39 



the soils were of the same depth, and if the roots penetrated as 

 easily in one as in the other. And in practice the conclusion is 

 shown to be just by the rapid exhaustion of sands overlying clay, 

 which at first give exceedingly luxuriant and fine tea, but which 

 become exhausted more rapidly than any other type of land in 

 Assam. Though, calculated in the non-sandy part of the soil, they 

 were very rich to start with, yet the total amount of plant food was 

 small, and was rapidly exhausted, while the roots were unable to seek 

 fresh supplies further afield because of the underlying clay. In the 

 example I have taken, however, it is otherwise. The soil No. i has 

 at least twenty feet depth of soil in which to range without meeting 

 clay or water, and hence produces luxuriant tea at the beginning, 

 and in addition lasts an almost indefinite time. 



We have therefore to consider in a soil analysis : 



1. The concentration of the soil constituents, i. e., the amount 



of the valuable matters in the part of the soil other 

 than sand. This determines the luxuriance of the 

 plants and probably the value of the tea. 



2. The absolute amount of the ingredients of value as repre- 



sented by the percentages in the soil as a whole. This 

 determines the permanence of the plant without 

 manuring, but only if taken in conjunction with. 



3. The available depth of the soil, which may counterbalance 



the apparent absolute poverty (No. 2) of the land, if 

 this depth is sufficiently great. 



These three ways of looking at the matter must all at least 

 be taken into consideration in interpreting the figures. If they had 

 been so considered in the past, they would often have led to a very 

 different manure recommendation, or to no recommendation at all. 



METHODS OF TAKING SAMPLES OF SOIL FOR ANALYSIS. 



In sending a soil for analysis, therefore, in order to satisfy 

 consideration No. 3, the character of the subsoil should always be 

 stated, and in addition the depth of the subsoil water, as these may 

 have a paramount influence in determining the meaning of the 

 analysis. The actual sample sent should always be taken in the 

 following manner. Dig a hole in the plot of which an analysis 

 is desired, fifteen inches deep, leaving a block in the centre of the 

 hole 6 inches square. This block is therefore 6" x 6" x 15* in 



