24 BOTANICAL GAZETTE [JULY 
residing in the seed substance at any moisture content between 
saturation and air-dry. The actual values for this range in Xan- 
thium seeds are given in the last column of table ITI. 
The results obtained by measuring the soils with measured seeds 
are of the greatest interest. While it is important to have means 
of determining in terms of atmospheric pressure the water-retaining 
power of soils at any degree of dryness between saturation and air- 
dry, it is still more important to understand the moisture relations 
of the plant to the soil at and just below the wilting coefficient. 
Does wilting occur because capillary soil forces and osmotic root 
forces reach equilibrium? If so, why should Briccs and SHANTZ 
(8) have found a uniform wilting coefficient for all kinds of plants 
when we know that root cells vary somewhat in osmotic concen- 
tration from species to species? And why should this uniformity 
fail in the presence of intense evaporation, as shown by CALDWELL 
(9), and by SutvE and LivincsTon (32)? These questions, and 
the discrepancies between the excellent work done at Washington 
and at Tucson can probably be answered intelligently, and explained 
in the light of these experiments. 
The moisture equivalent of a soil is the percentage of water left 
in the soil after centrifuging for a certain time under a force of 1000 
gravities, a force about equal to one atmosphere. We have been 
accustomed to find the wilting coefficient empirically, or to divide 
the moisture equivalent by 1.84. But we have not known how 
much greater are the soil forces at the wilting coefficient than at 
the moisture equivalent. From the results obtained, the pressure 
value of the wilting coefficient seems to be about 4 atmospheres. 
As the soil becomes drier and drier, the forces become greater 
and greater, until on the approach of air-dry conditions a very 
small change in moisture content makes a very large change in the 
forces involved. The increase of the force with decreasing soil 
moisture is shown graphically in fig. 5, which shows the curve of 
increasing force for the Oswego silt loam subsoil, and for no. 2/0 
sand. The difference in the curves for clay and sand is striking. 
In the clay, where the surface films are relatively thick, the force 
decreases very slowly for a considerable distance. But as the films 
become very thin, the surface force and forces of adhesion of water ° 
