A. Gray—Forest Geography and Archeology. 87 
tic and a Pacific. They may take hors names, for they are 
vise upon the oceans which they respectively border. 
so we have an intermediate isolated region or isolated lines 
of seit flanked on both sides by bare and arid plains,—plains 
which on the eastern side may partly be called prairies,—on the 
western, deserts. 
and less arid region, is larger, becaibii more diffused. Trend- 
ing westward, on and beyond the northern boundaiy of the 
United States, it approaches, and here and there unites with, 
the Pacific forest. Eastward, in Northern British territory, it 
makes a narrow junction with northwestward prolongations of 
the broad Atlantic forest. 
So much for these forests as a whole, their position, their 
limits. Before we glance at their distinguishing features an 
component trees, I should here answer the question, why the 
occupy the positions they do ;—why so curtailed and separated 
at the south, so much more diffused at the north, but still so 
strongly divided into eastern and western. Yet ‘I must not 
consume time with the rudiments of physical geography and 
foliage, its vast amount of surface which it cannot diminish or 
ta except by soli that weheng it lives, it is completely 
elplessly exposed to every atmospheric change; or at 
sem its resources for sdnpiatid are very limited ; and it can- 
not flee for shelter. But trees are social, and their gregarious 
habits give a certain mutual support. A tree i 
doomed, where a forest, once established, is comparatively 
secur 
so 
rees vary as widely as do other plants in their constitu- 
tion; but none can withstand a certain amount of cold and 
other exposure, nor make head against a certain shortness of 
summer. Our high northern regions are therefore treeless; 
and so are the summits of high mountains in lower latitudes. 
As we ascend them em we walk at first ie ruces and fir- trees. 
