A survey of the Phytogeography of the Arctic American Arcliipelago 45 
Geographic area. Greenland, northern America, northern and central Asia, 
northernmost parts of Europé. 
In the KH there is, besides the specimens mentioned above, one from Parrt's 
second voyage without locality, and in the NHM one from bis third; Rae, 1, c, p. 
215, records it from the interiör of the Melville Peninsula. It seems to evade the 
Silurian strata. 
15. Oalamagrostis purpurascens, R. Br. 
Distribution. Victoria Land: Minto Inlet, Anderson (KH). 
G-eographic area. Greenland (?), Arctic America, Rocky Mountains (?). 
After considerable hesitation I have resolved upon using the above name for 
the plant in question, although I am very willing to admit that it may not be 
possible to keep it up as a separate species. I have not, however, had any oppor- 
tunity of forming an opinion of my own about the species to which it should be 
united. This being the case I think better to use a name that certainly applies to 
it, than to follow some or another of the authors who have reduced it to a variety 
of some other species. Macoun, Cat., p. 206, is inclined to look upon it as belon- 
ging to C. silvatica, DC, Gelert in Ostenfeld, Flora Aret., p. 102, has thrown 
both together under C. arundinacea, (L.) Roth. Gelert has not, however, seen any 
american specimens, and I do not feel certain that the Greenland plant is the same 
as the american. If Gelert was right, it should be a species of very wide distri- 
bution, hut not strictly circumpolar, as it is not found in the eastern parts of 
America. If, on the other hand, only the specimens from arctic and western 
America are looked upon as belonging to one species, there will be nothing peculiar 
in its distribution. 
16. Oalamagrostis lapponica, (Wahlenb.) Hartm. 
Distribution. Baftm Land: Cumberland Gulf, Tatlor (KH). 
Geographic area. Greenland, northern America, Siberia, northern Europé. 
Tatlor, Pl. Baff., p. 332, has only one species of Oalamagrostis in his list 
which is called C. canadensis. Now I do not, indeed, doubt that he means the same 
plant as the specimen of the KH represents, but in the label it was styled »Agrostis 
rupestris, Willd.» Probably the labels have been in some manner mislaid, which 
might, perhaps, account also for the wrong name mentioned above under Agrostis 
horealis. If this conjecture is right, the special locality should be Cape Searle, as 
given in Taylor'8 paper. The name C. canadensis is often used for forms belonging 
to either C. lapponica or C. neglecta; I have seen a specimen collected by Rae in 
the arctic coast of the Continent, which had been successively named ^lapponica», 
>^stricta^, and »canadensis^. Certainly C. lapponica and C. neglecta, rather well 
defined as they are in Scaudiuavia, are very difficult to separate in other regions. 
The question if the name C. confinis, (Willd.) Nutt., is to be taken up instead of 
Lands Universitets Årsskrift. N. P. Afd. 2. Bd 9. 6 
