HENRY J. FRANKLIN. 475 



Habitat. — I have records of this species as follows : Maine 

 (Orono), Massachusetts (Amherst — a single 9 ), New Hamp- 

 shire (Webster, Durham, Crawfords, Mt. Washington and 

 Conway), New York (Ithaca), Washington (Mt. Rainier), 

 British Columbia (Kaslo and Metlakatla) and Alaska (Nush- 

 agak and Sitka). It belongs mainly to the Boreal region, 

 but it runs over into the Transition Zone somewhat. Exten- 

 sive collecting will probably extend the range of habitat 

 given above considerably. In the West, it will probably be 

 found to range south into Colorado and Northern California. 

 In the East, it seems to be about as common as ashtoni. It 

 is surely present in Vermont. 



I am strongly of the opinion that fernaldce is the female of 

 tricolor. I have evidence to support this opinion as follows : 



1. "Qoth. fernaldce and tricolor, as shown by their structure, are with- 

 out any very close allies in the New World, and they are the only spe- 

 cies which appear to stand thus isolated. 



2. They have about the same range of habitat, though tricolor has 

 been taken in Colorado v^\\\\q. fernaldce has not. 



3. The fernaldce females alone ever have a touch of ferruginous pile 

 on the sides of the fifth dorsal abdominal segment to correspond with 

 the pile of the same color on the apex of the abdomen of the tricolor 

 males. 



4. The closest allies of fernaldce known to me are the females of 

 the European species, P. quadricolor and P. globosus, and the males of 

 those same two species are the closest allies of tricolor, as shown by the 

 descriptions, figures of genitalia, etc. of Schmiedeknecht (Apid. 

 Europ., 1882, pp. 406 and 409 and Tab. 14, figs. 4, 5 and 6). There 

 are specimens of both sexes of both quadricolor and globostis, labeled 

 by Schmiedeknecht, in the collection of the Massachusetts Agricultural 

 College and I have compared them carefully with specimens of fer 

 naldce and tricolor. I find, by this comparison, that both quadri- 

 color and globosus, especially the former, are closely allied to both 

 fernaldce and tricolor. While I was unable to dissect any of these 

 specimens of either quadricolor or globosus, on account of their value 

 as authentic specimens, the genitalia of one of the quadricolor males 

 were protruding from the end of the abdomen and I was thus enabled 

 to compare even the genitalia of this species with those of tricolor. 

 They are very similar, as Schmiedeknecht's figure shows. While 

 quadricolor and globosus both differ noticeably, in some respects, in 

 coloration from fernaldce and tricolor, they all agree in a tendency to 

 have some ferruginous pile on the apical portion of the abdomen. 



TRANS. AM. ENT. SOC. , XXXVIII. 



