﻿1869]. 157 



for the process of once changing its skin, whilst G. ignita is fed up in 

 six days, during which it has found time to change its skin four times. 

 I was also struck with the similarity between the larva of Cfirysis and 

 that of Odynerus ; a similarity that seems to be a true and not a super- 

 ficial one. Throughout its existence, the larva of spinipes is yellow, 

 its viscera are tolerably visible through the integument, especially 

 portions of a yellow tortuous duct in the lateral dorsal region from the 

 fifth segment backwards. In Ghrysis, the larva is white, and its interior 

 is more masked by masses of white fat. The first spiracles, which, 

 though usually situated in the second (the head being the first), belong 

 properly to the third segment, are in Ghrysis at the anterior margin of 

 that segment ; but, in O. spinipes, they are actually in the second seg- 

 ment. The form of the head and parts of the mouth are very similar 

 in both. This resemblance between the two larva? is closer than that 

 between the larvse of spinipes and of the common wasp (Vespa vul- 

 garis) ; and, in those points in which the larva of Ghrysis least resem- 

 bles that of the Odynerus (e.g., form of jaw, distinctness of viscera as 

 seen through the skin, and colour), it resembles Vespa vulgaris more 

 than O. spinipes does. I have not been able to seize any characters to 

 distinguish the larvae of G. ignita, hidentata, and neglecta from each 

 other. 



G. neglecta spins a compact oval cocoon from 5 to 10 millimetres 

 in length, of a greyish-white and blackish silk, in layers, similar to the 

 cocoons of various ichneumons, Ophion for example ; this is surrounded 

 by a looser layer of brown silk, similar to that which loosely fills up 

 the rest of the cell of O. spinipes, and some remains of the little green 

 grubs are always to be found at the bottom of the cell. The cocoon 

 of G. ignita is rather longer than that of G. neglecta, of a much slighter 

 texture, and with hardly any loose silk about it. 



The cocoon of G. hidentata is contained in that of O. spinipes, the 

 cocoon proper occupying the lower half of the cell, and its roof being 

 an almost mirror-like diaphragm of brown gummy silk stretched across 

 the centre of the cocoon of the Odynerus, the walls of which, above 

 the diaphragm, are also covered by a thin layer of silk spun by the 

 larva of Ghrysis ; the lower part of the cocoon is in contact with the 

 Odynerus cocoon all round, and contains in its walls three vertical 

 whitish patches, rather thicker than the rest of the cocoon, which, 

 when removed from that of the Odynerus, is translucent. 



Like the larva of 0. spinipes and other hybernating Hymenopterous 

 larvse, that of Ghrysis shrivels to a certain extent after it has spun its 



