Rydberg: Notes on Rosaceae 87 



species, while Wolf regards both as varieties of P. pennsylvanica. 

 Wolf regards P. pseudosericea and P. glabrella as good species, 

 while Nelson regards the former as a synonym of P. bipinnatifida 

 and ignores P. glabrella altogether. When such a diversity of 

 opinion exists regarding the limitation of the species, and I in 

 my mind was just as uncertain which should be regarded as species 

 and which as varieties or forms, I treated all that had some char- 

 acters tolerably constant as distinct species. 



Both Wolf and Simmons * regard P. Sommerfeltii as a variety 

 of P. pulchella. Dr. Simmons states: "but there are in the 

 London collections, no original specimens from Ross's first voyage 

 and the specimens under the name of P. pulchella from Melville 

 Island, that I have seen, are really P. Vahliana, to which, however, 

 the description does not apply." In the old Torrey herbarium there 

 are some specimens of the Melville collections, gathered by Parry. 

 Two of these are P. Wahliana but one is P. pulchella as I under- 

 stand it, a plant with leaves densely silky on both sides, slightly 

 if at all tomentose beneath, and with narrow, linear, acute segments. 

 P. Sommerfeltii, as I understand it, has leaves green and almost 

 glabrous above, somewhat silky and densely white-tomentose 

 beneath, with oblong, rather than linear, and obtuse segments. 

 This is the common plant of Spitzbergen, but also found in arctic 

 America. 



The original Potentilla Sommerfeltii was collected by Keilhave 

 and found in Sommerfelt's herbarium; it may be the same as 

 P. Keilhavii Sommerf.,* which has always been regarded as a 

 synonym of P. pulchella. As it has been impossible for me to see 

 the original description of P. Keilhavii, I left it as a questionable 

 synonym under that species. 



If P. Sommerfeltii is regarded as a species, it would not be 

 out of place to regard P. pulchella elatior as such. It has been 

 customary to label all tall well-developed plants P. pulchella v. 

 elatior and all depauperate ones P. pulchella. Without regard to 

 size, there are evidently two distinct races (they may be called 

 species, varieties, or forms) known as P. pulchella, beside P. Som- 

 merfeltii discussed above. One is the typical P. pulchella just 



*Vasc. PI. Fl. Ellesmereland 479. 1906. 

 fMag. Naturv. II. 1: 244. 1832. 



