806 STUDIES IN AUSTRALIAN NEUHOPTERA, vii., 



to conceive how so large a larva could have managed to encase 

 itself in it. 



As regards the actual mode of spinning the thread, the mobility 

 of the anal papilla is very marvellous. Lying upon its back, the 

 larva can protrude this papilla to a great distance, and spin silk- 

 all round itself, even behind its head and thorax, with very little 

 effort. Most of the threads, however, ai'e spun transversely 

 across the larva, beginning with a small arc at the anal end, 

 then swinging out into a longer arc across the middle of the 

 abdomen, and finally making shorter arcs behind the back of the 

 thorax. All these transverse strands appear to be carefully con- 

 nected, on both sides, with the small platform already mentioned, 

 which itself is made to form a part of the outermost coat of the 

 cocoon. When the first complete layer of the sphere is finished, 

 the larva has its head bent over so as almost to touch the tip of 

 its abdomen: but the mobility of the anal papilla appears to be 

 but little decreased in this position. In spinning tlie inner layers 

 of the cocoon, the larva changes its position every now and 

 then, so that the anal papilla may cover all parts of the sphere 

 with an equally thick layer of silk. 



As found under natural conditions, the cocoon is an oblate 

 spheroid, with the two poles in contact with the sides of the 

 crevice in which it is placed, and the larva lying so that its 

 sagittal section is in the equatorial plane of the spheroid. When 

 spun, however, in an open tube, the cocoon is practically a sphere. 

 It closely resembles the cocoon of a Chrysopid, particularly that 

 of the genus Nothochrysa, from which it would not be possible 

 to distinguish it for certain. But, unlike most of the Chrysnpidoi, 

 Psi/chojjsis does not attempt to spin any external matter into 

 the outer mesh of the cocoon. Such a proceeding is, of course, 

 unnecessary, considering the position in which it is placed. The 

 diameter of the cocoon varies with the size of the larva, but is 

 usually about 5 mm. (Plate Ixxix., fig. 6). 



The Pupa. (Plate Ixxix., %. 7; Text-tig. 10). 

 As far as the external form is concerned, the true metamor- 

 phosis is that between larva and pupa. The latter is a pnpa 



