46 
that which has been cominonly used, as Gelert, in Ostenfeld, Flora Aret., p. 103, 
wiW have it, I shall not enter upon now. Macoun, Cat., p. ^05, keeps the olH na me. 
17. Aira csBspitosa, L. 
Deschampsia ccespifosa, Armstrong, Narrat., Hooker, Coll.; D. brevifolia, Brown, Chlor. Melv., 
HooKER, App. Parry II & III, Fl. Bor. Amer., Macoun, Cat. 
Distribution. Victoria Land: Minto Inlet, Anderson (KH); [Melville Peninsula 
(?), Parry (KH), J. C. Ross (LS), Edwards (ex Hooker, App. Parry II)]; Baffin 
Land: Port Bowen, Parry (LS, KH); iMelville Island, Sabine, Ross, Fisher (NHM), 
herb. Trevelvan (CH); North Devon: Cape Vera, Simmons (Stray Contr.); North Kent, 
SiMMONs (1. c); Ellesmerelaud: widely distributed, Hart, Greely, Simmons. 
G eographic area. Greenland, arctic and temperate America, Asia, and 
Europé, southern South America, Tasmania, New Zealand. 
Besides the specimens mentioned above there is in the NHM one more from 
Parry's third voyage without locality, hut probably from Port Bowen. All speci- 
mens I have seen from localities to the south belong to the var. arctica, (Trin.) 
Simm. So do also the Ellesraereland specimens from the southern coast and from 
Grinnelland, but at Fram Harbour on the eastern shore I collected another form, 
so widely diiferent that, as my material was rather imperfect, 1 did not recognise it 
as belonging to the same species, but after cousiderable hesitation identified it as 
a form of Aira fiexuosa. Afterwards Dr. Ostenfeld who bad a better developed 
and more numerous set of specimens from Northeastern Greenland at bis dispo- 
sal found it to belong to another variety of A. ccespitosa, viz., var. pumila, Ledeb. 
Having inspected the material in question, I think best to adopt Ostfnfeld'8 view 
(Danmarkexp., p. 12), even if it appears rather curious that the species 
should be represented by two so widely different forms, one of which, var. arc- 
tica, is connected with the type by a series of intermediate forms, whereas the other 
stånds quite isolated and might perhaps be looked upon as an arctic species, diffe- 
rentiated from the type of A. ccespitosa within its present range. That it does not 
belong to A. fiexuosa, I cannot but readily admit after having seen the material 
from East Greenland, and thus the latter species must be excluded from the flora 
of the Arctic Islands. 
18. Aira atropurpurea, Wahlenb. 
Distribution. North Somerset, Parry (NHM). 
Geographic area. Greenland, northern America, northern Europé. 
HooKER does not mention this species in App. Parby III, and in Fl. Bor., 
Amer. be has only one locahty for it, viz., Sitka. As, however, the plant in question 
has been found låter in a good many places, among others along the Arctic Coast 
(Pitllen), i think there may be nothing wrong about the specimen from North 
Somerset, probably collected at Fury Beach. 
