582 Plain, Prairie, and Forest. [ October, 
tered. That the winds on the Pacific coast are, in places, inimi- 
cal to the growth of forests seems very clear, but this does not 
seem to be the case in the Mississippi Valley, for we often find 
the most abundant growth of trees on the very spots which are 
most exposed to the force of the blast. It is on the rising ground, 
the knobs, knolls, and mounds which are scattered over the sur- 
face of the prairie, that those isolated clumps of timber, called 
groves, are most likely to be found. If the force of the wind 
were essentially inimical to the growth of trees, we should find 
them thriving, if anywhere, in the sheltered nooks, and to the 
leeward of the northwesters, that being the quarter from which 
the heaviest blasts come. This is not the case, nor is there any- 
thing, so far as we have observed, which would lead to the con- 
clusion that the force or direction of the wind has any sensible 
effect on the growth or distribution of the timber in the prairie 
region. 
The only climatological cause for the existence of the prairies 
which is worthy of serious consideration in this connection is the 
distribution of moisture, and by far the larger number of those 
who have written on this question have unhesitatingly asserted 
that in something connected with the rain-fall was to be found 
the really efficient agent by which the distribution of woodland 
and prairie has been effected. Exactly what this something 18 
seems, however, not easily to be made out. It must, as it would 
appear, be one of four things: either the annual rain-fall is on 
the whole deficient, or it is not favorably distributed through the 
seasons, or the climate is subject to cycles of drought, or there 
must be an excess of moisture. In one or other of these cat- 
egories the influence of the rain-fall must be found, or if not m 
one single condition of these here enumerated, then in some 
combination of them. : 
Let us first examine whether the average annual rain-fall 1s 
really deficient, so that the absence of trees over a considerable 
portion of the Mississippi Valley may be referred to this as 4 
cause. And to settle this question we have no better method 
These charts are, of course, for many districts only rough ap- 
proximations ; for it is, over a large part of the country, 
within a few years that statistics have begun to be collecte 
The data seem, however, to be sufficient for our purpose. If se 
we examine these charts, we find that for the typical prairie regions 
