30 CONNECTICUT GEOL. AND NAT. HTST. SURVEY [Bull. 



becomes increasingly more pronounced, until in such a fly as the psilid 

 shoAvn in Fig. 5, T, the Aentral prolongation of the postpedicel re- 

 sults in the formation of an extremely elongated segment bearing the 

 arista in the region which was originally terminal, but now has the 

 appearance of being dorsal or subbasal in position. 



Some of the segments which normally enter into the composition 

 of the arista apparently unite with the third antenna! segment, in- 

 stead, in some forms, as is the case in the stratiomyid shown in Fig. 

 5, K, in which the apparent third segment is divided into four "an- 

 nulations*', which probably represent four uniting segments. The aris- 

 ta of the South American tachinid shown in Fig. 5, S, is rather inter- 

 esting from the fact that its terminal segment is greatly enlarged. 

 The row of long branched hairs borne on the arista of the tsetse fly 

 shown in Fig. .5, W, gives it a verv peculiar appearance in the antenna ^ 

 of this fly. 



From the morphological standpoint, it is very unfortunate that 

 the former incorrect method of referring to the number of antenna! 

 segments is still retained by many modern dipterists, despite the f act , 

 that many of them have abandoned the old division of tlie Diptera into : 

 the suborders Nematocera and ''Brachycera" {sensu lato), which was ! 

 based upon a misconception of the number of segments composing the 

 antennae. When any of the "Brachycera" were found to have more 

 than three antenna! segments, those segments distal to the third were 

 formerly interpreted as secondary "annulations" of the tliird segment, j 

 and the small segments visible in the arista, etc., of some Diptera were i 

 regarded as subdivisions of an appendage of the supposedly three- j 

 segmented antenna. The number of antenna! segments in the brachy- i 

 cerous le^^tid (or rachicerid) fly, Rachicerus (Fig. 5, M), however,' 

 may be as liigh as 28, while in the great majority of the Nematocera 

 the number of antenna! segments is mucli less, and only a very few 

 Nematocera equal or surpass Rachicerus in the number of segments 

 found in the antennae. There are ten very distinct and well developed 

 segments in the antenna of such a stratiomjdd as the one shown in 

 Fig. 5, P, and no amount of arguing could convince an unbiased 

 morphologist that the well developed, separate segments beyond the 

 third are merely "annulations'' of the third segment. Even when tlie ' 

 terminal segments of the antenna are reduced to form a style, as in the • 

 asilid shown in Fig. 5, G, or an arista, as in the oscinid shown in Fig. . 

 5, Q, the segments can frequently be clearly seen, and it is morpholog- 

 ically incorrect to speak of such an antenna as three-segmented, al- 

 though it may be convenient to do so in taxonomic descriptions. 



7. Mouth parts 



The labrum, Zr, or upper lip, is morphologically a portion of the' 

 head capsule, while the true mouth parts (i. e., the mandibles, maxillae,, 

 and labium or united second maxillae) are highly modified limbs off 

 the gnathal segments entering into the composition of the head re-;- 

 gion, but since the labrum is functionally one of the mouth parts, it; 

 may be considered under this headinof. 





