Dec, 1905.1 Fall: Affinities of the Genus Tachycellus. 171 



As thus constituted the genus includes cognatus Gyll — placed as a 

 Bradycellus in our lists — and several allied forms, all of which 

 Ganglbauer has properly transferred to Dichirotrichus Duval on the 

 basis of the presence of a seta in the hind angle of the prothorax, 

 creating for them the subgenus Trichocellus, typical Dichirotrichus 

 having the male front tarsi pilose rather than squamulose beneath, and 

 the upper surface of the body conspicuously pubescent and rather 

 closely punctured throughout. The latter author regards Tachycellus 

 as merely a subgenus of Bradycellus, separable by its squamose middle 

 male tarsi, these being simple in the true Bradycellus ; a single small 

 species {similis Dej.) is referred to it. 



What now is the bearing of this upon the disposition of our own 

 species ? The interpretation of Seidlitz is totally different from that 

 of either Le Conte or Horn, and would exclude all the species we 

 now refer to the genus, substituting for them cognatus alone. The 

 view of Ganglbauer is more nearly in accord with that of our own 

 authors, and if we reverse in order of importance the antennal and 

 tarsal characters, the desirability of which is indicated above from a 

 study of our own species, the two become practically identical. The 

 question as to whether the differences in the modification of the male 

 tarsi are to be regarded as of generic orsubgeneric import is of course 

 a matter of opinion ; Ganglbauer holds the latter view ; I am inclined 

 to the former. Such characters are commonly held to be of rather 

 more than ordinary importance, and are almost invariably associated 

 with others which, if somewhat trivial in themselves, assume an added 

 importance in conjunction with the former; moreover, in the present 

 case, clearness of exposition is certainly facilitated thereby. 



Let us pass briefly in review the species of Tachycellus as they now 

 stand in the Henshaw list. It is first necessary to remove nebulosus. 

 An examination of the LeConte type during a recent visit to Cam- 

 bridge shows it to be a typical Bradycellus, closely resembling and 

 probably not distinct from rupestris. Nitidus is a very aberrant form 

 and cannot be retained in the genus, the only reason for so associating 

 it being the nearly glabrous third joint of the antennae. It is more 

 nearly related to Bradycellus, but does not properly enter there and 

 the proper course would seem to be to restore Casey's genus Glycerins 

 for this and several allied forms to be herein described. 



The remaining species agree in possessing the fundamental tarsal 

 characters of the genus, and all, with the exception of nigrinus, have 



