2^6 



Journal of Agricultural Research 



Vol. III. No. 3 



is given. In each series the results are averages of duplicate plots, 

 except for the controls, which represent the averages for four plots in 

 each case, all plots being one-fortieth of an acre in area. The tests in 

 1909 and 1910 included plots receiving four different quantities of nitro- 

 gen, four of phosphoric acid, and three of potash. Dried blood, acid 

 phosphate, and muriate of potash were used as fertilizers. 



Table XII. — Results of tests with cotton at Manning, S. C, to determine the influence 

 of fertilizers on tlw oil content of the seed 



The soil used in 1911 was very poor, as shown by the large increases in 

 crop yields produced by the complete fertilizers. The addition of all 

 three elements, nitrogen, phosphorus, and potassium, combined in vary- 

 ing proportions, gave in all cases considerably heavier seeds, with a 

 smaller percentage of hulls and a higher oil content in the kernels as 

 compared with the controls. With respect to the varying quantities of 

 the three fertilizer elements, increased applications of nitrogen had no 

 appreciable effect on the weight of the seed and only a slight eft'ect on 

 the percentage of hulls, but lowered considerably the oil content of the 

 kernels. Increased applications of phosphorus and potassium did not 

 materially affect any of these characters. The tests of the two preceding 

 years gave similar results. 



Pot-culture tests were made in 191 1 with the Peking variety of soy 

 beans, using the tile cylinders, previously described, filled with the 

 Arlington clay soil. In a series in which phosphorus and potassium 

 in a fixed ratio were added in three different quantities, the yields of 

 beans were greatly increased and the weight of the seed was not 

 changed, while the oil content was increased about 20 per cent. With 



