36 DISTRIBUTION AND MOVEMENTS OF DESERT PLANTS. - 
habitat with Parkinsonia microphylla, namely, the sides of the gravelly 
bluffs skirting the washes here. In the third physiographic area, also, 
a few instances are found where it occurs away from the washes, but 
probably these may be traced to the courses of former streams, which 
may be the courses of present underflow, or to places of comparatively 
extensive dissipation of currents within recent time. 
This species is much more abundant in the second physiographic area 
than the first, and the streams and streamlets are followed much farther 
up. It is also more numerously represented in the second than in the 
third area, but is most abundant in the fourth, or the wash bottom. By 
far the greater number of individuals stand in immediate proximity 
to the washes, where the periodic flow touches, or almost touches, the 
stem and undoubtedly reaches the roots. Of these a large number, pos- 
sibly a majority, stand either on the point of the prong formed by the 
junction or division of two washes, or on either bank immediately opposite 
such junctions or divisions. These places indicate the most copiously 
watered soil. 
By reference to the map (plate 17), it will be seen that along the main 
wash congestion, so to speak, takes place either at the principal bends 
or where the bed of the wash widens or divides and subdivides into numer- 
ous channels. Where the channels are comparatively narrow, straight, 
and without forks, the number of trees is comparatively small. ‘This 
and the foregoing paragraph indicate that this tree likes to have, it seems 
almost must have, its roots immersed in the periodical floods at points 
where the rapidity of their current is checked and their volume partly 
dissipated in the adjoining soil. 
At some points in the vicinity of Tucson, especially to the southward, 
on the road to the Sierritas, Cercidiwm torreyanum occurs on higher ground, 
and at higher altitudes to the eastward it takes its place with the general 
shrubby growth of the mountain slopes; but nothing has as yet been 
observed conflicting with the general fact that at this place it belongs 
to areas of transported soil, especially washes such as those of the Labo- 
ratory domain, where periodically the water-supply is very great, and 
probably never falls to the minimum of closely adjacent areas. An abun- 
dant water-supply and effective drainage are apparently the local factors 
determining its distribution. 
PROSOPIS VELUTINA. 
As already stated, the mesquite stands in sharp contrast to the creosote- 
bush in regard to distribution, the latter being distinctively a plant of 
erosion areas and the former of areas of deposition. There are occasional 
exceptions, or apparent exceptions, to this, as for example where, south of 
Tucson, a broad mesquite belt occupies a long, gentle slope from the 
Santa Ritas, joining the forest of the Santa Cruz flood-plain in a series 
