the Biology of some British Neuroptera. 523 



myself very disappointed to find that the first-instar larva 

 has no trace of two tarsal claws, or apparently of palpi, 

 though further work may reveal something. A forerunner 

 of Sisi/ra probing mud for Dipterous larvae, as does Osmylus, 

 might well have discovered in the sponge (which usually 

 contains such larvae) an easier method of obtaining food. 



Larvae live on the outside of the sponge and probe it at 

 intervals, or they m.ay enter the canals. The long setae 

 covering the ])ody often entangle debris which serves for 

 concealment. Full-fed larvae may be found all the year 

 round, and circumstantial evidence points to several broods 

 in the year, the main one being in May and June. This 

 year, when eggs were just hatching, in early August, full- 

 and half-grown larvae were also found in the same locality. 

 Nevertheless it should be stated that no pupae have been 

 taken later than July. 



When full fed, the larva leaves the water and spins a 

 small yellowish-white cocoon 4 mm. by 3-5 mm. of oval 

 shape. This consists of an outer, coarse, yellowish silk 

 network, and inside, a finer white cocoon. On the walls 

 of a bridge crossing a river or canal, cocoons may often be 

 found in great numbers. In one locality the larvae 

 regularly cross a tow-path ten feet to a bridge wall and 

 then ascend this to under the eaves of the bridge, twelve 

 feet from the ground, before spinning up. There is a 

 great mortality in the cocoon, in nature. About one-third 

 die as larvae and are found covered v/itli white mould. 

 Some die as pupae, and yet others are parasitised by a 

 Hymenopteron which lays an egg beside the resting larva. 



The pupa is not peculiar and is at first of the same 

 colour as the larva, later darkening and apparently emerging 

 in about two weeks. The pupa often leaves the cocoon 

 completely before disclosing the adult, but sometimes 

 remains within, after having as usual bitten a hole for its 

 escape. 



More evidence is necessary on several points. Larvae 

 may be found in winter, and probably hibernation is 

 normally as such, but I hope to make further observations. 

 The difficulty of breeding Sisym is only that of keeping 

 the sponge alive, which so far has proved impossible for 

 more than a week or so. The larva feeds upon both 

 Spongilla lacustris and Ephydatia Jiuviatilis. 



The imagines pair readily in captivity, shortly after 

 emergence, at dusk. There is no previous ceremony. The 



