1919J Pendleton: A Study of Soil Types 169 



With regard to the methods of mechanical analysis, one should not 

 overlook Mohr's work on The Mechanical Analysis of Soils of Java. 

 which gives an excellent discussion of the relative merits of the better 

 known systems of mechanical analysis. He describes a modified cen- 

 trifuge method preferred by him. 



Under a discussion of the physical constants of soils. Free 31 dis- 

 cusses the value of mechanical analysis as a soil constant, and shows 

 that there are three serious errors in the determination, all of which 

 impress themselves upon one making and using such analyses. They 

 are: "(1) disunity of expression; (2) failure to express conditions 

 within the limits of individual groups; and (3) failure to take 

 account of variations in the shapes of the particles." Vet he empha- 

 sizes, and rightly so, "that mechanical analysis is by no means useless 

 nor to be belittled as a means of soil investigation." 32 



Moistun equivalents. — This determination showed quite distinct 

 averages for the types, though there was considerable variation within 

 each of the types. Eliminating those samples shown to he non-typical 

 according to the mechanical analysis, the variation within the type is 

 reduced considerably. Yet it cannot be said that as regards this con- 

 stant that all soils mapped under a given type name, or even those 

 soils under a given type name which the mechanical analysis has 

 shown to be true to name, have closely similar moisture equivalents. 

 Briggs and McLane 33 express the belief that ultimately moisture 

 equivalent determinations will replace mechanical analysis in the 

 classification of soils, because the determination is simple and the 

 result can be expressed as a single constant. 



Hygroscopic coefficient. — The two heavy types show averages dis- 

 tinct from those of the two light types, but the wide and erratic varia- 

 tion within the type, together with the nearly universal failure of 

 Briggs and Shantz's formula 34 to convert these values into values 

 even approximating those of the moisture equivalent, leads one to 

 doubt the accuracy of these figures of the hygroscopic coefficient. It 

 is because of the ease of determining the moisture equivalent, and 

 because of the difficulties involved in correctly carrying out the hydro- 

 scopic coefficient, that the doubt is cast upon the latter determination. 



so Bull. Dept. of Agr., Indes Neerlancl, 1910, no. 41, pp. 33. 

 3i Free, E. E., Studies in Soil Physics, Plant World, vol. It (1912), aos. _. 

 3, 5, 7, 8. 



32 Ibid., p. 29. 



33Proe. Amer. Soc. Agron., vol. 2 (1910), pp. 138-47. 



34 U. S. Bur. PI. In<L, Bull. 230 (1912), p. 72. 



