390 



must be borne in mind when using the keys to this stage. Bombylius 

 is parasitic, or, speaking more precisely, inquihnous, in the nests of 

 the bee genus Andrena, and in the larval metamorphosis from an 

 active initial instar to a series of inactive instars presents an analogy 

 to Meloc, a beetle which lives under similar conditions. Structurally 

 the larvae of Bombyliidae closely resemble those of Asilidae, but, as 

 far as my limited material permits me to detei-mine, may be separated 

 from the larvae of that family by the much smaller head with a slight- 

 ly chitinized dorsal projection which is flattened anteriorly as shown 

 in Figure 2, Plate LVI, by the normally crescentic form of the entire 

 body, the head and the apex of the abdomen being slightly recurved 

 ventrally, and by the absence or reduction in size of the thoracic and 

 apical abdominal hairs. I have been able to detect the presence of 

 thoracic hairs in some larvae of Bombyliidae, and Chapman has re- 

 corded their presence in the European species Bomhylius major. He 

 makes no mention, however, of apical abdominal hairs, and Nielson 

 distinctly states that neither thoracic nor apical abdominal hairs are 

 present in Bombylius minor. I have found no apical hairs on any of 

 the larvae. The spiracles are small, the anal pair located upon the 

 penultimate abdominal segment and difficult to distinguish ; the lateral 

 abdominal pairs are absent. 



Pupa. — Like the larva, the pupa resembles that of the Asilidae in 

 its general appearance, but as far as I have seen there is invariably 

 present upon the central portion of the head-capsule, towards its lower 

 margin, a pair of stout thorns which are in some species very closely 

 approximated, or even fused except at apices. The only exception I 

 know of is Toxophora virgata as described by Townsend, but the 

 armature of the abdomen is typical of this family. Most of the 

 Asilidae which I have reared or have before me have at least 3 strong 

 thorns upon each antennal sheath, whereas I have found no bombyliid 

 that has more than 2. This character can not be depended upon for 

 the separation of the families, however, as Lcptogaster flaznpes, an 

 asilid, has no thorns on the antennal sheath. 



Imago. — The imagines of this family should be readily located in 

 the family by the use of the synoptic key on a previous page. Willis- 

 ton's "Manual" presents a very good key to the genera. 



HABITS OF THE IvARVAE 



All the species that are known in the larval stages are predaceous, 

 parasitic, or inquilinous. A brief summary of the species attacked by 

 the different genera follows. Argyramocha occurs in the nests of 



