Rvdberg : Notes on Rosaceae 49*3 



Potentilla fastigiata Nutt. was transferred from the GRACILES 

 group. Although it connects the CONCINNAE with this group and 

 with the CANDIDAE, and has a more erect stem than the other 

 species, I think, after all, that it should be placed here and that it 

 has its nearest relative in P. concinnaeformis Rydb. Dr. Wolf 

 retains it in the Graciles group and associates it with P. tomen- 

 tosa, P. oaxacana, and P. Nuttallii. Its nearest relative outside 

 the CONCINNAE is P. Candida Rydb., and a few of the specimens 

 cited by me under the latter in my monograph belong to P. fasti- 

 giata instead. It is evidently unknown to Professor Nelson, for 

 under P. gracilis, he has the following remark: "(This may be a 

 composite species; as here used it includes the following, which 

 are not readily discriminated: P. fastigiata Nutt.; P. pulcherrima 

 Lehm.; P. Blaschkeana Turcz )" The leaves of P. fasti- 

 giata are silky-villous with rather long hairs on both sides, only 

 very slightly tomentose beneath. 



A species related to this and P. concinnaeformis is described in 

 the North American Flora under the name of Potentilla Hassei. 

 It differs from P. fastigiata in the broader, broadly obovate leaf- 

 lets and the oblong instead of linear-lanceolate bractlets, and from 

 P. concinnaeformis in the dense many-flowered inflorescence, the 

 densely pubescent stem, and the oblong bractlets. Besides the 

 type given in the North American Flora, I have seen the following 

 specimens, referable to it : 



California: Head of Stanislaus River, 1903, Hall & Chandler 

 4778. 



There has been no change made in Potentilla concinnaeformis , 

 P. oblanceolata, and P. bier xnata since my monograph, except that 

 the range of the last one has been extended to Wyoming, where 

 it has been collected by Professor Nelson. Dr. Wolf admits all 

 three as species, although he had not seen specimens of any of 

 them. Concerning the reduction of P. obovatifolia Rydb. to a 

 variety of the first one, see my remarks on page 493. 



A further study of P. concinna and its variety divisa brought to 

 light facts that seemed to me sufficient to warrant the raising of 

 the latter to specific rank. 



In the North American Flora, I transfer Potentilla quinquefolia 

 Rydb. to this group. In my monograph I had placed it with P. 



