EXPERIMENT STATION BULLETINS. ' 511 



location of water obtained in tlie preceding series of experiments was 

 procured from column of soil with uniform moisture content. As will 

 be shown subsequently there is no doubt whatever, but that this maxi- 

 mum thermal translocation of water in the various soils would have been 

 far greater if the moisture content of the cold column of soil was less 

 than that of the warm column of soil. In nature, as already mentioned, 

 the moisture exists in a gradient form and consequently the movement 

 of w'ater is upward, and the forces of the factors which cause this up- 

 ward movement are increased during the night. Hence, ivJiile the amount 

 of thermal translocation of icatcr during a single night in hare soils 

 under field conditions may not he as great as that ohtained in the fore- 

 going series of experiments, yet it mill he quite appreciable ; and since 

 the process is repeated, the sum of ivater translocation of all the nights 

 during the vegetative season, ivill prohohly he considerable. 



The moisture content at which the maximum thermal translocation of 

 water occurs, or what has been designated as the thermal critical mois- 

 ture content, is very significant and needs further consideration. It 

 would be of very great interest to know, for instance, what the thick- 

 ness of the water film around the particles is at this degree of mois- 

 ture. This thickness could be calculated if all the soil grains were solid 

 and spherical. The particles of the soils used, however, — and these are 

 the commoner types of agricultural soils — are neither spherical nor solid. 

 Nearly all the particles in agricultural soils can be said t© be irregular in 

 shape; some of them are solid and enveloped with a colloidal coating; 

 others are compound aggregates or ''crumbs'' and are porous; and still 

 others, mainly of the peat nature, are of a sponge structure, and are 

 necessarily porous. The particles of a soil or soils may be classified 

 under two categories: (1) particles Avhicli are solid and have only an 

 external surface, and (2) particles which are partly or wholly porous 

 and possess both an external and internal surface. In the case of the 

 solid and cleaned surface particles, the water film is spread over the 

 surface, but in the case of the solid particles coated with colloids, or 

 the mineral flocules and the organic particles, the film of water en- 

 velopes, theoretically, their whole external surface, and also water per- 

 meates their internal surface. The single solid mineral grains, which 

 may compose the compound particles, may be cemented together in a 

 way analogous to that found in a piece of sandstone, in which case the 

 water exists only in the interstices and not as a complete film around 

 each particle. Furthermore, whether the soil grains are solid or spheri- 

 cal, or compound and porous, the water film is not uniform in thickness 

 over the entire inner surface of the soil mass, but thickena more at the 

 capillary angles between the particles. 



In view of these considerations, therefore, it was considered useless 

 to attempt to compute the thickness of the film as many investigators 

 have done. Furthermore, in view of ihe nature of the soil particles as 

 discussed above, it docs not appear sirictly proper io define the capillary 

 icater in the soil as many writers do as a thin film overspreading the par- 

 ticles and thickened into a luaist-lilie form at their points of contact. 

 Hence, a new definition of capillary water is needed. 



If we are to accept the theory which has been used to explain the fore- 

 going phenomena of thermal translocation of water that the soil 

 possesses ar very great attraction for water, that (his attractive force is 



