1898-1902. No. 2.] VASCULAR PLANTS OF ELLESMERELAND. 45 



visible, the surface is more or less shiny, bright and smooth. That a 

 denser covering of black glandular hairs (especially on the scape and 

 sepals) should characterize Z). octopetala in distinction to D. integri- 

 folia, as Hooker (1. c.) and Rosenvinge (1. c.) intimate, hardly holds true. 

 Among the Danish Greenland specimens in the Copenhagen herbarium, 

 there are, besides those from Kingigtok at the Vaigat, which the latter 

 author mentions, many more that have densely glandular scapes and 

 sepals. Also some of the specimens which I collected at Egedesminde 

 and Godhavn have a rather dense glandular covering, and still more 

 is this the case with specimens from Foulke Fjord, and with most EUes- 

 mereland specimens. When Pursh (1. c, p. 350), says that the flowers 

 are only half as large as in D. octopetala, this must have been acci- 

 dentally the case with the specimens after which he has made his des- 

 cription. The flowers of D. integrifolia are not really so small, even 

 if they never become so large as the biggest ones of D. octopetala, and 

 they are mostly a little smaller than the medium size of the flowers 

 of the latter species. Besides they are not so purely white, but rather 

 often have a faint touch of yellow. 



It appears from the preceding, that it is a rather difficult task to 

 make a distinction between certain forms of both species in districts 

 where both grow. But such common areas are restricted to the border- 

 regions of the two species, and the principal foundation for the distin- 

 guishing of D. integrifolia as a separate species lies, in my opinion, 

 apart from its having proved constant in cultivation (Hooker, 1. c), in 

 its geographical distribution. To a plantform, which has a great con- 

 tinual area of distribution, where it totally excludes another allied one, 

 which on the other hand, is the ruling one in other equally wide 

 tracts with similar conditions of life, I must, for my part, concede 

 the right of being looked upon as a good species, even if the mor- 

 phological characters are not so very distinct. 



Nathorst, 1. c, p. 24, also has a D. octopetala L intermedia, which 

 he looks upon as a connecting form between the two species, or perhaps 

 a hybrid. Specimens of it seen in the State Museum at Stockholm, 

 have proved to belong to a broad-leaved and dentate form of B. inte- 

 grifolia. I have seen plants that could be reckoned as this form in 

 several localities and they are also represented in my collections (f. i. n. 

 4218), as well as such as could easily be taken for D. octopetala 

 (f. i. n. 2345, found in deep shade under a protruding rock). However I 

 also found another variety which seems to deserve a name: 



