26 CROP PEST COMMISSION OP 



containing a large-sized pupa were taken one hundred and ten 

 wings and forty-eight heads of horseflies, but no other insects. 

 The heads in every case were those of females. 



On July 5, I counted the remains of insects taken from a 

 burrow which contained a partially grown larva. This count 

 showed forty-eight wings of horseflies and two wings of a large 

 mosquito (Psorophora ciliata}. Two specimens of the horseflies 

 were Tresh and had every appearance of having been but re- 

 cently brought to the nest. 



On July 14, I unearthed a pupa in the southwest corner of 

 the Station field, and found with it one hundred and twenty-two 

 wings of horseflies and two of another fly (Pseudopyrelia cor- 

 mcini). 



Another pupa taken in the same place had in the burrow 

 v/ith it forty-eight wings of horseflies and six of the screw- 

 worm fly. 



I also took a male horsefly in a good state of preservation 

 from a burrow containing a partially grown larva. This is the 

 orly instance where the male horsefly was observed among re- 

 mains found in the nests. 



Once in' a while nests were found in which the larvae were 

 ciead; in at least one instance it appeared as though a certain 

 species of little red ant* was responsible for this condition, for 

 numbers of them were feeding in the nest, both on the dead 

 l^rva and on the remains of insects found with it. The ant was 

 determined for me by Prof. Wm. M. Wheeler, of the American 

 Museum of Natural History. 



The nests are not stored and sealed up as is the case with 

 f:orne wasps, but the flies are carried in by the female as they 

 *;re needed for food. It appears that the burrow is opened, flies 

 carried in and the burrow closed again. How often this opera- 

 tion is repeated I did not discover, but larvae of all ages, as 

 well as pupae, were taken from closed burrows, and in some 

 ol the closed burrows nearly fresh horseflies were found, in- 

 dicating that the opening had just been filled. When the bur- 

 rows lending <lo\vii to the larva* or pupa 3 were closed they could 

 IK- followed by the color of the material iised in filling, as this 



