1878.] Distribution of Timber and the Origin of Prairies. 95 
combined with the increased roughness of the surface of the 
country, may also go far toward explaining the timber belt 
observed in the bluff region. This becomes more obvious 
when we remember that the east side of the Missouri bot- 
tom lands, in the particular counties through which the tim- 
ber belt passes, abound in sloughs and lakes, which are kept well 
filled by numerous springs and small streams, which fail to make 
_ their way across the bottom land to the river. 
4 These causes also explain the occurrence of timber on the 
= west face of the bluffs, where the moisture in the air counteracts 
the severe drainage of the abrupt slope and the intense drying 
action of the afternoon sun, which render other portions of the 
__ bluff-face almost destitute of even grass. 
(4.) It remains to explain, if possible, the distribution of tim- 
d ber and prairie in the alluvial valleys or bottom lands.’ Here we 
have usually a slightly different soil, layers of clay preventing the 
ready drainage of many parts. These conditions render much of 
the surface too wet at all times, while other places are too wet in 
spring and too dry in late summer. These areas are found mostly 
_ in the eastern portion of the Missouri river bottom, because the 
surface there is lower and receives the water from springs and’ 
freshets from the bluff-region. The occurrence of trees along 
__ the streams and on ridges along old channels, on the other hand, 
_ may be explained partially by the inequality of surface, making 
_ the drainage of surplus water possible, so that moisture around 
_ the roots is more constant than elsewhere on the bottoms. The 
_ prairie fires, moreover, have been more efficient over the dryer 
~ portions of the bottom lands than elsewhere, because of the 
_ greater growth of grass and the free sweep for winds north and 
south 
V: 
That the constancy of the humidity of air and soil is the most 
important factor in the formation of forests seems supported by 
further considerations taken from a wider field, which may be 
briefly indicated as follows: This theory explains the timber in 
areas of less rain-fall in northern Michigan and Wisconsin by 
their being traversed by moist winds, and not as subject to the 
hot rays of the sun as the southern portions of the same states. 
It explains the occurrence of timber upon mountains by a sub- 
stitution of altitude for latitude in the preceding cases. It 
