H. A. Tempany 327 



draw interesting deductions regarding the amount of colloidal clay 

 contained in any sample of soil from a knowledge of the Hnear shrinkage. 



It has already been shown that the observed linear contraction in the 

 case of any particular soil depends directly on the content of colloidal 

 clay, consequently it is obvious that if the maximum contraction 

 exhibited in the case of pure colloidal clay is known it should become 

 readily possible to calculate the proximate content of colloidal clay in 

 any particular example. To efEect such a calculation it Avill be merely 

 necessary to multiply the observed linear shrinkage per cent, by the 

 reciprocal of the Umiting value for the shrinkage of colloidal clay and 

 by 100. 



The value of this factor on the basis of the above result is 4-25. 

 Calculations performed in this manner will of course be only approximate 

 but at the same time they are capable of comparison with known data 

 obtained by other means. 



It is a well-known defect of the ordinary method of physical analysis 

 that while they are capable of effecting the separation of the coarser 

 particles of the soil with considerable refinement of accuracy they do 

 not give any exact indication of the amount of colloidal material con- 

 tained in the soil, a factor which profoundly affects its agricultural 

 value. In such physical separations the whole of the colloidal material is 

 contained in the four fractions designated silt, fine silt, clay and organic 

 matter ; usually the amount of colloidal material contained in the silt 

 fraction is small but in the writer's experience, especially in the case of 

 soils containing much colloidal material, it is impossible completely to 

 free the silt fraction from colloidal matter. In the fine silt and clay 

 fractions the bvilk of the colloidal material is always contained, but on 

 the other hand it frequently happens that much material of a non- 

 colloidal nature is also brought down with the colloidal material, and 

 it is impossible by the ordinary methods of physical analysis to give 

 expression to the actual amount of colloidal material present. 



From the foregoing it is obvious that the calculation of the content 

 of colloidal clay by means of a factor of the type indicated above, is 

 capable of being checked inasmuch as, the whole of the colloidal material 

 being contained in the fractions indicated, the total content of colloidal 

 clay cannot exceed the total content of the four fractions in question 

 and should be generally less. In point of actual practice since the bulk 

 of the colloidal material is contained in the fine silt and clay fractions,' 

 it is sufficient for purposes of comparison to compare the content of 

 colloidal clay calculated from the observed linear shrinkages and the 



22 2 



