226 (Octobfr, 



successful, more frequently Avlion exerted from one side. The appearance 

 of tlie egg on the leaf was very niucli the same as that I recently de- 

 scrihed in the case of Tlirinax mixta — viz., that they were loosel}^ on the 

 sui'face and would easilj'^ fall off. But the method of attachment was 

 different in the two cases, though in both dependent on an incision hy 

 tlie saw. 



In Thrlnax a minute portion of the egg-shell was trapped between 

 the sides of a small veitical incision in the leaf. In P. pavula a pocket 

 was formed in much the same manner, as with those sawflies that place 

 the e^^ in a cavit}^ fully sufficient to contain it, beneath the raised 

 cuticular film; but the pocket in the case of F.pavida is much too 

 small, both in length and depth, to contain the Q^^^, so that the ^'^•^ 

 ]>laced in it touches both leaf surface and the under surface of the 

 cuticular film, but so as to leave the greater part of the q^^ exposed, as 

 is sliown diagramatically in fig. 3 in a cross-section of the ^^^^. Fig. 5 

 shows three remains of pockets after the removal of the eggs. The 

 cuticular film of the leaf, coming up one side of the Q%^, prevents the 

 fgg being removed, if pressed from tliat side ; but if pressed from 

 the other, not unfrequently it comes away unhurt, and leaves the 

 cuticular film c loose as a little white flap. Sometimes, as in fig. 1, a 

 portion may remain still holding the q^^ if the pressure for its removal 

 is witldield in time. In other cases, as in fig. 4, a portion of tlie film 

 remains attached to the egg, 



I thought it might be interesting to examine how the eggs of 

 Ncmatus ribesii were attached. Unlike P. ^^(ti'ida, whose eggs are on 

 the flat of the leaf, these are laid, like those of T. mixta, visuall}" all 

 along a rib, and it is interesting to note that the method of attachment 

 is essentially that adopted by T. mixta and not that of P. f avid a. 

 I found the eggs to be, if anything, more firmly attached than in either 

 of these species, so that I only succeeded in freeing several by a process 

 of tearing the leaf to j^ieces. It then apjoeared that an extremely narrow 

 fknge of the egg-shell was pinched or trapped in a slit in tlie leaf tliat 

 held it so tightly, and was so closed as to be practically invisible when 

 the Q^^f^ was removed. In fig. 6 a camera sketch of the egg (side view) 

 has added at d what seemed to be the relative proportions of the trapped 

 flange. But the true dimensions are not to be guessed or measured with 

 •any great accuracy in the released g%%. 



The death of the eggs of P. pavida, when the leaf begins to witlier, 

 confirms the view that fixity of tenure is not the only object in tlie 

 method of attachment, but tliat these eggs, apparently quite outside 

 the plant, derive from it nutriment of some sort, just as do those that 



