204 THE entomologist's record. 



about •8iiun. at one end, and slightly less at the other. The surface 

 is covered with a very much raised, but at the same time delicate, cell 

 network, the thin edges of the dividing walls giving the eggs an appear- 

 ance of being covered with a " bloom." In colour they were, when 

 freshly laid, a pale yellow, but have now assumed a deep, dull, orange 

 shade. The eggs are deeply depressed on both sides, and the cut oft" 

 appearance of the micropylar end appears to be mainly, if not en- 

 tirely, due to a similar depression at this end. 



Larva : A fully detailed description appears to be unnecessary, 

 thanks to Mr. Burrows' camera drawings, in which the special features 

 will be much more easily understood, than from any account, however 

 lengthy, which I could give. In the first stadium the lateral flange on 

 abdominal segments I to 5 is enormously developed, but I cannot 

 detect the spiracles, which must be hidden in, or by, the deep 

 depressions and folds of the skin. The positions of the specially 

 developed hairs, which are presumably used for attaching the fragments 

 which form the coat, very closely, if not exactly, agree with those of the 

 allied P. .s)iiarai/dana. The primary tubercles, so far as I can ascertain, 

 are arranged thus, i and ii set trapezoidally on abdominal segments, 

 transversely on meso- and metathoracic segments. These bear long 

 hairs on most, if not on all, segments, trumpet-topped. I cannot trace 

 iii on the segments bearing the extremely elevated flange, and I am not 

 certain that I rightly identifiy it on the other segments as a compara- 

 tively small tubercle, bearing a small, simple, tapering hair. The 

 long and highly specialised hairs on the flange I identify with iv and 

 V. The large trumpet-topped hair I take to be iv, and the long taper- 

 ing, somewhat serrated hair shown on the lower edge of the flange, I 

 suppose to be v. On the segments where the flange is not developed, 

 they would appear to be small and inconspicuous simple hairs. Beyond 

 the difference in the shape of the tops of the specialised hairs, and the 

 much greater extent, to which the lateral flange is developed in P. 

 jiKstiilata, this species shows a fairly parallel structure with P. 

 xinarajidaria, with which it also agrees in the shagreened or spicular 

 character of the skin surface. I^ut in the remarkable, development of 

 what IMr. Burrows calls the " battledore " hairs, jnistidata shows 

 independent specialisation. These hairs, which appear to me to be 

 "flask-" shaped, I take to be secondary characters, and not develop- 

 ments of any existing primary hairs. I see that Mr. Burrows has slightly 

 exaggerated their size in relation to the specialised primaries in his 

 drawing, evidently with a desire to make them more distinct. The 

 fact of their occurrence upon the thoracic and abdominal segments 6 

 and 7 proves that they are not directly correlated with the remarkable 

 development of the spiracular hairs (iv and v), and the upward shift- 

 ing of the spiracular flange, and its great development on the 1st to 

 5th and 8th abdominals. This view is also supported by their absence 

 on the young larva of /'. suianiiidaria. Their purpose presents an 

 interesting enigma. The only theory which occurs to me being that 

 they have a buoyant or balloon-like use, when the larva drops, as it 

 probably does, on a thread, similar to the supposed use of the bulbed- or 

 balloon-based hairs of Psiliira imDiacha, Pnrthrtria dis/iar, and some, at 

 least, of the " Footman " larva^ when in their young stage. In the adult 

 larva, the condition of development of the base of tubercle iv into a special 

 process for the attachment of the silk threads is well shown by Mr. 



