304 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxi. 



more narrow and liable to be flat or even somewhat hollowed inward. 

 The form of the eighth (terminal),plate is iisuall}^ more or less trian- 

 gulai- and tlic outline of its posterior edge is of systematic value. 

 Tufts of hairs are frequentl}^ present at the sides of the hinder plates 

 beneath, and in Isodontia rows of stiff hairs along the posterior margins 

 of these plates ai'c characteristic of the subgenus. In some species 

 the fourth and fifth ventral segments each have a median area densely 

 sericeous in nature and usually of a dark color. The genitalia of the 

 male often protrude somewhat between the last dorsal and ventral 

 plates and in some species are so large as to show their structure quite 

 well even when drawn in as far as possible. 



Wings. — The wings are quite large, and, though sometimes hyaline, 

 are usually more or less colored, either in part or entirely, as already 

 described. It has seemed best in this paper to follow the nomencla- 

 ture of the veins and cells used by Cresson and others, but drawings 

 have been included which name the parts according to the Comstock 

 system, these having been obtained through the kind assistance of 

 A. D. MacGillivray, of Cornell University.' (Plate VII, hg. 8;' Plate 

 VIII, tig. 10.) 



Fore wings. — The radial cell is elongated, rounded at its outer end. 

 Separating it from the costal celh is a well-developed stigma. Three 

 closed cubital cells are present, except in cases of abnormal venation, 

 lying between the costal cell, the stigma and the radial cell in front, 

 and the first and third discoidal and second apical cells behind, 

 the vein between these last and the cubital cells being the cubital 

 vein. Of the three closed cubital cells the first is uuich the largest. 

 The second and third vary in size according to the position of the 

 transverse cubital veins which separate them. In some subgenera the 

 first and second transverse cubital veins, which run approximately 

 pai-allel, are so near each other that the second cubital cell is much 

 longer ])etween the radial and the two discoidal cells than it is in the 

 other direction, a condition usually expressed as "higher than })road." 

 In P)'ofc/-osj>/itM'this cell has about the same diameter in each direction, 

 while in Jsodoidia the breadth tends to be notici-ably greater than the 

 height. The thii-d cubital cell is rougldy triangulai- in form, the third 

 transverse cubital vein passing at first oblitiuely outward and forward 

 from the cuVntal vein, then l)ending inward and joining the radial vein 

 not far from where the second transverse cubital vein unites with the 

 latter. Two of the three cells immediately behind the cul)itals are 

 closed and are termed the "'first and third discoidal cells." while toward 

 the tip of the wing from the last named is the unclosed second apical 

 cell, which lies posterior to the outer portion of the thii'd cubital cell. 

 Sepai'ating the two (first and third) discoidal cells and the second apical 

 cell are two recurrent veins, the first of which arises posteriorly from the 

 anterior outer angle of the second discoidal cell, which lies posterior 

 to the first discoidal cell, the second recurrent vein arising from the 



