1 64 



Journal of Agricultural Research 



Vol. V, No. 4 



it is somewhat greater in soils with higher than with lower colloidal con- 

 tent, and (3) that it increases with the rise in moisture content. 



By comparing these results with those obtained with columns of soils 

 of uniform moisture content, some very striking contrasts are revealed. 

 The previous results show, for instance, that the maximum thermal 

 motion of water occurs at a definite but comparatively low moisture 

 content and that the value amounts in some cases to more than 3.50 

 per cent. The above data show, however, that the maximum movement 



'IVafer mouec/ from mo/s^ so// a^ ^W"c. fo e/ry so// 3/ 0°c. 

 - « ., .. „ ,. .- O'c. " '- " "^O'c. 



20°c. 

 O'c. 



O'c. 

 .' 20°c. 



Fig. 8.— Curve showing the percentage of moisture moved from a moist and warm column to a dry and cold 

 column of Miami silt loam, and from a moist and-cold to a dry and warm column of Miami silt loam. 



of water from the moist and warm column to the dry and cold column 

 of soil takes place at the highest water content and that in the majority 

 of ca§es the percentage of this maximum water translocation is only 

 one-half as great as in the former case. 



These apparent differences seem to be easily explainable. The increase 

 of water movement from moist and warm soil to dry and cold soil with a 

 rise in water content is natural and only goes to prove that the water 

 is held by the soil with low moisture content with great force, and con- 

 sequently it can not be extracted readily and extensively by a greater 



