Rydberg : Notes on Rosaceae 501 



pharga, and P. uniflora are included. Dr. Simmons states further: 

 "In the Index Kewensis, it is referred to P. multifida L., where its 

 right place seems rather to be." This statement was made after 

 Dr. Simmons had seen the original specimens of Bunge. It is not 

 unlikely at all that its relationship is with P. pinnatifida, but the 

 plate illustrating it does not resemble so much P. pinnatifida as 

 P. Hookeriana. The latter is also intermediate between the 

 Niveae and the Multifidae groups, both in habit, the congested 

 inflorescence, and the fusiform style. I can not agree with Dr. 

 Wolf in regarding P. altaica as merely a form of P. nivea pinnati- 

 fida in the sense he uses the name. I refer the following specimens 

 to P. ni pharga: 



Mackenzie: Fort Good Hope, 1861-2, I. S. Onion (type). 



Alberta: Rocky Mountains, Drummond 368. 



Utah: Sierra La Sal, 1899, C. A. Purpus A, C, L, and H. 



Greenland: Vajat-shore, Disco, Morten Pedersen 113; Un- 

 artuarsuk, Disco, Morten Pedersen; Onjigsak, Disco, Morten Peder- 

 sen 233. 



Potentilla Pedersenii Rydb. was based upon P. subquinata 

 Pedersenii Rydb. In describing the latter I had only one collec- 

 tion in flower, viz., Morten Pedersen 470, and the description was 

 drawn wholly from that. Laying too much stress upon the pe- 

 culiar rootstock of this species and almost overlooking more 

 essential characters, I referred carelessly to it several sterile speci- 

 mens with a similar rootstock. These do not belong to it and 

 Dr. Wolf is fully correct in his criticism of me for basing the 

 variety on a mixture and for including in it specimens which he 

 refers to P. nivea and var. subquinata. The type specimen he 

 regards as a small-flowered P. Wahliana. With this I can not 

 agree. It is true that it has something of the habit of P. Wahliana, 

 the long hairs, although not yellowish, of that species, and the 

 pubescent upper surface of the leaves; but it does not have the 

 large, broad, and overlapping petals of P. Wahliana, or the oval 

 or elliptic obtusish bractlets characteristic of that species and P. 

 villosa. It has the flowers of Potentilla nivea. It could be a 

 hybrid of P. nivea and P. Wahliana. 



Dr. W 7 olf regards both P. uniflora and P. Hookeriana as vari- 



