ROLAND HAYWARD. 37 



As compared with other Carabidpe, the species of Bembidium are 

 decidedly beh)w the average size, the length varying, ronghly speak- 

 ing, from slightly less than one-tenth to about seven-twentieths of 

 an inch, or from two to nine millimeters. Most of them are riparial 

 in their habits, occurring under stones and refuse near the water's 

 edge along streams, the shores of ponds, or on the sea-shore. Some 

 are found under bark, amongst old leaves, or in moss, while a few 

 occur almost everywhere. 



With regard to the characters which have been made use of in 

 grouping the species, but little need be said. 



The eyes are usually large or moderately large and prominent. 

 In a few species, B. obloiigulam, B. laticeps and their allies, they 

 are small and flattened, a character which enables us to isolate them 

 from other Bembidia. 



The grooves on the front between the eyes, whether single or 

 double, and whether parallel or convergent, are of use in the sepa- 

 ration of groups. 



The mouth-parts, although affording us the means of defining the 

 tribe Bembidiini, have otherwise yielded but few characters. They 

 vary to a great extent within the limits of the genus, but the differ- 

 ences, as a rule, become evanescent. The outer lobe of the maxill?e, 

 whether biarticulate or not, and the form of the mentum tooth have, 

 however, been of service. 



The surface of the head, whether alutaceous or smooth, is of value 

 only in the separation of species, and even then must be used with 

 care. 



The thorax, whether pedunculate or not, as well as by its form 

 and the f )veate impressions at the base, whether uni- or bistriate, 

 has afforded several useful characters. 



The position of the dorsal punctures, whether situated on the third 

 stria of the elytra or on the third interval, is of very great impor- 

 tance, while their number is also of use in the definition of groups. 

 In nearly all our species they are confined either to the third interval 

 or to the third stria. In B. Icevigatiim and B. semistriatiim, how- 

 ever, each representing a separate group, there are irregular rows 

 of small setigerous punctures on all the intervals. In the table of 

 groups which follows these are placed, together with those in wliich 

 the punctures are confined to the third interspace, in the category : 

 "third interval with dorsal punctures," in contra-distinction from 

 those in which they are situated on the third stria. 



TRANS. AM. ENT. SOC. XXIV. FEBRUAKY, 1897. 



