78 MASS. EXPERIMENT STATION BULLETIN 165. 



Considering the table as a whole we find that more calcium is remoA'ed 

 from the limed ends than from the mihmed. The effect of the application 

 of lime in 1913 is plainly shown by the increased amount of calcium re- 

 moved from the limed ends in 1913 over 1912 by the normal and 23^^ nor- 

 mal solutions. Further, in 1912 no increase is noted in the amount of 

 calcium removed from the limed plots with an increase in the concentra- 

 tion of ammonium-sulfate solution used. In 1913, however, an increase 

 is noticed with each increase in the concentration of the ammonium- 

 sulfate solution. With the imlimed ends of the plots no such increase is 

 noticed either year, except between the one-tenth normal and normal 

 concentrations. 



Studying the variations between the different plots we find the chief 

 difference to be the smaller amount of calciimi removed from the two sul- 

 fate plots 6 and 8, both Hmed and unhmed ends. 



In the fight of results already reported by other investigators as to the 

 action of ammonium sulfate on calcium in the soils, the foregoing results 

 indicate that ammonia replaces calcium in the soil, the amount varying 

 with the concentration of the ammonium sulfate used and the amount of 

 available calcium in the soil. The smaller amount of calcium removed 

 from the two ammonimn-sulfate plots indicates that the previous fertiliza- 

 tion with anmionium sulfate has depleted the soil of this element. That 

 sodium nitrate has had a protective action on the calcimn in the soil is 

 shown by the larger amount of calcium removed from the unlimed end of 

 the nitrate plot 1 than from the no-nitrogen plot 7. The presence of the 

 soda has prevented the soil from becoming poor in calcium. When larger 

 quantities of fime are present, and the concentration of the ammonimn- 

 suKate solution is high, the protective action largely disappears, as is 

 shown by the amount of calcium removed from the hmed end of plot 1 in 

 comparison with plot 7. 



Sodium and Potassium removed. 

 Owing to the large mass of ammonium suKate and the small amount of 

 sodium and potassium in solution it was difficult to make exact determina- 

 tions of these elements. Several methods were studied, but the method 

 of weighing the sodium and potassium as the double chlorides, and then 

 detennining the potassiiun as the chlorplatinate was finally adopted. The 

 results obtained are as accurate as could be expected under the circum- 

 stances. 



