380 RyDBERG: NOTES ON ROSACEAE 
of the styles. It is practically the same character as I used in 
distinguishing the genera, only that I placed more importance 
on the position and Dr. Wolf on the form of the style. I also 
took the stamens into consideration, which Dr. Wolf only in- 
cidentally mentions. Dr. Wolf’s RHOPALOSTYLAE correspond 
to my genus Dasiphora. His NEMATOSTYLAE correspond to 
Sibbaldiopsis and Comarum, together with several groups not 
American and therefore not treated by me. These two subsections 
constitute his section POTENTILLAE TRICHOCARPAE. It is evident 
that Comarum (Potentilla palustris of his monograph) should not 
be counted in this section, as the carpels are perfectly naked. 
He associates P. palustris, a herbaceous plant with creeping root- 
stock and glabrous achenes, with P. Salesowiana, a shrub with 
hairy achenes. While the latter is in its right position in the sys- 
tem, the former is not. I shall discuss this further under the 
genus Comarum. To Dr. Wolf's POTENTILLAE GYMNOCARPAE 
belong the rest of the subsections. (CLOSTEROSTYLAE correspond 
to the genus Drymocallis and LEPTosTyLAE to Argentina. The 
subsections CONOSTYLAE and GOMPHOSTYLAE show so many inter- 
gradations, a fact admitted by Dr. Wolf, that there is no ground, 
in my opinion, for keeping them apart. They constitute what 
I have called Potentilla. In the main points Dr. Wolf and myself 
agree, the only difference being that what he calls subsections, 
I call genera. J can not help, however, but accuse Dr. Wolf of 
inconsistency, for Sibbaldia, Fragaria, Horkelia, Comarella, and 
Stellariopsis, all admitted by Dr. Wolf, are none of them better 
genera than these subsections. 
Now let us take up the different groups of Potentilla in the 
order they are in the North American Flora. 
TORMENTILLAE 
This contains six species, of which two, Potentilla reptans and 
P. procumbens, are introduced. Dr. Wolf admits only one North 
American species, regarding P. pumila and P. simplex as varieties 
of P. canadensis. Potentilla caroliniana was evidently unknown 
to him. Regarding P. pumila, he states that I regarded it in 
1898 as a distinct species but withdrew the rank in 1899 (re- 
fering to the Bulletin of the Torrey Botanical Club for that year, 
