REVIEW AND DISCUSSION. 123 
their seasons of the plants composing the two widely different biological 
groups of this association. 
To avoid unnecessary diversion the parasitic, symbiotic, and miscel- 
laneous introduced species are here omitted from discussion. Excluding 
these, there are in the four natural groups that have been described 12 
well-defined associations of plants exhibiting habitat preferences which, 
on the evidence obtained from careful observation, must be correlated 
first of all with soil-water, but in some cases quite as clearly with tem- 
perature. Other factors, especially aeration and percentage of alkali salts, 
apparently have a dominant influence in one or more of these associations. 
As already noted, it has been observed that in several of the associa- 
tions plants of widely different physiological requirements grow in close 
proximity. Some of the more conspicuous cases are the Brgelowra, which 
on the flood-plain covers the ground under the mesquite, and Parkin- 
sonia microphylla, which is one of the most constant associates of the 
sahuaro on the hill. It has also been observed that some species, notably 
the creosote-bush and one or more of the salt-bushes, are by no means 
confined to their own special habitats, but grow even more vigorously 
in quite different situations beyond their limits. These phenomena are 
significant and are referred to in a later section. 
EDAPHIC RELATIONS. 
The conclusions arrived at in the study of the plant associations of 
Tumamoc Hill and the adjacent valley, though based on long-continued 
observation, nevertheless required confirmation. The work, therefore, 
was extended so as to include detailed mapping and more thorough 
study of the local distribution of certain species, comparative observa- 
tions in the Gila Valley and elsewhere, temperature determinations, and 
finally a special investigation of the soils of the Laboratory domain. 
Examination of the maps (plates 13-18) in connection with the text, 
shows that, in the local distribution of the species selected, character of 
soil and, in certain cases, aspect preference, are evidently two most impor- 
tant factors. The preference of Encelia farinosa and Cereus giganteus 
for southern exposures is strikingly shown by the maps (plates 13, 15, 
and 16). Equally clear is the fact, shown by plates 14 and 18, that 
the creosote-bush and mesquite exhibit no aspect preference whatever. 
On the other hand, the sharp delimitation of the areas of thickest growth 
of these two species, in connection with the great differences of soil on 
the flood-plain and the Tumamoc slope, renders it hardly possible to 
avoid the conclusion that soil relations have determined their respective 
habitats. Through all the careful work which was done in connection 
with the mapping of distribution of species, these two facts, the deter- 
mining influence of soil conditions on the one hand and of aspect on the 
other, have been constantly emphasized. 
