1919.] ex 



to the terebra, whence it is dropped, probably one mass for each egg. 

 I have noted finding one such mass still attached to the fly, but similar 

 pellets in some numbers lay on the floor of the jar. 



It is curious that P. j^ini, whose eggs hatch not very long after 

 they are laid, should protect them with a thick covering, and P. sertifer 

 should not do so, though her eggs have to remain in situ unhatched 

 thi'oughout the winter. 



Amongst the few sawflies whose egg-laying I have watched, these 

 are the only ones that remove actual material from the cavity formed to 

 receive the es,^, and I note that, in mounting the saws for examination, 

 the basal mass (the bones and sinews through which the saws are 

 worked) is very solid and massive. The saw itself is rather short, very 

 robust, and armed with a row of teeth on each section, that are stronger 

 than in most sawflies, and seemed calculated to rasp out material as they 

 are retracted. 



Both species cast their larval skins immediately before spinning up. 



Plate I, fig. 3, shows the uiethod of feeding gregariously that is 

 followed by P. pirn in its first larval instar, and less constantly in the 

 second. If a larva be separated from its companions and isolated, it does 

 not thrive and usually dies, but I did not experiment on this point 

 sufiiciently often to say that this is alwa3^s the case. One can easily 

 understand how a solitary larva of a species that lives gregariously in a 

 nest or under a wel), such as, for example, TJiauinntopoea pityocamjya, 

 the pine-processionary moth, is unable to thrive when deprived of the 

 protection en303fed with its fellows, but it is less easy to see why it 

 should be so in species such as P. pini and ClaJius viminalis. It 

 seems very possible that the process of feeding on a narrow strip of 

 tissue is very difficult, unless the adjacent strip is being at the same 

 time removed by its neighbour. 



Many of these sawflies have a curious habit, the use of which seems, 

 very obscure. If a few $ $ of these are placed together in a jar or 

 cage, with a good deal of room where any collision would seem unlikely 

 unless activeh' sought for, it is not long before some, and a little later 

 nearly all, are mutilated by the loss of one or more legs or antennae, 

 bitten off by their neighbours. This occui-s with Cladius viminalis, 

 Phadinoceraea micans, Phymatocera aterrima, and both these species 

 of Pteronus. For example, Mr. Green sent me five $ $ of P. sertifer 

 on October 25th, 1917 ; these were dead on October 30th, having been 

 placed in a jar with branches of Pinus sulvestris. Three still had their 

 legs and antennae perfect, one had a 2nd leg wanting, another two legs 

 abbreviated. 



