298 PROCEEDINGS OF THE 



July, and not one larva could be seen. It has, however, yet to 

 be observed whether there may not be two broods in a year, though 

 certainly with me the flies did not make their appearance till 

 the following March ; and this fact rather militates against the 

 supposition of the species producing a double brood in a year. 



The cocoons are somewhat larger than those of N. Vallis- 

 nieri, and the colour is rather darker. 



The larvae changed to pupae in March ; and in about fourteen 

 days after the flies emerged. The pupa is greenish white, with 

 the antennae, feet, and wings, white. 



In February I noticed in the bottle where the cocoons were 

 kept, a number of small white grubs, which at the time I did not 

 examine particularly, believing them to pertain to the Ichneumonidae 

 or Chalcididae. In a short time they changed to pupae, and it was 

 then observed that they were not Hymenopterous, but Dipterous 

 insects. Their history seems to be as follows : — 



The Dipteron is an internal parasite of the Saw-fly larva; the 

 latter retains sufficient vitality to spin its cocoon, inside of which 

 the fly-larva completes the destruction of its victim. In February 

 it quits the cocoon, a proceeding rendered necessary by the 

 mouth organs of the fly not being sufficiently strong to cut 

 through the cocoon. In the earth it changes to a pupa. When 

 I described one of these it had the thorax fuscous and shining ; 

 the abdomen white, but with a tinge of green ; the antennae, 

 wings, and feet, black ; but probably at an earlier stage their 

 colour was white. The limbs are laid along the sides of the 

 breast, the feet crossing over the abdomen. The pupa is enveloped 

 in a thin pellicle, which retains its shape after the fly has left it. 



The larva quits the cocoon by a rough, irregular hole, made in 

 the side ; and, from finding more than one hole in some of the 

 cocoons, it seems probable that two or three of the Diptera 

 are sometimes parasitic on one Saw-fly larva. Nearly a third of 

 my cocoons were infested with these parasites. 



Mr G. H. Verrall tells me that the fly is a Sciara, probably 

 S. humeralis, Zett. ] or confinis, Winn. I am also indebted to 

 Mr Verrall for the information that no species of the genus has 

 been noted as a parasite hitherto ; but one species (S. pulicaria, 

 Mg.) has been reared from rotten galls on Salix pentandra ; and 

 anotlier species ( S. tUicola, Lw.) is a true gall-maker. 



The female Saw-fly has the antcnni\3 a little shorter than the 



