88 DISTRIBUTION AND MOVEMENTS OF DESERT PLANTS. 
well moistened with each period of high water. The water readily pene- 
trates to great depth, and appears to be fairly well held, at least in the 
deeper layers. It is probable, although no direct evidence is at hand, 
that these soils possess a subterranean water-table, at least during many 
months of the year, and it seems not unlikely that the larger perennials 
may obtain moisture from this source. 
(4) The soil of the river flood-plain is of unknown depth, without caliche 
or other impervious hardpan. At the time of the digging of the well for 
the Laboratory water-supply, an opportunity was presented to observe the 
conditions of this soil from the surface layers to depth of permanent water. 
The surface layers are composed of a clay loam which possesses a 
water-holding power of 38.5 per cent of its dry weight. At the well site 
this loamextends to adepth of about 5 metersand then gives place to strati- 
fied sands and gravels, with some water-worn bowlders at a depth of about 
g meters. Permanent water, in coarse sand, was met at a depth of about 
12meters. Nearer the river channel subterranean water occurs ata less depth. 
At the time the well was dug, March, 1906, a number of determina- 
tions of the water-content of this soil were made. Samples were taken 
in closed bottles, weighed, dried in an oven at a temperature of from 
105° to 110° C. Table 3 gives the moisture condition of the soil at this 
time in terms of percentage calculated on the dry weight of the soil. 
TABLE 3.—Morsture Conditions of River Plain, March, 1906. 


Nature Water 
bs atare of soil. content. 
Meters. Peet. 
0.20 Loam. £250 
3.30 Loam. 19.0 
4.00 Loam. 2250 
5.10 Loam. ee | 
Sues Sand. rae 





From table 3 it is apparent that the water-content increased from 
the surface downward to the limit of the loam, and that in the sand the 
water-content was low, too low to be of much benefit to plants, the roots 
of which might penetrate to such a depth. No determinations were 
made in the vicinity of the water-table, but it may safely be assumed 
that for some distance (perhaps a meter or two) above this the sand 
usually contains a considerable amount of water. 
The relatively great depth of the water-table makes it appear that the 
general vegetation of the flood-plain derives little, if any, of its water- 
supply from this source. Some trees here, especially the native mesquite 
(Prosopis velutina), may send roots deeply enough into the soil to reach 
layers kept permanently moist from the underground flow, but the natural 
herbaceous growth on the plain probably derives all of its water-supply 
