68 



Journal of Agricultural Research 



Vol. XV, No. 2 



with that of distilled water. The results, in terms of air-dry soil, ob- 

 tained with a sample from plot 4, were : 



This shows that, under the conditions maintained, the N/j solution 

 had about 2.7 times the solvent effect of water alone; and the normal 

 solution about 3.1 times that of water. In other words, trebling the 

 concentration of the salt solution increased its solvent effect little more 

 than one-eighth. 



Hence, as the N/j solution introduced no more ammonium chlorid 

 than could be handled with a fair degree of convenience in the analytical 

 operations, a solution of that concentration (17.6 gm. of salt to 1,000 

 cc. of water) was used in the remaining studies with the solvent. 



The average results for nine samples of each soil were, in terms of the 

 air-dry soil and corrected for blank : 



Percentage of 

 potash (K2O). 



Plot I, untreated o. oo936±o. 00012 



Plot 4, dressed with potash 01872 ± . 00039 



In order to ascertain the relative solvent effect of a second extraction 

 with this solution after two of the treatments, one for each plot, included 

 in the above averages, the once-extracted soil residues were carefully 

 washed with a little water to remove the excess of the salt solution, 

 dried on the filter, and again subjected, as before, to the action of i 

 liter of the fresh salt solution. The results obtained in the respective 

 extractions were : 



Percentage of 

 potash (K2O). 



Plot I, first extraction o. 01002 



Plot I, second extraction 00198 



Plot 4, first extraction. 02210 



Plot 4, second extraction 00208 



The interesting fact appears that the amounts of potash , removed in 

 the second extraction were almost identical for both soils, and in both 

 cases were much less than the quantities removed by the first extraction. 

 In order to check this finding, a second portion of plot-4 soil was 



