8 



UNITED STATES NATIONAL MUSEUM BULLETIN 224 



egg and tamp the whole with the end of the abdomen 

 and the outstretched spines. Melin (1923) says that 

 Philonicus Loew in the Asilinae uses the spines as a 

 sort of brush. 



It will be noted that the larger the fly the larger 

 and stouter the spines become ; thus in small and weak 

 flies like Brachyrrhopala Macquart and Stichopogon 

 Loew they are slender and delicate. I attach no great 

 significance to the supplementary setae that are present 

 in some of these smaller genera, and in some therevids. 

 They are found only in the smaller, weaker flies and 

 are found laterally and ventrally to the spines. Per- 

 haps they represent the type of bristle that has become 

 modified into these so-called ovipositoral spines. 



In summary, these five tribes of the Dasypogo- 

 ninae — the Dioctrini, Phellini, Chrysopogonini, Dama- 

 lini, and Laphystini — seem never to have acquired 

 these ovipositoral spines, except rarely and always in 

 an analogous form, and are here regarded as general- 

 ized Asilidae. Together these tribes comprise 28 per- 

 cent of the Dasypogoninae genera. 



The second important acquisition in the subfamily 

 Dasypogoninae is the anterior tibial apical spine. This 

 structure is not a true morphological spine, but is a 

 stout, thickened or swollen bristle, which has become 

 spinelike ; it is appropriate to refer to it as a spine to 

 avoid comparison with stout setae or bristles often 

 present on these flies. However, in some genera it may 

 be accompanied by or borne on a spinelike process or 

 outgrowth from the apex of the tibia. 



In at least the larger forms and possibly in all asi- 

 lids which show it, this spine appears to be an adaptive 

 mechanism, possibly useful as a small, raptorial clamp 

 in holding straggling prey until the proboscis and 

 its quieting juice can be inserted. These spines may 

 have developed at a time when the salivary juices 

 of the fly were less potent in their paralyzing and 

 stupefying effect. This spine may be polyphyletic in 

 origin. There are several distinct types as well as de- 

 grees of development. Once emerged, it is not un- 

 common to find three coincident types of adjacent and 

 associated structures. The first is the extension of the 

 base of the spine into a strong protuberance, reaching 

 a maximum in Molobratia Hull (Dasypogon of au- 

 thors). The second is the coincident appearance of 

 opposing denticles on the base of the anterior basi- 

 tarsus. These are frequently found but may be quite 

 absent. The third type is the swelling of the basitar- 

 sus, with or without denticles, into a conspicuous flange 

 or mound, reaching a peak in Neolapams Williston, 

 Megapoda Macquart, and allies. 



Because the foretibial spine itself is of several types, 

 as well as in several stages of development, it has been 

 criticized as a distinguishing character. The follow- 

 ing major types of such spines are recognizable: 

 First, the slender, twisted or sigmoid spine as is found 

 in Cophura Osten Sacken and its allies; it has little 

 and sometimes no corresponding basitarsal develop- 



ment. The second type is the stout, well developed 

 spine with at least a moundlike, denticulate, basitarsal 

 swelling. The third type is a stout, straight spine at- 

 tached to a strong protuberance which extends beyond 

 the tibial apex with correspondingly exaggerated basi- 

 tarsal process as above cited. These second and third 

 types may represent continued specialization from the 

 same stock as Cophura Osten Sacken, the cophuroid 

 genera possibly having arisen from ancestral forms, 

 which have been retarded ; or at least arisen from those 

 which have retained a generalized form of the spine. 

 Another view is that these several types of spines in 

 the Dasypogonini represent separate polyphyletic or 

 analogous developments of the foretibial spine. 



Thus far I have had reference to the several types 

 of anterior tibial spine only as seen within the tribe 

 Dasypogonini. I have relegated Codula Macquart to 

 the Chrysopogonini on account of its notopleural 

 spine. Members of this genus lack completely any 

 trace of the tibial spine, from which I conclude that 

 the other genera in the tribe acquired the tibial spine 

 independently. Again, Leptarthrus Stephens, with a 

 small, bent, tibial spine and no trace of ovipositoral 

 spines has been assigned to the Dioctrini. Here again 

 it is believed that these two exceptions may possibly 

 represent parallel acquisitions of the tibial spine. If 

 this is true we may consider that the tibial spine has 

 arisen three or more times in the Dasypogoninae, and 

 once in the Megapodinae. 



In this study the Dasypogoninae are divided into 

 ten tribes. The Dioctrini, with all cells of the wing 

 generally open and with a trace of the antepenulti- 

 mate palpal segment, is placed lowest, and followed by 

 the Phellini, the Chrysopogonini, the Damalini, and 

 the Laphystini. The Stichopogonini are considered 

 nearest the Dasypogonini. The spined and not spined 

 genera of the higher Dasypogoninae have been sepa- 

 rated into the Dasypogonini and the Stenopogonini. 



I reject the tribal name Saropogonini used by some 

 authors and based on the spined, front tibia, because 

 the name should be Dasypogonini. Some but rather 

 less than half of the unspined Stenopogonini can be 

 separated in the male sex by the not rotate terminalia. 



In any case, the presence or absence of the spine is 

 extremely useful in classification. Megapoda Mac- 

 quart and its allies have been raised in this study from 

 tribal status to subfamily rank. I am inclined to re- 

 gard the Laphystini as the highest of the not spined 

 Dasypogoninae, because of their reduction in the seg- 

 ments of the abdomen, but it is possible that the Dama- 

 lini are more specialized. They are also a very suc- 

 cessful group, and, together with the Stichopogonini, 

 they dominate such habitats as sandy shores and mud 

 flats. In the higher Dasypogoninae, some 75 genera 

 lack foretibial spines, while approximately 50 genera 

 have acquired them. 



The principal line of cleavage in the Laphriinae has 

 to do with the chitinization of a medially fused and 

 complete postmetacoxal arch. It separates the Atomo- 



