CH. VIl] 



Comparative Values 



139 



equal to nitrate of soda or sulphate of ammonia, while 

 nitrate of lime was rather better, but owing to its hygro- 

 scopic nature it was less easy to handle. On an average 

 of all British experiments if the nitrogen of nitrate of 

 soda is valued at 100 that in sulphate of ammonia would 

 be 95 and in cyanamide 90. In experiments where each 

 substance is tested on one plot only, the results can only 

 be relied upon to within 10 per cent, in any one season, 

 or some 5 per cent, over several seasons. For finer work 

 it is necessary to repeat the plots 4 or 5 times in the 

 same field each year, and to ascertain from the results 

 exactly what is the error of experiment. The method of 

 doing this is described by Wood and Stratton in the 

 Journ. Agric. Science, vol. iii. p. 107, a paper which the 

 student should read. 



Table V. Effect of various 7iitrogenous manures on 

 different crops. Rothamsted: yield per acre 1909-1914 



* Twelve tons of dung per acre was appUed to all the Mangold plots. 



