246 BOTANICAL GAZETTE [OCTOBER 
dry and saturated soil the water-retaining power is calculated in 
the usual way (3). 
The actual soil moisture content was determined by the weigh- 
ing and drying method, the temperature of drying being 105-110" C. 
The samples were taken, by digging, at depths of 15 and 3o cm. 
for all soils but that of the Larrea slope. For the latter soil, on 
account of its shallowness, the depths were 10 and 20 cm. Samples 
for each soil type were taken from the same area of a few square 
meters, the small excavations being immediately refilled. No later 
sample was taken from within 50 cm. of the soil thus disturbed. On 
December 13 irrigation water made it necessary slightly to change 
the position of the station for the soil of the floodplain, otherwise 
the stations remained the same throughout the whole period. 
Samples were taken at intervals of about 10 days. Three samplings 
were omitted for all soils during the month of June 1908. A few 
other omissions of single samples occur. 
The specific gravities of these fours soils are approximately the 
same, there being almost no content of organic matter. Therefore 
the moisture contents are calculated on the basis of the dry weight 
of the soil and are here approximately comparable. 
The careful work of Mr. J. C. Brumer, Mr. E. E. SHERFF, 
and Professor J. E. Krrxwoop, who performed the operations of 
collecting, weighing, and drying the soil samples, as well as that of 
keeping the rain record, is here gratefully acknowledged. 
The accompanying graphs (figs. 1-4) have been constructed 
for each soil separately; the lesser depth is represented by a thin 
line and the greater by a heavy line. The precipitation record 
was kept for the hill station alone, but there is no reason to doubt 
that the rain record for the other stations would have been essen 
tially the same had it been actually taken. For easy comparison, 
the rainfall is represented below the graphs of soil moisture simply 
by vertical, shaded blocks, the length of each block denoting the 
relative amount of precipitation which took place during the 
corresponding ten-day period. The figures directly above these 
blocks denote the precipitation in centimeters, that is, the relative 
length of the blocks. The data from which the moisture graphs 
are constructed are placed upon the figures directly above the base 
