MyniuUon formicarms from the, Pupa. 55 



giving one the impression that it had been stung and 

 paralysed at this juncture. Only once did I notice sand 

 being thrown about by the larva, and even on that 

 occasion I could not see for what purpose this was done. 

 It should be stated that a trap was never well formed — 

 possibly because the sand supplied was too mobile for the 

 purpose. Often the larvse would wander about during the 

 night and make circular furrows in the sand, which were, 

 there is little doubt, intended for traps. While in my 

 possession the larvae seemed to be usually buried well 

 beneath the surface of the sand. 



At length one day threads of a gummy nature were 

 noticed along the sand and across the box containing the 

 larvse — this waste of material seeming to be ill able to be 

 spared by so small an insect which had to produce so large 

 an imago. Gradually there arose from the surface of the 

 sand a spherical cocoon consisting of grains of sand 

 cemented together by this gummy material (liquid silk). 

 This cocoon was formed from below, a small section of 

 a sphere first appearing, the sphere itself taking shape 

 as the work progressed, and being when completed about 

 nine-sixteenths of an inch in diameter. Having but the 

 one cocoon, it was not possible for me to examine the 

 pupa within. When a little later the emergence took 

 place I was pleased to find that the imago was of good 

 size, not having suffered apparently from the long fast in 

 the larval stage. 



In the case of the Chrysopas (Green Lace-wing Flies), 

 which are near relatives of the ant-lions, we have a tiny 

 spherical cocoon, quite ethereal in appearance and evidently 

 made of silk alone. Before disclosing the imago the pupa 

 of Chrysopa neatly cuts for itself a circular door. It then 

 leaves the cocoon entirely and afterwards the disclosure of 

 the imago takes place outside. The pupa of Myrmcleon 

 formicarius does not do this, but after making an aperture 

 at the top of the cocoon, protrudes the fore part of its 

 body only, somewhat in the same way as the pupa of 

 the Burnet-moth {Zygxna trifolii) does. Emergence of 

 the imago then takes place from a dorsal slit, the delicate 

 pupal skin remaining half-projecting from the orifice in 

 the cocoon. 



As previously mentioned interesting changes take place 

 in the mandibles as the ant-lion proceeds from stage to 

 stage of its life-cycle. These I have illustrated by means 



