554 Rydberg : Rocky Mountain flora 



ones are not so elongated as in that species. From C. undulatus 

 it differs in the narrower segments of the leaves, the glabrate upper 

 surface of the same, the somewhat purple-tipped inner bracts, 

 and the inconspicuous dorsal ridge. 



Utah: Fish Lake, near Twin Creeks, Aug. 8, 1905, Rydberg 

 & Carlton 7499 and 7487; Beaver City, 1877, Palmer 273. 



Colorado: Grand Junction, June 15, 1900, Mrs. Stokes. 



Neither of the supposed parents was collected at the same 

 date and locality, but both are found in Utah and Colorado. C. 

 pulchelliis was collected by Carlton and myself in the neighborhood 

 of Marysvale, 7ios. 7016 and 7179. 



A good deal can be said about the treatment of this genus in 

 the New Manual of Botany of the Central Rocky Mountains. 

 Many of the species, reduced to synonymy, have little or no re- 

 lationship to those of which they were made synonyms. 



Under Cardiius americamis we find the following synonyms: 

 C. Cejitanreae Rydb., C. erosus Rydb., and C. griseus Rydb. 

 The first is a pure synonym. \Mien the name was proposed I was 

 following the Madison amendments of the "Rochester Code," 

 and according to those amendments an older varietal name in- 

 validated the name C. americanus. C. erosus I now think is a 

 hybrid of C. americanus and C. coloradensis . C. griseus, on the 

 contrary, is a good species, easily distinct from C. americanus, 

 and its bracts have no dilated erose tips. In my Flora of Colorado 

 I included in it at least two hybrids of C. americamis with this 

 species and related ones. This may have given Professor Nelson 

 a wrong idea of C. griseus. 



Under Carduus Hookerianus we find as a synonym C. Oster- 

 houtii. C. Hookerianus is, so far as I know, not found within the 

 United States. It is from the Saskatchewan region. I, as well 

 as others, have referred specimens from Colorado to it, but all 

 these belong to a form of C. scopulorum. This may have been the 

 reason why Nelson has made the latter a variety of C. Hookeri- 

 anus under the name C. Hookerianus eriocephalus. C. Osterhoutii 

 is not closely related to C. Hookerianus, but more so to C. scopu- 

 lorum. I think it is a hybrid of that species and C. griseus. 



Under Carduus Hookerianus eriocephalus we find the follow- 



