14 DISTRIBUTION AND MOVEMEN’S OF DESERT PLANTS. 
especially Sueda moquint, grow more luxuriantly than in the salt-spots 
themselves, but this is a physiological response characteristic of plant we 
in general, and can hardly be taken in evidence. 
The question of soil preferences involves great difficulties, due in part 
to the apparently contradictory deportment of many plants that have 
been cited in evidence, and still more to lack of sufficiently extended 
observation and experiment. The interesting experiments of Kearney 
and Harter (1907) on the comparative tolerance of various plants for the 
salts common in alkali soils shows clearly the great complexity of the 
subject when the relation of any plant to a mixture of alkali salts is under 
investigation. On the other hand, an illuminating discussion of the 
soil preferences of alpine plants by Fernald (1907) has brought forward 
so many incontrovertible facts, established by years of careful observa- 
tion, that it no longer seems possible to doubt that the chemical com- 
position of underlying rocks has been an important factor in determining 
the distribution of the remarkable ecological group to which his study 
has been directed; the whole question, however, calls for experimental 
investigation. 
(5) THE WASH; PALO VERDE=-CATCLAW ASSOCIATION. 
At the foot of Tumamoc Hill, on the northwest, there is a ‘‘ wash,” 
formed by the union of a number of similar ones which come down from 
the long slopes to the westward, the system presenting throughout its 
extent the usual features of a dry watercourse, with sandy bottom, along 
which a dense vegetation of shrubs or low trees marks its course even 
when seen at a distance (plate 5). The wash with its branches consti- 
tutes a drainage system for the west slope of Tumamoc Hill on the one 
hand and the east slope of the nearby hills of the Tucson Range on the 
other, forming the natural channel or run-off into which water runs or 
seeps from the adjacent banks. Though its bed is usually dry, there 
are abundant indications of the presence of water below the surface, 
especially in the character and habits of its vegetation. In the first place, 
certain species, notably the creosote-bush, attain here a size and vigor 
of growth in striking contrast with the low, straggling individuals of the 
slopes nearby; and, secondly, other species, particularly the mesquite, 
everywhere taken as an indicator of water, are conspicuously present. 
Both of the species just named are prominent constituents of the vege- 
tation of these washes, but since their characteristic habitats are else- 
where, other species have been chosen to give their name to this asso- 
ciation, viz, the palo verde (Cercidium torreyanum) and catclaw (Acacia 
greggit). 
At present, as the process of base-leveling slowly proceeds, these 
species are advancing from the flood-plain toward the adjacent hills. 
Cercidium torreyanum (plate 6) is already found far up one of the gulches 
