101 



it is laid, but it supplants that of Odynerus spinipcs, probably because being 

 laid by the side of that of the wasp the larya to which it gives birth destroys 

 the Odyntrus egg. This is, however, only conjecture, but, what I can vouch 

 for, is, that a few days after the mother wasp has closed her cell stored with 

 green grubs, it contains a young larva of C. nefflecta busily eating that store, 

 whilst no trace remains of O. spinipes larva. Early in July the ,larvEe of O. 

 spinipes and of O. neglecta are to be found in separate cells fuU fed and spin- 

 ning their cocoons. As the season advances the later-stored cells appear to 

 escape the attack of C. neglecta, for in the middle of July, whilst 0. spinipes 

 is still busy in storing, there are few or no specimens of C. neglecta to be seen. 

 On the other hand, C. bidentata is now abundant, though its oviposition has 

 hardly begxin. As my observation of C. ignita throws some light on the 

 economy of C. neglecta, I shall here relate some of the facts it has brought to 

 light. On July 17th I observed a nest of O. parietum with one cell open, 

 containing a nearly complete supply of lepidopterous larvje. A Chrysis ignita, 

 flying about, settled beside the cell, and, after a brief examination with her 

 antennae, wheeled round, and, introducing her abdomen into the cell, rested 

 for about twenty seconds, doubtless in the act of oviposition. I now regi-et 

 that I did not then examine the contents of the cell in order to ascer'ain 

 the fate of Odynerus parietum's egg. Three-quarters of an hour later O. parie- 

 tum had closed the cell with the usual earthen pellets. I examined this cell on 

 the 19th, two days after, when I found a larva of C. ignita a quarter of an incli 

 long, together with several of the lepidopterous larvfe stored by the wasp, but 

 found no trace of either egg or larva of the latter. On the 23rd, six days 

 from the date of oviposition, the chrysis larva had eaten all the store and was 

 fuU-fed. I obtained evidence of its having cast its skin three times whilst under 

 observation, and from the analogy of C. bidentata I believe it had done so four 

 times altogether. The stored larvae had all been devoured, their heads alone 

 remaining, just as when eaten by the wasp grub. The larva then spun a cocoon, 

 which I knew to be typical of G. ignita. The rapidity with which it had fed up 

 was extraordiuary. None of my neglecta or bidentata fed up so rapidly, but the 

 warm sunny wall on which parietum had built her nest may partly account for 

 this, my larvje of the other two species having been kept comparatively cool. 

 Towards the end of July, spinipes and neglecta are represented only by odd 

 specimens which have survived the mass of their brethren, though bidentata is 

 still to be found somewhat plentifully. 



Chrysis bidentata when about to deposit her eggs searches for a full- 

 grown larva of O. spinipes, either just before or immediately after it has 

 spun up. 0. spinipes on the completion of her burrow fills up the moiith 

 with clay long before the most accessible cells can contain full-grown larvae ; but 

 it happens that, in a large proportion of cases, the wasp meets with some acci- 

 dent and her burrow remains uncompleted, the latter cell of which is thus only 

 protected by the diaphragm of clay which was to serve as a party wall between 

 it and the succeeding cell had the wasp lived to complete her work. Such 



