360 



the larvEB like those of Notodonta ziczac, dromedarius &c., change 

 to a bluish grey, the beautiful saddle-like marking with its yellow- 

 border appears much darker, but is blended into the general tint. 

 They next spin a thick web, and then eat into the bark or wood on 

 and in which they form the cocoon. The particles eaten off are forced 

 through the web, giving it the appearance of the substance to which 

 it is attached, so that it is not easily discovered until winter's rains and 

 winds have bleached it. The cocoon is generally erect, the head of the 

 insect being uppermost. The escape of these beautiful insects from 

 the cocoon is as mysterious as the manner in which they form it : how 

 so delicate a creature can, without injury, break through so hard a 

 substance, is one of the wonders of Entomology. I have observed 

 that the perfect insects generally come out towards evening ; they re- 

 main quiescent much longer than several of our Lepidoptera before they 

 begin to expand their wings, and previously to doing so they rarely 

 remove far from the cocoon. The larva of this genus is attacked by a 

 very small parasite of a metallic green or bronze colour. So far as 

 my experience goes, it does not appear to attack the larva until the 

 latter has formed its shell-like covering, and has entered the pupa 

 state, or is about doing so. The parasite then drills a small hole in 

 the cocoon, through which it deposits its eggs. It has been suggested 

 that this hole probably affords the means of egress ; this cannot be 

 the case, as the hole is funnel-shaped, being larger at the top than at 

 the bottom. On opening one of these cocoons in August last I found 

 it full of the fly ; and a few days afterwards I discovered a cocoon, 

 which, from the freshness of the edges of the hole, I concluded had 

 been recently perforated ; by means of a powerful glass I could per- 

 ceive that the tenant was still in the larva state, and at work. This 

 cocoon I opened the middle of last month, and found that the larva 

 had gone into the pupa state and was full of small cocoons, from which 

 no doubt the same flies had been produced, but they had all escaped 

 i'rom the cage where the cocoon had been placed, the hole made by 

 the parent fly having aff'orded them the means of egress. — Alfred 

 Lambert; 6, Trinity St., Southwark, July 7, 1342. 



JOriN VAN VOOI{ST. ff\ aHK^^iS.^'x7 PATERNOSTER ROU 



