102 



. It will be seeii, from this table that there came to be, at the 

 eiKl'O^fp^daysj a marked inequality in the distribution of the 

 water^olubie salts which could be recovered by washing in dis- 

 tilled water. Instead of the perfect uniformity which existed 

 at the time the cylinders were charged with the soil, the 70 days 

 of capillary movement has resulted in very large concentra- 

 tions, especially of the nitrates, chlorides and sulphates, at and 

 near the surface. The nitrates, for example, range from; 20.4 

 parts per million at 15 to 18 inches below the surface to 1140 

 parts per million in the surface inch of the mulched surface; 

 and from 15.2 parts at the bottom to 1076.3 parts at the top, 

 where the soil was firm. 



It will be seen that the water-soluble salts in the upper layer 

 of the loose soil of the mulcheed cylinders are often greater than* 

 in the corresponding layer of the unmulched or firm surface. 

 This difference results partly from the fact that the weight of 

 soil in the loose condition is less than where the soil was firm, 

 and hence the same amount of salts brought into the surface 

 inch represents ,a larger number of parts per million of the 

 soil. Because the data of the last table, p. 101, are not fully 

 comparable, on account of the differences stated, there have 

 been brought together in the next table the absolute amounts of 

 the several water-soluble salts which were recovered from, the 

 respective levels. These amounts are obtained by multiplying 

 the observed dry weights of soil recovered from each layer by 

 the parts per million taken from the last table. As the weight 

 of the soil was obtained in pounds the results are in pounds, 

 but on account of multiplying by parts per million they are one 

 million times too large, and are stated in this way to avoid long 

 decimals. In studying the data of this table it will be needful 

 to bear in mind that because there are approximately 3 times the 

 amount of soil in the 3-inch layers that there are in the l-inch 

 layers, the amounts in the 3-inch layers appear to be relatively 

 higher than is the case. To avoid this confusion the totals for 

 the three 1-inch layers are also given in the table. The data of 

 lines 0-3, 3-6, etc., to 15-18, show the distribution of the sev- 

 eral salts in the respective levels in a strictly comparable man- 



