32 AMERICAN COLEOPTERA. 



O^ THE SPECIES OF BElIBIDir^VI OF AMERICA 

 NORTH OF MEXICO. 



BY ROLAND HAYWARD. 



The present paper is the result of a study, extending somewhat 

 over three years, of the species of Bemhidium occurring in the 

 United States and Canada. This is one of the few remaining genera 

 of the family Carabidse which has not been revised in recent years, 

 the latest work upon the subject being a division of the genus into 

 groups by the late Dr. LeConte (Proc. Acad. Phil., 1857, p. 3). 

 Seventy-five species are enumerated in that paper ao known to him, 

 of which sixty-five are here regarded as good species. Since that 

 time a large number have been discovered, others then unrecognized 

 have been identified, and there are many in collections which are 

 here described for the first time, so that the number which I have 

 been able to verify is one hundred and twenty-four, while there are 

 about a dozen described from North America by Mannerheim, Mots- 

 chulsky and Chaudoir which remain unknown to me. 



It has, therefore, been impossible for the collector to identify his 

 species without reference to the larger typical collections, especially 

 as the descriptions are so scattered as to be practically inaccessible 

 to the majority of students, wdiile many of them are so meagre as to 

 afl^ord almost no assistance. I have here attempted to clear up some 

 of this confusion. 



The genus Bembidium, as at present constituted, contains a very 

 large number of species, which present a considerable diversity of 

 form. Many of these diflPerences, particularly in the form and struc- 

 ture of the mouth-parts, are such as would be regarded as generic 

 in other parts of the family. As a natural consequence, numerous 

 eflx)rts have been made, especially by the older writers, to subdivide 

 the genus generically, so that, as a result, a number of groups have 

 at various times been raised to the rank of genera. Many of these 

 were based upon entirely superficial characters, some of them of ab- 

 solutely no value even for the definition of groups, while the char- 

 acters of nearly all of these so-called genera have gradually become 

 evanescent with the discovery of new and intermediate forms, so 

 that one by one they have been re-united with Bembidium. For an 



