Stray Contributions to the Flora of Greenland I—V. 37 
and thick pedicels, and its tendency to become (see fig. 7) spirally bent. 
The flowers of the Greenland plants are not oblate subglobose, like the 
var. iodandra, but distinctly urceolate, light pinkish, distinctly paler 
than in Greenland forms of A. polifolia. Anthers — in herbarium spec- 
imens — are brownish, but not pale-brown as is said about the main 
form. 
I have not yet seen mature fruits, but only found a very few old 
ones, showing that the -plant fruits in Greenland, at least in favour- 
able seasons. 
Besides the differences given by FERNALD and quoted above, there 
seems to be yet another. In A. polifolia, as it is described from Europe 
and as specimens at hand from Europe and Alaska show, the stems are 
very richly provided with roots, arising as WARMING I. c. has shown, 
in the axils of the leaves. The Greenland specimens are root-bearing 
right up to the lowermost leaves. The Greenland specimens of A. glau- 
cophylla do have roots on the subterranean parts of the stems, but 
in markedly less degree, and in two herbarium sheets from America 
there are none at all. One sheet (Gray Herbarium Nr. 24289, from 
Sphagnous boggy swale bordering a lake in Nova Scotia) shows stems 
without a single root, measuring 21—25 cm. up to the lowermost leaf. 
In the other (Gray Herbarium Nr. 26926, from turfy and peaty knolls 
in limestone barrens in New Foundland, leg. M. L. FERNALD, B. LONG, 
and B. H. Dunsar) there are none either. The bare parts of the stems 
under the lowermost leaves are 3—6cm. long. These last specimens 
are without flowers, but probably belong to the variety iodandra, as 
the habit matches well with the description and as they came from the 
same region and kind of soil as the variety. 
The geographical Distribution of the Species. 
In America A. polifolia is generally considered an arctic plant, 
extending very locally south to the Adirondack Mountains, Lake Huron 
and southern Alaska. Mostly the floras give the distribution quite 
broadly: “from the Atlantic to the Pacific and to the Arctic Ocean”. 
From literature available here a few really arctic occurrences may be 
mentioned: the Atlantic coast of Labrador, Kotzebue Sound and St. 
Lawrence Island, to which may now be added West Greenland, in 
68°47’ N. But in Eurasia this species cannot correctly be labelled arctic. 
Generally speaking, in Europe and Asia it occupies the northern belt 
of Coniferous woods and most of the deciduous forest regions. South- 
wards, it stops when the steppe-belt begins, or it is only to be met with 
in isolated alpine or boggy situations, see, f. inst., BRAUN-BLANQUET 
in his treatment of the Ericales in Hear: Illustr. Flora von Mittel- 
