348 AMERICAN HYMENOPTERA. 



queens, nine workers and one male. I have dissected the 

 male and compared its genitalia carefully with those of many 

 other specimens. The other nest contained eight queens 

 and thirty-three workers. This is the gentlest and least ready 

 to sting of all the bumble-bee species which I have had to 

 deal with in the living condition. This seems peculiar, as B, 

 vagans, which seems to be its nearest ally, is exceedingly 

 ferocious. 



The females of this species can readily be separated from 

 those of all its nearest allies by the dark lower portion of the 

 pleura, and the males can readily be distinguished by the 

 general coloration already described and by the strongly 

 protruding nuces of the genitalia. In collections, the fe- 

 males have long been confused with those of vagans and 

 af finis. 



Had I never taken a male in the nest, I should have said 

 that these males and females should go together as they 

 have the same range of habitat, are of corresponding abund- 

 ance throughout that habitat, belong to the same group and 

 are most closely related, as shown both by coloration and by 

 structure, to the females and males respectively of the same 

 species, and as there are no other females or males, belong- 

 ing to the same group and within the same habitat, which 

 have not been satisfactorily mated. 



Cresson, in his original description of the type male, says 

 that the third dorsal abdominal segment had "a slight ad- 

 mixture of black." All the males having the third segment 

 noticeably darker (though yellow for the most part) than the 

 first two, which I have dissected, have had vagans genitalia 

 and, for this reason, I think it not impossible or even improb- 

 able that the type or perpiexus was a male vagans. If this 

 surmise is correct, then this species should take the name 

 hudsonictis and the name perplextis should go into the syno- 

 nomy of vagans. As I have not seen the type specimen in 

 question, however, I will let this species stand as perpiexus 

 until we get more certain knowledge about the matter. 



I have seen several queens in the Harris collection which 

 had been named Bombus elatus F. by Harris. 



