328 INSECT ARCHITECTURE. 



inspected narrowly, a few threads of silk may be 

 seen interspersed through the whole. In size, it is 

 not larger than the egg of the gold-crested wren. It 

 has been considered by Brahni a puzzling circum- 

 stance, that this cocoon is usually perforated with 

 one or two little holes, as if made by a pin from 

 without; and Kirby and Spence tell us that their 

 use has not been ascertained.* May they not be left 

 as air-holes for the included chrysalis, as the close tex- 

 ture of the cocoon might, without this provision, prove 

 fatal to the animal? Yet, on comparing one of these 

 with a similar cocoon of the large egger-moth [La- 

 s'wcampa Qncrci(s), we find no air-holes in the latter, 

 as we might have been led to expect from the 

 closeness of its texture. We found a cocoon of a 

 saw-fly (Trichiosoma)y shout the same size as that 

 of the egger, attached to a hawthorn twig, in a 

 hedge at JVew Cross, Deptford, but of a leathery 

 texture, and, externally, exactly the colour of the 

 bark of the tree. This was also without air-holes. 

 The egger, we may remark, unlike the dock-weevil 

 or the bee-grub just mentioned, can work her co- 

 coon without any point of attachment. We had a 

 colony of these caterpillars in the summer of 182.5, 

 brought from Epping Forest, and saw several of 

 them work their cocoons, and we could not but ad- 

 mire the dexterity with which they avoided filling up 

 the little pin holes. The supply of their building 

 material was evidently measured out to them in the 

 exact quantity required; for when we broke down a 

 portion of their wall, by v,ay of experiment, they did 

 not make it above half (he thickness of the previous 

 portion, though they plainly preferred having a thin 

 wall to leaving the breach unclosed.']" 



Several species of caterpillars tliat spin only silk, 



* Erahm'slns. Nat. 2SD, andKirby oiulSpeiice'sIntr. iii. 223. 

 t J. R. 



