Pupal suspension of Thais. 207 



one eminence. The head and prothorax are stretched out 

 when the attachment of the loop is being worked at, the 

 first pair of legs being one on each side of the strand ; when 

 the centre of the loop is under the spinneret the thorax is 

 bent back from the twig, but the head is bent down to 

 bring the spinneret against the loop in front of the 2nd 

 legs, and the 1st pair of legs become practically invisible. 

 In two specimens watched the whole process took about 

 an hour and a half, in a room at about 64°. One wondered 

 all the time how the loop when finished was to get back 

 to its place. This however is managed very simply and 

 very quickly. After finishing, as it had appeared to finish 

 several times before, one of the lappings of the side of the 

 loop with binding threads, instead of sliding slowly round 

 and bending the head down slowly as it went, to the posi- 

 tion it takes when at the central point, it gave it this 

 position at once, i. e. before moving round, so that the head 

 went under the loop somewhat to one side, and as it then 

 gradually assumed the median position, the thread lay 

 across the middle of the front of the head. At the same 

 time, however, as it assumed the median position, it bent 

 back the head and curved the thoracic segments backwards, 

 so as to bring close together the back of the head and the 

 dorsal thoracic humps. Then the thread became slack 

 over the head, and slipped back into its place. In one 

 instance, the thread caught, on the side on which the 

 manoeuvre was made, the tops of the subspiracular warts 

 of the meso-and meta-thorax, bending them back, the larva 

 rested here a little, probably from some inadvertent move- 

 ment of mine that alarmed it, and then by a simple turning 

 and stretching movement the loop fell into its place. One 

 asks, Why does the loop always fall to this identical spot, 

 how does the larva manage to make the necessary move- 

 ment so exactly ? This question is justifiable as an 

 expression of admiration that the larva should always 

 make precisely the same movement, but is foolish in not 

 seeing that the same movement will always produce the 

 same result. In fact a little, but very little, latitude is 

 possible, as one could see that the result would be the 

 same had the movement been a little more or less ample, 

 etc. Nevertheless, the length of the loop and the place 

 from which it starts must be very exactly related to the 

 length, thickness, and movements of each larva. 



Another point may be noted, whilst the larva is at work 



