A survey of the Phytogeography of the Arctic American Archipelago 153 
chaiige of species between western America and eastern Asia wliich has actually 
taken place. 
Farther eastward we find the whole Arctic Archipelago changed into one 
contigiious landmass, no more separated from the Continent and only intersected by 
a few bays and fjords, while less thau thousand feet of elevation are enough to 
connect it with Greenland över the present Smith Sound and most parts of its 
northern continaation. Every channel now existing between the Polar Sea and 
warmer oceans on the american side of the Polar Regions thus has been closed up 
in late tertiary and early pleistocene time; it is also possible, and even rather 
probable, that similar conditions ha ve prevailed farther east, viz., that most part of 
the subraarine ridge now crossing the northern Atlantic from Greenland over Iceland 
and the Fseroes to Scotland has at the same time been elevated, so as to form a 
nearly continuous bridge. The existence of such a landlocked Polar Sea, without 
any warm current reaching it and without possibility of excharging its icemasses 
and cold water into warmer oceans, must have been of a very great importance 
for the climate of its encircling lands, and even if some channels existed on the 
east atlautic side, the ice and water of the large expanse between America and 
Asia must have been quite stationary. If these circumstances, in conuection with 
the considerable uplift of the land, especially in North America, have been enough 
to call the glaciation into existence, as Upham, Bell, and Leconte thiuk, or if 
other factors have also contributed to produce it, is a question which I small not 
venture to express an opinion about, the more so, as it is the effect and not the 
cause of the iceage that is of interest here. 
Gradually, however, as the elevation of the northern lands proceeded, the 
congenial climate of the tertiary era, when woods of evergreen trees grew beyond 
the 82" parallel, was replaced by a more and more severe one, and the old floras 
were compelled to retreat step by step southward along the different ways that 
stood open for them. The ancestors of the present circumpolar flora, wherever 
they came from, certainly were the last to fly before the encroaching landice to 
still habitable areas in the south. The question about regions of retreat in the 
north, already once referred to, 1 shall come back to låter, now it must be seen 
how the inlandice gradually took possession of wide areas in North America. During 
the maximum of glaciation the whole Dominion of Canada, from the Atlantic up 
to the valley of the Mackenzie, was covered by one continuous icesheet. The 
Southern horder, which is only of secondary interest here, lay far down in the 
United States, the northern seems to have coincided rather closely with the present 
Continental shoreline, with the exception of Boothia Fehx and the Melville Peninsula. 
About the western horder the american glacialists hold somewhat different views, 
some, as Chamberlin (in Geikie, Great Iceage) supposing it to have left a narrow 
strip of land free of ice, hmited on the other side by the separate Cordilleran 
ieesheeet, others again, as Wright, Iceage in N. Am., making them confluent far 
up in the Mackenzie valley. But even the wide eastern icesheet was not continuous 
