276 T. Holm — Studies in the Oyperacem. 



illustrations, we suppose that they may be regarded as members 

 of this section : the monostachyous C. hakkodensis Franch., 

 O. rhizopoda Maxim., C. heteroclita Franch. and C gralla- 

 tovia Maxim, all from Japan; O. Yallis pulchrce Phil, from 

 South America, and C. acicidaris Boott from New Zealand ; 

 finally the pliostachyous C. haimatostoma Nees, C. psychro- 

 phila Nees and C. macrogyna (Turcz.) from northern India. 

 Boott himself considered V. macrogyna as being an ally of 

 C. sempervirens, but states that in the former the uppermost 

 spikes are sometimes androgynous. In regard to C. hirtella 

 Drej., C. hcematostoma and C. psychrophila Boott compares, 

 also, these with the Ferruginece Tuckm., as representing on 

 the mountains of northern India those analogous species found 

 on the Alps of Europe. C. grallatoria is, according to Maxi- 

 mowicz, a close ally of C. lejocarpa. 



The Stenocarpce constitute a small, very limited section, and 

 the types are either alpine, arctic or confined to the coasts of 

 the Pacific from Bering Sea to Japan. (J. misandra exhibits 

 the widest range in geographical distribution ; besides being 

 circumpolar, this species occurs on the mountains of both 

 northern and middle Europe and on the Kooky Mountains as 

 far south as middle Colorado. The central forms, excepting 

 C. dblata from Vancouver Island, are confined to the European 

 Alps or, if we include those cited by Boott, to the mountains 

 of India. Finally, as mentioned above, there is a species in 

 South America and one in ]Sfew Zealand, both with a single, 

 androgynous spike, which Mr. Kiikenthal considers as belong- 

 ing to this same section, with him the Frigidce of Fries. 



There is, thus, considered from a geographical viewpoint, 

 a wide gap between the "hebetate*}" and "centrales;" yet, 

 as we have stated above, there is one of the latter which 

 inhabits Vancouver Island ; moreover by including the Japa- 

 nese representatives of " hebetatce" these are not so very remote 

 from their higher developed allies on the mountains of northern 

 India. The Pyrenees and middle European Alps possess, on 

 the other hand, types which are very isolated and are only con- 

 nected with their northwest American allies through C. 

 misandra. The most gradual development of the section may, 

 thus, be traced in Asia and in the northwestern corner of 

 America, where all the types occur : hebetates, centrales and 

 desciscentes. But we have no species of this section in Europe, 

 which as C. lejocarpa may point towards some fundamental types 

 of the recent time, and the formce centrales of the European 

 mountains, C. sempervirens, etc., may be considered as repre- 

 senting the remnants of an old center of which the earlier 

 types have disappeared long ago, but leaving a group of species, 

 the highest developed in the section. 



