78 



The Rothamsted Wheat Experiments. 



After making allowance for the seasonal variations in 

 each of these periods, it appears that, in fairly average 

 seasons, the mean produce of the experimental plots fairly 

 represents the average produce ; that in seasons of unusual 

 abundance the experimental plots indicate too high a figure, 

 and that in seasons of great deficiency they give too low a 

 figure. Still, it is, upon the whole, concluded that there is 

 no better basis for estimating the annual yield of the country 

 each year than that of the average produce of the same 

 selected plots as heretofore relied upon ; but that, as hereto- 

 fore, some judgment must be exercised each year, according 

 to the characters of the season, in deciding whether to 

 adopt the actual figure indicated by the experimental plots, 

 or in which direction, and in what degree, it should be 

 modified. 



TABLE XVII. SHOWING THE DEVIATION OVER EACH SEPARATE PERIOD 



FROM THE ADOPTED AVERAGE OF THE WHOLE PERIOD TAKEN AS 100. 



In illustration of the difficulty of this inquiry, it will be 

 appropriate to notice at this stage the nature of the data 

 upon which the investigators had to rely for information 

 upon the following points : 1. The area tinder wheat. 

 2. The average yield of wheat per acre. 3. The aggregate 

 home produce, and the amount of it available for consumption. 

 4. The imports. 5. The population. 6. The average con- 

 sumption of wheat per head of the population per annum. 



1. From 1852 to 1865 inclusive, it was necessary to rely on 

 estimates alone in fixing the area under wheat in England 



