278 Bovie: NoN-AVAILABLE WATER IN SOILS 
Not only this, but the root hairs, with their cellulose walls, 
are absorbents, and undoubtedly engage in selective absorption 
with the soil grains, so that there are ratios of distribution, between 
the soil grain, the film of soil water, and the cell wall." Physio- 
logical processes in the plant will continually disturb the equilib- 
rium of distribution. A series of changes must then follow before 
the equilibrium is again established. It was with the hope of 
getting some data regarding the nature of this equilibrium that 
the experiments described below were undertaken. 
EXPERIMENTS AND RESULTS ON THE INFLUENCE OF SALTS ON 
NON—-AVAILABLE WATER 
From a number of plants used in preliminary experiments, 
wheat was selected as most suitable. Its superiority lies in the small- 
ness of its seed, so that it early becomes dependent upon the soil solu- 
tion, and in the tenderness of its leaves, which makes it very sensi- 
tive to a deficient water supply. Some of the plants were grown 
in paraffined baskets; others in flower pots dipped in hot paraffin. 
Both the baskets and the pots seemed to give good results, but the 
pots were easier to prepare. Later, when it was found, by weighing 
the empty pots, that they had slowly absorbed water, their use 
was discontinued. Glass tumblers were then tried, and were found 
very satisfactory. The roots spread evenly through the soil, show- 
ing no tendency to mat about the inner surface of the glass. 
In the first experiments, the plants were grown in river-washed 
sand, but in the experiments from which the data given below were 
taken, the plants were grown in clean crushed quartz, obtained 
from a supply house. The chemical analysis of the quartz, fur- 
nished by the supply house, is as follows: 
Per cent. 
POBOINES SHIGA Or a5 on cheer rag Se ie areas l eed Se gw 99.97 
PFO CPTI, CAINE gk tava Pe nee ie auc ks 
100.00 
The fresh samples were very white, but, after standing, a yellow 
tint, just perceptible when old samples were compared with new, 
was noted. This was probably due to thin films of iron oxide 
which formed on the surface of the grains.!. The physical analysis 
is given in the table below: 
