward segments instead of 

 being bent dorsally, are bent laterally, and the head is 

 against the side of the 2nd and 3rd abdominal segments. 

 In all these positions the loop seems to be fairly tense. 

 When the head is bent to one side, the girth passes over 

 the middle of the 2nd abdominal segment and the middle 

 of the 1st thoracic, the portion of the larva between 

 these two positions being in front of the loop, the rest 

 behind it. In the median position, there is perhaps a 

 large proportion of the 1st thoracic segment in front of the 

 loop. Indeed the head only might be regarded as behind 

 the loop. It is observable, that during this process the 



