E. J. Russell and J. A. Prescott 109 



These are plotted in Fig. 6, and the two curves superpose. The 

 24 hours' treatment then, much as it altered the soil, did not alter its 

 adsorptive capacity. 



There remains the possibility that the solution obtained at the 

 end of 24 hours affects the equilibrium differently from that obtained 

 after Time = 0. There is considerable a 'priori justification for such 

 a view, inasmuch as the addition of alumina to the nitric acid depresses 

 the amount of P2O5 extracted, and may therefore very well increase 

 the amount of adsorption. 



This was tested by treating two lots of the same soil (a) with N/10 

 HNO3 that had already been in contact with the soil for 24 hours, and 

 (6) with fresh HNO3 of equivalent concentration. If the cause of the 

 difference of adsorption lay in the changed condition of the liquid 

 then adsorption from the 24 hours' liquid should be greater than that 

 from the fresh acid. 



Experiment showed that this actually happened. The curve 

 obtained with fresh acid lay below that obtained with the 24 hours' 

 extract, the two being related in the same way as those for Time = 

 and for Time = 24 hours (Table X, Fig. 7). 



Conclusions. 



The results enable us to explain what happens when a soil is shaken 

 with a dilute acid in the ordinary process of soil analysis. 



The acid dissolves out such phosphorus compounds as it can, and 

 different acids have much the same effect at equivalent concentrations : 

 nitric, hydrochloric and citric acids give the same results; sulphuric 

 acid, however, gives a somewhat higher result. 



A reverse reaction at once sets in, however. Some of the P2O5 is 



withdrawn from the solution in spite of the presence of excess of acid. 



The process is an ordinary adsorption and obeys the usual law expressed 

 1 



by the equation xj = /ic^. Its extent varies with the different acids; 



it is much more marked in the presence of nitric than of citric acid. 



The amount of PgOg actually determined by the analyst is, therefore, 

 not the true amount dissolved, but the difference between these two 

 wholly distinct actions. 



It is now obvious why the amounts of "available P2O5" determined 

 by extraction with dilute acids shows such great variations in different 

 methods of analysis, and so little correlation with the actual quantities 



