124 DISTRIBUTION AND MOVEMENTS OF DESERT PLANTS. 
Comparative studies, especially those carried out in the upper Gila 
Valley in connection with the Bureau of Soils, U. S. Department of 
Agriculture, show the existence there of the same topographic features, 
though on a larger scale, and the same general grouping of plant asso- 
ciations as in the Santa Cruz Valley. There is, moreover, a striking 
correspondence between the location of the various plant associations 
and that of the independently mapped areas marking soil distribution 
and character. An extended study of these areas, supplemented by a 
similar study in the Salt River Valley, has led to the conclusion that 
water-content, percentage of alkali, and drainage, that is to say aeration, 
are the three soil factors that have most obviously influenced the local 
distribution of desert plants and governed their association in definite 
habitats. 
The soil investigations conducted on the Laboratory domain by Dr. 
Livingston have thus far been directed chiefly to a determination of the 
water-content, but in this respect they confirm to a noteworthy degree 
the conclusions based on comparative observation. 
Four distinct types of soils are distinguished, namely, those of the hill, 
of the Larrea slope, of the wash, and of the flood-plain, corresponding 
closely with the grouping of plant associations already described. It is 
shown that the soil of the hill absorbs relatively most water and holds 
it with great tenacity. It dries only by the action of plants and by 
direct evaporation at its upper surface. The soil of the flood-plain holds 
as much water at the start, but dries downward as well as upward; that 
is, the dry soil below absorbs water from the surface layers, so that it is 
.not to be expected that this soil will remain wet, at a foot depth say, 
as long as the soil of the hill. The soil of the Larrea slope is much like 
that of the hill, but has so much angular gravel mixed with it that its 
absorbing and holding capacity are alike low. It has excellent drainage 
conditions, and even if once thoroughly saturated does not remain so 
for a long period. ‘The wash, as indicated by determinations of water- 
content, is even worse off than the Larrea slope, but its perennial plants 
apparently root deeply enough to get at the underground flow. 
Within these several areas modifications of conditions are pointed out 
which correspond closely with observed facts of habit and distribution. 
Thus the underground flow of the wash and the flood-plain is available 
to such deeply rooting species as the acacias and the mesquite, and 
these areas accordingly are their special habitat. The fact that on these 
areas the soil near the surface dries out relatively rapidly corresponds 
with their occupation by two biological groups of plants very different 
from the deeply rooted perennials and different from each other, namely, 
low perennials of distinctly xerophytic character, and various annuals, 
which are commonly able to complete their short cycle of development 
before conditions of extreme drought set in. Neither of these plant 
