E. J. Russell and J. A. Prescott 87 



to each, and they are rapidly replaced in the thermostat and shaken 

 for the proper time. In the case of the so-called "Time = 0" experi- 

 ments the bottles were not put back, but were vigorously shaken for 

 a few seconds and allowed to stand a further few seconds for settling. 



The estimation of the phosphate. 



The liquid is rapidly filtered through asbestos packed on a Buchner 

 funnel; 250 c.c, or if necessary 500 c.c, of the filtrate are evaporated 

 to dryness, and the residue treated as already described by one of us^. 



The action of dilute acids on soils. Relation between the concentration of 

 the acid and the amount of P2O5 extracted. 



Schloesing and de Sigmund have both stated that the amount of 

 P2O5 extracted from a soil by acids of increasing concentration is constant 

 over part of the curve, but we failed to find evidence of this. Working 

 exactly as they did, viz. at laboratory temperature, we obtained the 

 results given in Table II, and plotted in Fig. 1, which certainly might 

 be made to fit a curve like theiis, but could equally fit any other. Under 

 the more precise conditions of the thermostat we obtained the smoother 

 results given in Table III, and plotted in Fig. 2. 



In all cases there was a considerable excess of acid left after 

 the experiment, though the amount that had been neutralised was 

 considerably more than corresponds with the calcium carbonate (which 

 in this case was only small) and the P2O5 extracted. 



The action is complex: aluminium and silica are both found in 

 solution and in increasing quantities as the action is prolonged. But 

 the final solution is always strongly acid, and readily dissolves calcium 

 phosphate and attacks fresh soil dissolving out more phosphate: the 

 variations in amount of PgO^ dissolved cannot therefore be attributed 

 to exhaustion of the acid. 



The amounts of PgOg extracted by sulphuric, citric, and hydrochloric 

 acids are given in Table III, and plotted in Fig. 2. They increase with 

 the concentration of the acid, but not quite proportionately, and the 

 curves are apparently not straight lines over their whole course, although 

 they usually are nearly so. There is no sign of a break anywhere, and 

 in particular no indication of the horizontal run shown in de Sigmund's 



1 Prescott, This Journal, 1914, 6, 110. Extended experience with this method has 

 demonstrated its great accuracy for small amounts of P2O5 such as occur in soil extracts. 



