370 



on the penultimate segment, which in some Asilidae is very short and 

 appreciably sunken. In Mydaidae and Asilidae there are very small 

 functionless (?) spiracles on the mctathoracic and basal 7 abdominal 

 segments. In Bombyliidae the head is much less heavily chitinized 

 than in the other families, but the presence of the anal spiracles on the 

 penultimate segment indicates the relationship with the others. The 

 abdomen has no distinct spiracles except the apical pair, and the apical 

 segment has no hairs. 



Pupa. — The armature of the head is usually sufficiently distinct 

 from that of any other brachycerous family to distinguish the pupa, 

 as no other possesses strong thorns on the antennae such as are 

 shown in the figures for three of the families included in Asiloidea 

 (PI. LIU, Fig. 3 ; PI. LIV, Fig. 7 ; PI. LVI, Fig. 3). The only genus 

 of this superfamily known to me that does not have these thorns 

 is Lcptogaster, but here the characteristic abdominal armature of the 

 Asiloidea — consisting of strong curved thorns alternating with much 

 weaker bristles in single transverse series — readily separates the pupa 

 from those of Empidoidea. The pupae of Therevoidea have only one 

 thorn on the base of each antenna and in some other respects differ 

 from those of Asiloidea. 



HABITS OF LARVAE 



The larvae are either predaceous or parasitic, and may be con- 

 sidered beneficial — with the exception of those that are parasitic upon 

 parasites of injurious insects, such as Tiphia, the parasitic enemy of 

 white-grubs. The habits of the various genera are dealt with under 

 the different family headings. 



HABITS OF IMAGINES 



The imagines of all Asilidae known to me are predaceous, feed- 

 ing upon other insects. Some authors have recorded Mydaidae as 

 feeding upon insects, but I have not seen the species known to me 

 doing so. The imagines of Bombyliidae are, so far as I know, flower- 

 frequenters, feeding upon nectar. 



Family MYDAWAB 



The insects constituting this family have been considered as enti- 

 tled to superfamily rank by Osten Sacken. Yerrall has conceded their 

 claim to separation from the other superfamilies in Brachycera, but 

 links the Scenopinidae with them under the superfamily name 

 Dermatina. Brauer, in one of his earlier papers, attempted to associate 



