494 Rydberg : Notes on Rosace ae 



madrensis is overlooked by Dr. Wolf. P. sanguinea is described 

 as new in the North American Flora. 



Aureae 



As treated in the North American Flora, this group contains 

 nine North American species. If one follows Dr. Wolf in laying so 

 much stress on the form of the style, all the American species, except 

 Potentilla maculata and P. Langeana, should be excluded from the 

 Aureae group. In this group the styles should be somewhat 

 thickened upward, instead of downward. In some species the styles 

 are perfectly filiform, not thickened either way. The result is 

 that a distinct line can not be drawn between the Conostylae 

 and the Gomphostylae of Dr. Wolf's monograph. In the 

 Aureae, Dr. Wolf has such diverse species as Potentilla maculata, 

 P. elegans, P. Robbinsiana, and P. gelida. Of these, P. Robbin- 

 siana at least does not have the styles thickened upwards. The 

 Ranunculoides group of Dr. Wolf is still worse. It contains not 

 only the thick-leaved and thick-rhizomed Mexican and Central 

 American species, of which it was made up in my treatment, and 

 my Subviscosae and Brevifoliae groups, but also such diverse 

 species as Potentilla acuminata Hall (related to P. saxosa), P. 

 flabellifolia Hook., P. fragiformis Willd., P. Townsendii Rydb., 

 P. Palmeri Th. Wolf, and P. Sierrae- Blancae Rydb. Potentilla 

 emarginata Pursh is placed with the Aureae, while the closely 

 related P. fragiformis is placed in Ranunculoides. P. gelida 

 Mey. is put in the former, and P. flabelliformis Hook., which S. 

 Watson and other students of Potentilla have not been able to. 

 distinguish from it, is put in the latter group; P. perdissecta 

 Rydb. or, as Dr. Wolf calls it, P. diversifolia var. decurrens (Wats.) 

 Th. Wolf, is placed in the Multijugae, while P. Ranunculus 

 Lange, which can be separated from it only by the different root- 

 stock (see below), is placed in the Ranunculoides. P. Town- 

 sendii and P. Palmeri and their relationship I have discussed 

 before. Dr. Wolf's grouping in this case therefore is very artificial 

 and unsatisfactory. 



I think that the group as constituted by me is more natural, 

 although it could be subdivided into three subgroups. Potentilla 

 Sierrae- Blancae stands alone, is not so closely related to the rest, 



