288 BoviE: NON-AVAILABLE WATER IN SOILS 
roots of the plants in the higher concentrations came into contact 
with a much smaller proportion of the total surface presented by 
the soil grains. Yet, the percentage of non-available water, re- 
maining in these tumblers, was no greater than that found in the 
lower sets, which indicates a water transfer from the more remote 
parts of the tumbler to the root systems of the plants. At first, 
when the soil was very moist, this transfer might have been made 
by the movement of the liquid water over the soil surfaces, but as 
these films became thinner, the movement would become slower, 
and at last, when the films became discontinuous, the movement 
would stop altogether. The only method of transfer, then, would 
have been in the form of vapor, and before the soil moisture could 
have been reduced to as low a percentage as was found in these 
experiments, this kind of transfer evidently must have taken place. 
But little is known concerning the vapor pressure of soil mois- 
ture. It may be expected, however, to vary with the shape of 
the film surfaces, with the surface tension, with the thickness of 
the film, and with the concentration of the solutes present. If, by 
a change in any of these conditions, the vapor pressure of the soil 
moisture about the roots is lowered below that of the surrounding 
soil, there will be a movement of water vapor toward the plant 
roots, and then a condensation. More than this, the water which 
permeates the plant roots together with their root hairs has a 
vapor tension and thus the roots must either give off water in the 
form of vapor or take it in, as they are in an atmosphere with a 
vapor tension either lower or higher than that of their own mois- 
ture. The common practice of growing plants in a damp chamber 
is evidence of this. 
In the above-described experiments, the air cavities, in the 
quartz, were lined with projecting root hairs, which were often so 
numerous as to give the cavity the appearance of having walls 
covered with a delicate fur coat. 
The writer has undertaken some further experiments in regard 
to the relation of plants to vapor tension; while no numerical 
results have been obtained, the preliminary experiments show that 
plant roots are very sensitive to variations in vapor pressure. 
In regard to the influence of the concentration of soil moisture 
upon vapor tension and water transfer, the following experiment 
