104 
Journal of Agricultural Research 
Vol. XXIV, No. a 
By analogy we may say that the same relation exists between grassland 
or sage'brush and the lowest type of forest, commonly called “wood¬ 
land’' by foresters, since it is obvious that there can be no important 
change in climatic conditions in the small space between the centers of 
development, and that the development of the forest is made possible 
by the slight soil changes resulting from elevation, surface erosion, and 
leaching, all of which maintain a younger soil. 
The establishment of the forest experiment stations in the western 
United States, beginning with that at FlagstaflF, Ariz., in 1908, gave 
unparalled opportunity for the collection of forest and climatological 
data over a period of years. As a result, studies similar to the present 
one have been initiated in Arizona, California, and Idaho, and the study 
of forest and herbaceous vegetation has been carried on in connection 
with a number of experiments at the Utah Station, where gazing prob¬ 
lems occupy the attention first. The strictly forest studies, however, 
are for the most part not yet ready for publication. 
Pearson {14) at the Flagstaff Station early investigated the effect of 
yellow pine forests upon local climatic conditions, by securing data in 
the forest, the edge of the forest, and “parks” (grassy, meadowlike 
openings) of considerable extent. A^ile the climatic conditions recorded 
by Pearson are interestingly compared with our own, this study can not 
be said to throw very mudi light on the conditions governing different 
forest t5^es. 
However, Pearson {ij) has recently made available the results of 
observations at a series of stations in the San Francisco Mountains, in 
a very comprehensive way, and we shall have reason frequently to com¬ 
pare his conditions with our own. 
We have, similarly, had access to an unpublished report by Larsen ^ 
on the conditions of Montana and Idaho, which has been extremely 
helpful in giving comparable data. 
The problem of the prairies in the Middle West, and their physical re¬ 
lation to the occasional forested areas, has received considerable atten¬ 
tion, and this problem is not too remote from our own entirely to lack 
interest. On ^is subject may be considered the work of Shimek (22), 
who concludes, regarding Iowa conditions: 
I. Exposure to evaporation as determined by temperature, wind, and topography 
is the primary cause of the treelessness of the prairies. 
and 
3. Rainfall and drainage, while of importance because determining the available 
supply of water in both soil and air, are not a general, determining cause, both fre¬ 
quently being equal on contiguous forested and prairie areas. 
Shimek also dismisses fires as a cause of the absence of forests.^ It is 
believed that the later conclusions of Weaver and Thiel (2d, 28 ), with ref¬ 
erence to Minnesota, are essentially in agreement with this. The point 
which seems to have been overlooked here, and in all similar discussions, 
is that forests occur usually on the slopes of ravines or on hillsides, where 
the old soil is being rejuvenated by a secondary erosion and where, 
even with less moisture than in the heaviest soils, lie availability may be 
greater. Considered from this angle, the occurrence of forests in the 
prairie region is exactly parallel to their occurrence on the first mountain 
elevations at the edge of the plains. 
* Lars] 9 n, J. a. ci,imatic study of forest types, district I. U. S. Dept. Agr. Forest Serv., unpub¬ 
lished report, 1918. 
