1904] SPALDING—THE CREOSOTE BUSH 131 
creosote bush on the mesa and foot-hills is still transpiring consider- 
able quantities of water. The amount transpired appears to stand 
in direct relation to the amount of water available in the soil where 
the plant is growing, as is indicated by the following comparison of 
percentages of moisture given off by the soil when air dried. 
Samples of soil were taken at depths of 20 to 30°" below the 
surface from points near the plants on which the transpiration experi- 
ments were conducted. In each case the sample was weighed, then 
left in a shallow basin in the air, exposed to sunlight, but protected 
from draughts of wind, for three days, after which the weighing was 
repeated. It was found that the soil from the laboratory hill, taken 
at a depth of 30°™ below the surface, lost by air-drying during this 
period 8 per cent. of its weight, while that from the plain near the foot 
of the hill, which was much drier and in which the creosote bushes 
were evidently suffering from lack of water, taken from a depth of 
20 to 25°™, lost at the same time 3 per cent. Another sample from 
the hill lost by heating over an electric stove 12 per cent. of its weight. 
The days when the drying was done the relative humidity of the atmos- 
phere ranged from 20 to 27 per cent. 
It is of course essential that much more extended and critical 
work in this direction should be carried out. Meantime the impor- 
tant fact is established that after months of excessive drouth the soil 
in which creosote bushes were living, taken only a few inches below 
the surface, gave up when air dried 3 to 8 per cent. of its weight of 
watery vapor, while a considerably higher per cent. was driven off 
by heat. This fact being proven, our interest chiefly centers in the 
capacity of the plant to utilize the available soil water after it has 
been so greatly reduced. This involves a study of the root system. 
By way of ascertaining first general facts, the roots of creosote 
bushes were examined by carefully removing the earth in which they 
were growing, and then following their ramifications as far as possi- 
ble. This is not an entirely satisfactory procedure, inasmuch as it 
is quite impracticable to follow the finest roots to the end without 
breaking them off. It is possible, however, to lay bare so large a 
part of the root system as to obtain a clear view of its direction of 
growth, mode of branching, and other characteristic features. Fig. 5 
isa photograph of two seedlings of Covillea, a few months old, that 
