C. L. Whittles 167 



number first, followed by the appropriate class number. This is sum- 

 marised in Table II. They point out that each type has a correspondinrj 

 type into which it can merge by inappreciable degrees, the transforma- 

 tion being effected by interchange in the proportions in the ingredients 

 which are of primary and secondary importance respectively, e.g. 1 , 2 

 to 2, 1. On the other hand, they can diverge far from each other in 

 character. Other relationships are pointed out of which the most im- 

 portant perhaps is the change in the secondary ingredient, as for example 

 from i, 2 to 1, 3. 



Table II. Hall and Carpenter s Classifivaiion of Soils. 



The earliest reports of mechanical analyses were often illustrated by 

 photographs of the actual fractions obtained. The sand, silt, etc., were 

 placed in small phials of uniform size, and the predominance of the bulk 

 of the material in certain grades served to give a more vivid idea as to 

 the meaning of the figures of the analysis. The method at that time had 

 its value, for the purpose of a mechanical analysis was not so well known 

 then as now. Hilgard's text-book (21) and the early bulletins of the U.S. 

 Department of Agriculture contain illustrations of this type. 



It is only a short step from the use of the actual fractions to that of 

 shaded blocks. In Whitney's report on the tobacco soils of Maryland (57) 

 a general summary of the types is given in diagrammatic form in which 

 three grades are distinguished by different shading, viz. sand, silt and 

 clay. 



In order to compare several soils or soil types on the same diagram 

 the method of plotting the percentage of each grade against an arbitrary 

 scale of grades has frequently been adopted. Differences are here far 

 more obvious than resemblances, and their correct interpretation is 

 almost as difficult as that of the figures themselves. The plotting of the 



Journ. of Agric. Sci. sn 12 



