246 SYSTEMS OF PERMANENT AGRICULTURE 



of land, the experiment was conducted in triplicate; and even 

 each of the triplicate tests was in a sense duplicated in that a double 

 comparison is made between the two forms of phosphorus, the test 

 with yard manure being entirely independent of the test with stall 



manure. 



For convenience the average yield of each crop is given by plots 

 for each series of plots separately. Thus, as an average of four corn 

 crops in Series A, plot 15 with yard manure alone produced 41.5 

 bushels, and plot 2 with yard manure and raw phosphate produced 

 54.9 bushels, showing by direct comparison a gain of 13.5 bushels 

 due to the raw phosphate. Further comparison shows average 

 gains of 2.1 bushels of wheat and .58 ton of clover hay by raw 

 phosphate and yard manure above the yields made where un- 

 treated manure was used. 



A similar comparison shows average gains of 5 bushels of corn, 

 3.9 bushels of wheat, and .37 ton of hay by raw phosphate and stall 

 manure above the yields where stall manure alone was used. 

 Acid phosphate also produced marked gains, the average gross 

 increase being somewhat greater than with the raw phosphate, 

 but the net profit being slightly less on Series A. 



Attention is called to the fact that 8 tons of manure per acre 

 have been applied every three years to all manured plots. This 

 does not do full justice to the phosphate plots, because these plots 

 have yielded as an average about one fourth more produce than 

 the plots receiving manure alone, and from this increased produce 

 about one fourth more manure can be made in regular farm prac- 

 tice. Consequently, after the first rotation, the applications of 

 manure should be larger on the phosphate plots in proportion to 

 the produce of the previous rotation; whereas, to apply equal 

 amounts of manure to plot 15 and plot 2, for example, means 

 essentially that some of the produce from plot 2 is used to make 

 part of the manure that is applied to plot 15. 



In the above comparison to determine the effect of the phos- 

 phorus used, the yields with manure alone are subtracted directly 

 from the yields with manure and phosphorus. As an average of 

 many tests, this direct method of comparison is perhaps as good as 

 any indirect method, but where a small number of tests on only a 

 few fields are to be considered, probably an indirect method of 



