6 



UNITED STATES NATIONAL MUSEUM BULLETIN 224 



fully fused and complete laterally. Moreover, there- 

 vids have a bristle-type empodiuin, which, together 

 with venational similarities, suggests that they may 

 not be far from the asilid line of ancestry. 



As far as the general sequence of events with regard 

 prosterna, I draw the following conclusions: 1, The 

 complete presternum is generalized and ancestral in 

 asilids and probably in all Brachycera; 2, the need 

 for strength of muscle attachment in asilids coping 

 with vigorous, struggling prey has perhaps retarded 

 the definite trend towards a breakdown of the pro- 

 sternum ; and 3, the presternum is in process of being 

 broken down in most if not all groups of asilids today, 

 because of the premium on flexibility of the anterior 

 subraptorial legs. This conclusion is strengthened 

 by the fact that within the same genus the larger the 

 asilid, of those asilids which have a reduced pro- 

 sternum, the greater, proportionately, is the amount of 

 membrane surrounding the presternum. I further 

 point to the condition in Coenomyia Latreille, illus- 

 trated in text-figure 1, which shows the presternum at 

 a point of just complete separation from the lateral 

 propleuron. 



As regards the conclusions to be drawn from the 

 state of the presternum with reference to the phylo- 

 geny of groups within the asilids themselves, one must 

 proceed with caution, because of the widespread tend- 

 ency within a family for suprageneric groups to proceed 

 along parallel lines. The presternum is most variable, 

 at the present time, in the Asilinae, as observed by 

 Clements (1951), but here again it appears to be com- 

 plete in the more primitive elements, except that it is 

 not clearly possible usually to point to definitely prim- 

 itive genera. For instance, Nerax, new genus, might 

 be regarded as specialized with respect to behavior of 

 the veins at the apex of the wing, yet the arrangement 

 of the anterior branch of the third vein, and the fre- 

 quent vestiges of former connecting veins, complete in 

 some related forms, represent a primitive trait. In de- 

 termining the relationships of genera and groups of 

 genera, the total picture of their development as it 

 emerges from a study of the whole insect must be taken 

 into consideration and I fully agree with Clements that 

 the presternum cannot be used, except in indicating the 

 affinities of a genus. I have used it only once in this 

 study to differentiate between genera, for instance, to 

 separate the African Leptogastrinae genus Ammophilo- 

 mima Enderlein, from the Asiatic Lagynogaster Her- 

 mann, and then only in connection with other char- 

 acters. Broadly speaking, the character of the pro- 

 sternum does help to define the tribes Laphystini and 

 Atomosini because it is universally complete in these 

 groups, or at most shows but slight emargination in the 

 former tribe. 



In this work the family Asilidae is subdivided into 

 five subfamilies: Dasypogoninae, Megapodininae, La- 

 phriinae, and Leptogastrinae, which we hold to consist 

 of one interrelated group, and lastly, the Asilinae, which 

 is considered to be somewhat more distantly related to 



the other four. In point of size in terms of species 

 (the species count as given is approximate), the bulk 

 of all Recent robber flies fall into the Dasypogoninae 

 and Asilinae. 



The Dasypogoninae contain about 1,837 species, 

 whereas the Asilinae contain about 1,852 species. In 

 the Megapodinae are 33 species, a few of which older 

 names may be synonyms. The Leptogastrinae contain 

 about 252 species, to which 24 others have been added by 

 Martin, 1957. In the Laphriinae there are about 763 

 species. 



The Dasypogoninae are here divided into the fol- 

 lowing tribal associations: Dioctrini with about 112 

 species, Phellini with 7 species, Chrysopogonini with 

 18 species, Damalini with 119 species, Stichopogonini 

 with 141 species, Thereutrini with 7 species, Enig- 

 momorphini with 1 species, Laphystini with 93 spe- 

 cies, Stenopogonini with about 829 species, and the 

 Dasypogonini with about 510 species. 



The Laphriinae are divided into the tribes of 

 Laphriini with 446 species, the Andrenosomini with 

 117 species, the Atomosini with 185 species, and the 

 Ctenotini with 15 species. 



I call further attention to the proportions within the 

 Laphystini, where the genus Laphystia Loew has 42 

 percent of the species in this tribe. In the Damalini 

 we find Damalis Fabricius in the Old World with 43 

 species and Holcocephala Jaennicke in the New World 

 with 40 species. Collectively they total 69 percent of 

 the tribe or about the same proportion as Leptogaster 

 Meigen in its subfamily; divided they each total 33 

 percent and 35.8 percent of the totals within the tribe. 



The Megapodinae in this work have been treated as 

 a subfamily. 



It is of interest to note that the large genus Laphria 

 Meigen contains about 254 species or about 33 percent 

 of the subfamily Laphriinae; and that in the tribe 

 Atomosini, the large genus A tomosia Macquart contains 

 58 species or about 33 percent. Again, the tribe Andre- 

 nosomini contains the large genus Andrenosoma Ron- 

 dani, which is about 37 percent of the species. 



However, we find that in the subfamily Leptogas- 

 trinae with about 276 species there is a single, large 

 genus Leptogaster Meigen with about 210 species or 

 almost 76 percent of the total. It is not, however, sur- 

 prising to find some genera of flies unusually large, as 

 for example Anopheles Meigen, Tabanus Linne, Li- 

 monia Meigen, and it does not necessarily follow that 

 they should be broken up into numerous genera and 

 subgenera. Sometimes the recognition of well marked 

 and distinct species groups serves just as useful a pur- 

 pose and may more accurately reflect the true state of 

 affairs. 



The Asilinae are divided into the tribe Ommatini 

 with approximately 187 species and the tribe Asilini 

 with about 1,665 species. 



In this work the Dasypogoninae, with their open, 

 marginal cell and 2-segmented palpus, are regarded as 

 the most generalized members of the family. Two other 



