WILLISTON: APIOCERID^. II5 



oc-erus, Laphria, Lampria, 2^x1^ Dasyllis, the palpi are distinctly two- 

 jointed. In Leptogaster, there is but a single joint. This character 

 may be expressed as follows : 



A. Palpi one-jointed. 



B. Marginal cell of the wing open — Leptogastrin^. 

 BB. Marginal cell closed — Asilin^. 



AA. Palpi two-jointed. 



C. Marginal cell of the wings open — DasypogoninjE. 

 CC. Marginal cell closed — Laphrin^e. 



The diminution of the number of palpal joints has been produced, 

 evidently, by a greater specialization, and the general tendency in 

 diptera is toward their entire loss. It may be safely assumed that the 

 Asilinae and Leptogastrinae represent a more specialized type than do 

 the Dasypogoninae and Laphrinae ; a specialization further seen in 

 the change from styliferous to the aristoid antennae. On this ground 

 alone, it seems to me, the relationship of Apiocera with the Dasypo- 

 goninae is more intimate than with the Asilinae. It is true that we 

 find the palpi sometimes wanting in families of lower type, even as 

 they may be wanting in Saundersia, a genus closely related to Dejaenia, 

 where they are very large. On the other hand, I predict that two- 

 jointed palpi will be found very seldom in those genera having a well- 

 developed arista. 



In MallopJiora, the palpi are very stout, cylindrical and hairy, and 

 nearly as long as the palpifers. In the other genera of the Asilinae 

 they are smaller. In Laphria and Lampria they are small, one might 

 almost say sometimes rudimentary in the latter. Everywhere in the 

 family their size seems to have some relation with the hirsuteness of 

 the species. In many of the Dasypogoninae they are strongly devel- 

 oped, though perhaps nowhere so much as in MallopJiora. In all 

 cases they seem to be appendages of the palpifers, to which they are 

 attached without articulation. The palpifers are in all cases stout, 

 chitinous rods, much larger and stouter than in any of the forms pre- 

 viously described.. Generic, and even specific, differences are shown 

 in their shape, sometimes slender on the distal extremity, sometimes 

 hooked or barbed, or with angular projections like a Roman battle- 

 axe. They indicate a more distinct function than appears in the more 

 simple forms occurring in the families already mentioned. The galea 

 is strongly chitinous, with its base only, membranous. The terminal 

 portion, corresponding to the labella, is variable in shape, sometimes 

 extending far back on the upper side, and separated from the portion 

 behind it by a more translucent line, as indicated in the section shown 

 in the plate. The labella, however, are always horny, not at all 



