206 Dr. T. A. Chapman on the 



The following notes made whilst actually watching the 

 larva may help to explain the spinning of the girth, though 

 they have some incoherence from the circumstances of their 

 production. 



Cassando-a, when building its loop, has it quite ventral, 

 it passes round above the marginal wart of 2nd and 3rd 

 thoracic and 1st abdominal and between the 1st and 2nd 

 pairs of legs. There is, however, only one phase in this 

 process to which this applies, viz. when a thread of silk is 

 being added to the loop at exactly its highest point. When 

 the spinneret is against either side of the loop or working 

 at the point of attachment, the loop is free from all contact 

 with that side of the larva, or just touches the marginal 

 wart of 1st abdominal, whilst it is fully stretched on the 

 other side, over the 2nd thoracic leg, the three warts (2nd, 

 3rd thoracic, 1st abdominal), and touches the ventral 

 prominence of 1st abdominal. The attitude of the larva 

 being that of a curve or twist that brings these parts 

 directly away from the twig of attachment, whilst those 

 that the loop is free from are crowded together towards 

 the twig, the three warts having their apices close together, 

 and so that on the curvature relaxing they expand 

 again into that side of the loop and hold it whilst the 

 other side contracts and frees itself from the loop in its 

 turn. 



The larva moves very leisurely, and with some to and 

 fro movement, so that one traverse of the loop takes about 

 three minutes and the movements of fastening the end of 

 each thread to the twig about one minute ; but between 

 each complete traverse usually at least one partial journey 

 is taken, i. e. from the twig for about one-third of its 

 length and then back again, and along this piece especially 

 towards the end of the process a good deal of local spinning 

 i.s done which covers this thicker portion of the loop with 

 an outside binding. When the loop is finished, the central 

 third consists of a number of threads more or less separate, 

 or at least apparently separate, straight, parallel and 

 uncomplicated. The end portions are thicker, and bound 

 together as one strong strand. 



As the larva moves its head from side to side, the loop 

 slips to and fro, or rather perhaps the larva slips to and 

 fro on it, the loop always taking the position described on 

 the convex side of the larva,, and lifting out of it on the 

 concave, as the warts and le^s are contracted together into 



