OHIO FIELD EXPERIMENTS 445 



is some evidence which indicates that more or less of the effect of 

 the potassium salt is due to indirect action rather than as plant 

 food. Thus the sodium nitrate on plot 17 produces distinctly 

 better results than the oil meal or dried blood on plots 21 and 23. 

 (About twice as much potassium was applied in the Pennsylvania 

 Experiments, but the sodium nitrate shows superiority over dried 

 blood during the second 1 2-year period, after the supply of humus 

 has probably become somewhat depleted. If the blood nitrogen 

 failed to become available with sufficient rapidity, one would 

 expect it to produce cumulative benefits like the manure, to some 

 extent, but such is not the case.) 



In the Ohio experiments the average effect of potassium has been 

 nearly the same, whether applied alone ($4.89 on plot 3), in addi- 

 tion to phosphorus ($5.66 on plot 8), or in addition to both nitrogen 

 and phosphorus ($5.70 on plot u), although nitrogen and phos- 

 phorus without potassium (plot 6) produced an increase of $24.63, 

 or 59 per cent above the unfertilized land ($41.73). More than half 

 of this increase must be credited to phosphorus alone, and less than 

 half to nitrogen after phosphorus; while, in reverse order, nitro- 

 gen gets one fourth and phosphorus three fourths of the credit. 

 Nitrogen alone produced an increase of only $6.73 (plot 5), and it 

 is questionable if this increase is not in part due to indirect action, 

 such as increasing the availability of the soil phosphorus, the effect 

 on the clover crop being as marked as on the other crops. As an 

 average the effect of nitrogen and potassium together is only 

 $7.73, leaving a net loss of $10.15; but the addition of 20 pounds 

 of phosphorus, under this most favorable condition, pays back this 

 loss and adds a net profit of $10.05, making a gross increase of 

 $22.60, or almost ten times the cost of the phosphorus, which 

 certainly establishes well the fact that, if the farmer can supply 

 the nitrogen in clover or in manure, and liberate the potassium 

 etc. by means of the decaying organic matter, there must be large 

 profit from the use of purchased phosphorus. (In fine-ground rock 

 phosphate the 20 pounds of phosphorus will cost about 60 cents, 

 at present prices.) 



Phosphorus is evidently the limiting element on all plots where 

 76 pounds of nitrogen have been supplied, practically no increase 

 being produced by the extra nitrogen on plot 12, and twice as much 



