ON COLLOID CHEMISTRY AND ITS INDUSTRIAL APPLICATIONS. 735 
reaction is the typical adsorption shown by colloids, and can be 
expressed by the equation which has been found to fit so many 
of them. It is considerably influenced by the acid, being greater 
in the presence of the mineral acids than of the organic acids. 
The amount of phosphorus compound actually brought out is the 
difference between the direct and the reverse action. Thus, 
hydrochloric acid dissolves out a certain amount of phosphate, 
but considerable adsorption takes place, so that the net amount 
_left in solution is only small. Citric acid dissolves out the 
same amount of phosphate, but there is much less adsorption, 
and therefore the amount left in solution is markedly greater. 
The difference between the various dilute acids les not in their 
solvent power, which is similar for all, but in their influence on 
the adsorption process. The observed net effect of the acid on 
the soil is therefore expressed by the ordinary adsorption curve. 
The reaction between dilute acids and soils thus appears to be 
essentially a displacement of absorbed material by something 
which is itself absorbed; and it falls into line with other displace- 
ments from colloids. 
If this view is correct, it follows that acids need not be used in 
soil analysis, at any rate for the extraction of bases: any agent 
capable of being absorbed by the soil would serve equally well. 
Ramann has used a solution of ammonium nitrate and finds 
it dissolves at least as much potassium, calcium, ete., as an acid, 
and in some respects is more satisfactory. This method of 
extracting the bases from soils promises to be useful and to clear 
up many of the difficulties in soil analysis. 
The water relationships of soils. 
The earlier soil physicists, regarding soil as a mass of mineral 
matter, began their studies of the water relationships of soils by 
treating soil as a mass of sand, or, in mathematical analysis, of 
small bullets or marbles, and investigating the distribution of 
water by surface tension. The conclusions obtained were not 
wholly in accordance with facts. The facts observed in the 
evaporation of water from soil were particularly difficull to fit 
in with expectations, and numerous breaks in the curves seemed 
to indicate the existence of a variety of critical points and 
co-efficients. Keen has shown, however, that the phenomena 
become much clearer when regarded as colloidal effects, He 
found that the relationship of water to soil differs considerably 
from its relationship to sand. The evaporation of water from 
sand, silt, china-clay, and ignited soil proved to be relatively 
simple, and could be explained by the known laws of evaporation 
and diffusion. But the evaporation of water from soil could not: 
it was more complex. Instead of the simple proportionality 
between water content and time observed in the case of sand, the 
curves for soil were more exponential in type. The difference was 
traced to the soil colloids, and it disappeared when the soil was 
ignited and the colloidal properties lost: the curve then became 
identical with that obtained for sand. The influence of the 
colloids has so far only been expressed empirically, but it is 
