Junes, i9i6 Hessian-Fly Parasites 375 



were almost identical in size. As was the case with E. allynii, the 

 mandibles of the newly hatched lan^a appeared to be heavier, more 

 powerful, and somewhat larger than the mandibles borne by the second- 

 instar larva. Also, the head shield appeared to be more heavily chiti- 

 nized in the first instar than in the later ones. Beginning with the 

 second instar, the successive pairs of mandibles apparently increase 

 fairly uniformly in size with each molt. In the mounts where five 

 pairs of mandibles could not be found, those which were found corre- 

 spond in size and shape to some one of the pairs in the complete series 

 and it was evident that certain pairs had been lost in making the mount. 

 No more than five pairs were found in any one case. All the findings 

 lead to the conclusion that larvae of M. destructor normally pass through 

 five instars in making their growth. 



The relative sizes of the molted mandibles are shown below. The 

 measurements represent the distance in a straight line from the tip of 

 the mandible to the shoulder where it suddenly enlarges into the broad 

 base. 



Molt No. Length of mandible. 



I o. 014 mm. 



2 014 mm. 



3 020 mm. 



4 024 mm. 



5 032 mm. 



The larvae develop readily on Hessian-fiy lar\'ae and pupae, both in 

 flaxseeds and in glass cells, unless the host pupa has nearly completed 

 its development. Several newly hatched larv^ae in flaxseeds and glass 

 cells containing Hessian-fly pupae which were nearly developed killed 

 the pupae, but died from lack of sufficient food to complete their growth. 

 The larvae are evidently cannibalistic upon occasion. In one flaxseed 

 which had been exposed to ovipositing females, a young larv^a of M. 

 destructor was found which had been feeding, as also the shrunken remains 

 of another young lar\^a. Evidently the healthy larv^a had found and 

 killed the other and was feeding on the Hessian-fly lar\'a when the flax- 

 seed was dissected. Full-grown larv'se in glass cells punctured and killed 

 eggs and larvae of M. destructor which were placed in the cells with them. 

 Larvae of M. destructor were able to become full grown by feeding on 

 larvae of M. suhapterus also. 



The periods required by 36 larvae to make their growth when reared in 

 glass cells varied from 7 to 1 1 days. Cool weather appeared to make 

 growth slower. After becoming full-grown the majority of the lar\-ae of 

 M. dcstnictor reared in glass cells remained quiescent for months, though 

 still alive and able to wriggle vigorously when touched. Lar\'ae reared 

 in flaxseeds exhibited the same characteristic. In other words, the larvae 

 seem to have a tendency to estivate and hibernate until another warm 

 season before pupating. Larvae of A/, destructor were actually found to 



