E. J. Russell and J. A. Prescott 



73 



in the soil becomes more marked after 24 hours than after shorter periods. 

 It is shown later, however (§ 15), that the discrepancy is not real. We 

 are therefore justified in speaking of the absorption as an adsorption. 

 8. The different acids have markedly different effects on the adsorp- 

 tion of phosphates from solutions of sodium phosphate. Adsorption 

 goes on readily in the presence of hydrochloric and nitric acids, but 

 is notably smaller in presence of citric acid in equivalent concentration. 

 It may be inferred, therefore, that the greater net action of citric acid 

 in comparison with hydrochloric or nitric acids is not due so much 

 to a greater solvent power, but to a greater power of reducing adsorption. 

 Thus we should expect the actual solvent action of these acids to be 

 more nearly alike than is indicated by the net action. 



y = 20-2 C2~4 , rtlO^^ 



21-2 C3~2 



8-8 C?. 



10 20 30r40 50 60 



PoOj mgm. per 1000 c.c. of solution 

 Fig. 3 a. 



0-5 10 



Log C 



Fig. 3 6. 



Fig. 3 a. Amounts of P2O5 adsorbed by soil Agdell B in presence of di£ferent acids. 

 The log. curves are given in ffig. 3 b (see p. 95). 



9. This expectation turns out to be correct. Adsorption can be 

 eliminated almost entirely by arranging the experiment so that the 

 phosphorus compound is removed from the soil as soon as it is dissolved. 

 This is readily accomplished by diffusion. The practical difficulties 

 are overcome by shaking the soil with 2 % agar solution, and pouring 

 the suspension quickly into glass tubes so as to cast it into sticks. 

 These are then placed upright in beakers containing the various acids : 

 diffusion of acid into the stick, and of dissolved P2O5 out of it, rapidly 

 takes place, and after some ten or more changes the diff usate is practically 

 free from phosphate. Under these circumstances the amounts of 

 phosphorus compound dissolved out from the soil are substantially 



