Relation of Nonavailable Water to Hygroscopic Coefficient 59 
not entirely dead until May 13. The cylinder was opened on 
that date. Xo free water was present in the surface soil and 
there was practically none in the subsoil. The former contained 
— 4.3 per cent free water and the latter 0.4 per cent. The only 
roots that had entered the subsoil extended about two inches be- 
low the surface foot. They are shown in figure 13. The dry 
matter in these roots amounted to 0.003 gram. 
The explanation of the lack of difference in growth during 
the first month is evident from the above. As long as the plants 
in all six cylinders had their roots confined to the surface foot 
of soil, all had about equally favorable conditions and made 
v VI xx XIX XIV XIII 
Fig. 13. Roots of Red Fife wheat; experiment of 1910. 
similar growth, but as soon as the roots were developed to the 
bottom of the surface foot the plants in XIII fell behind the 
others ; the roots did not penetrate the subsoil on account of the 
absence in it of free water. The final dryness of the subsoil as a 
whole was due simply to its not having had any water to give up 
to the plants, while that of the surface soil was due partly to the 
loss thru transpiration but chiefly to direct evaporation. The in- 
crease of 0.2 per cent in the average moisture content in the five 
feet of subsoil may be due to a slight distillation from the moist 
surface soil early in the experiment, but it is probably rather to 
be considered within the limits of the experimental error. 
