Relation of Nonavailable Water to Hygroscopic Coefficient 67 
foot. The development of the roots was similar to that in XV, 
except that there were many more in the second and third feet 
and very few in the sixth. It seems probable that under favor- 
able weather conditions more or less grain would have been 
formed. It is interesting to note that in this cylinder, as in XIV 
and XV. the moisture in the second, third, and fourth feet had 
been reduced much below the wilting coefficient. 
Cylinder with 12 per cent eree water in the subsoil. — 
Cylinder XXII differed from XVII in that the subsoil carried 17.8 
VIII VII XXII XVII XV XI 
Fig. 15. Roots of Kubanka wheat; experiment of 1910. 
per cent total water. Two plants in this ceased to grow about the 
middle of March and gradually declined, altho neither was 
actually dead even as late as June 8. Each of the other two 
plants developed 3 tillers and each produced a spike, the one con- 
taining 18 grains but the other forming none. The death of both 
plants was due to the high temperature of June 12. The subsoil 
still contained much free water, which increased from 1.1 per 
cent in the second foot to 8.9 per cent in the sixth. The roots ex- 
tended to the bottom of the cylinder, but those in the lower por- 
