ROLAND HAYWARD. 193 



The genus Tachys exhibits a considerable range of variation 

 several groups having at different times been elevated to the rank 

 of genera. As elsewhere expressed with regard to Bembidium,, it is 

 my opinion that no satisfactory results can be obtained in generic 

 subdivision until all the species of the genus, at least, shall have 

 been studied, if not, indeed, all those of the tribe Bembidiini. 



As here recognized, the genus includes all the members of that 

 tribe in which the anterior tibire are obliquely truncate at apex, 

 Pericompsus, which differs from Tachys only by the deep, entire 

 marginal stria of the elytra, being regarded as a group of the latter 

 genus. 



Several other characters not seen in Bembidium may be men- 

 tioned, which, although not possessed by all the species, are observa- 

 ble in most of them. 



Among these the sutural stria of the elytra in nearly all is deeper 

 posteriorly and presents the appearance of being recurved at apex, 

 the recurved portion being parallel either with the margin or with 

 the suture, in the latter case often hooked at tip. This is due to the 

 fact that the other dorsal strue are usually feebly impressed and 

 almost or entirely effaced at apex, what appears as the recurved 

 portion of the sutural stria being in reality the remnant of the 

 apical portion of the fifth, seventh or, in one group, probably the 

 third stria. In those species in which the elytral stria? are more 

 distinct and entire, or nearly so, this character is less marked, and 

 in anescens is entirely lost. 



The scutellar stria of the elytra, which is present in at least all 

 our North American species of Bembidium, is entirely absent in 

 those of Tachys, and as Duval gives this as one of the distinguishing 

 characters of his group XV, which is the equivalent of Tachys, it 

 probably applies to the exotic species. 



In most Tachys the second and third joints of the antenna? are 

 subequal. Rarely, however, the second is longer than the third. In 

 Bembidium the reverse is the case, though in a few species the third 

 is but slightly longer than the second. 



Most of the characters employed in the system here proposed 

 have been long known and have become the common property of 

 science, some are brought into greater prominence than heretofore 

 in the grouping of species, while a few, so far as I am aware, are 

 here suggested for the first time. No claim is made that the system 

 will apply to the large number of exotic species that have been de- 



TKANS. AM. KNT. SOC. XXVI. (25) JANUARY, 1900. 



