A survey of the Phytogeography of the Arctic American Archipelago 163 
and therefore we find the latest traces of habitation as well as the present settle- 
ments situated wliere local causes have contributed to produce open water for a 
comparatively long part of the year. But the same causes will contribute to a 
local glaciation and thus be adverse to plantlife. 
On the whole we ma}' most probably look upou the flora as stabile, and even 
if a few speeies may be retreatiiig, it is on the other hand not improbable that 
new immigrants are still arriving. The rather large nuraber of speeies, stopping 
at the coast of the Continent, may some day be found to have crossed the narrow 
straits o ver to the southern islands, and speeies found only in single locahties on 
these islands may perhaps be viewed as such late arrivals. 
The stages of plantimmigration in the Arctic Archipelago and their 
traces in the present flora. As pointed out in the preceeding the most impor- 
tant area of rofuge during the Glacial period was to be found west of the Mackenzie, 
and most probably also the subsidence of the postglacial time first set in to the west. 
Certainly the melting down of the icesheets began in the west, and proceeded eastward 
so as lo make the Labrador icesheet the last existing of the wide inlandices of America 
(Tyrell, NW Canada, p. 390). Thus the first chances for reimmigration of plants 
to the Archipelago must have beeu such as to allow a dispersion eastward along 
tlie coast of the Continent and over to Banks and Victoria Land Under such 
circumstances speeies which have wandered along that line must have beeu the 
first to reaeh the Arctic Islands, and as they have had the longest time for 
spreading, they ought to have reached farther over the Archipelago and over larger 
parts of Greenland than speeies wandering along ways which were opened first at 
a låter period. Actually thig western element takes a very prominent part in the 
composition of the flora of the Archipelago, and even in Greenland a rather consi- 
derable uumber of speeies shows a distribution, only to be interpreted by an early 
postglacial immigration over the western islands, to Ellesmereland, and from there 
to Greenland, whére the plants in question have reached more or less far to the 
south, some even to the southernmost extremity of the land. For the purpose of 
pointing out the distribution of this group of plants the following table (VI) is 
The first column of the table accounts for the distribution of the speeies on 
the Continent of America. The signs used here are to understand as follows : 
»W — » means that the plant goes south of the arctic parts in western America, for 
instance as alpine in the Rocky Mountains, but that it bas, as far as known, an 
eastern borderline somewhere in the Barren Grounds. When an »E» is added, it 
signifies that the speeies is distributed to the south also in eastern America, »NE» 
that it reaches the Hudson Bay region, or even Labrador, but is decidedly more 
northern on the east side of the Continent than on the western. »NW»,»NW— », 
and »NW~NE» purport a principally arctic range Of the Arctic Islands Elles- 
mereland alone has got a column, as the distribution there is most apt to Ihrow light 
