THE ROTHAMSTED EXPERIMENTS 383 



The nitrate nitrogen practically paid for itself as an average of 

 the first 30 years, but left a deficit of about $i a year for the 

 subsequent 25-year period, of which, moreover, the last 15 years 

 show an annual loss of $1.49. 



Phosphorus added to nitrate has paid for itself and 60 per cent 

 net profit as an average of the 55 years, and the effect of phosphorus 

 is apparently increasing where applied in this connection, which 

 practically amounts to using it in addition to both nitrogen and 

 potassium, assuming that the sodium has power to liberate potas- 

 sium from the soil. If the nitrogen were secured from the air by 

 clover, and if the potassium were liberated from the soil also by 

 clover, plowed under directly or in manure, it is easy to see that 

 applied phosphorus would be still more profitable, especially if 

 the 29 pounds were applied in raw natural phosphate at a cost of 

 87 cents instead of in acid phosphate costing $3.48. 



It should be remembered always that computations based upon 

 increases compared with the yields from unfertilized land may 

 indicate profits that would not be wholly realized if the total yield 

 of the unfertilized land is not sufficient to pay for its own cost. 

 In other words, if it costs more than the value of 14.8 bushels of 

 barley to secure that yield, then the financial deficit from the un- 

 fertilized land must also be overcome before any profit can be had 

 from the use of fertilizers. 



Furthermore, in planning systems of permanent agriculture, we 

 must also consider whether the apparent increasing gains are due 

 solely to improvement resulting from soil treatment, or in part to 

 the general depletion of the unfertilized land. Probably nothing 

 is more difficult for the average landowner to realize than that 

 what appears to be profit is in part at least taken from his own 

 capital. This is very clearly illustrated in the Hoos barley experi- 

 ments. Thus, with nitrogen on plot Ai, during the 15 years (1892 

 to 1906), there appears to be an average increase in yield of nearly 

 8 bushels per acre above the unfertilized yield; but, by referring 

 to the average for the first 10 years (1852 to 1861), it will be seen 

 that the unfertilized yield has decreased by more than 12 bushels. 

 On this basis, as an average of the last 25 years, the apparent in- 

 crease from nitrogen is wholly represented in the decrease in pro- 

 ductive power, and consequently in the decrease in value, of the 

 unfertilized land. 



