48 The Rothamsted Wheat Experiments. 



which would be in a much drier state in the spring and 

 summer than during the autumn and winter. 



From the table it may be seen that it was not until after 

 the experiment had been carried on for eight years that the 

 spring-sown nitrate showed a superiority over the autumn- 

 sown ammonia-salts ; but during the first three years there 

 were no minerals, and there was much less nitrate, applied to 

 the nitrate plot (9a). During the next three periods of eight 

 years each, the produce by the nitrate was considerably in 

 excess ; though in six of the last eight years the ammonia- 

 salts were applied in the spring. Over the thirty -two years 

 the nitrate gave an increase over the ammonia-salts, con- 

 taining an equal quantity of nitrogen, of 3| bushels per acre 

 per annum (36 J bushels as against 32|). Another note- 

 worthy feature is, that the spring-sown nitrate supplying 

 861b. of nitrogen gives exactly the same total average yield 

 (36^ bushels) as the chiefly autumn-sown ammonia-salts 

 supplying 1291b. ; and, taking the total produce (grain and 

 straw), which is by far the best measure of the power of a 

 manure to promote growth, it is seen that the 861b. of 

 nitrogen as nitrate of soda produced 69821b., whilst the 1291b. 

 of nitrogen as ammonia-salts produced 68321b. Although 

 there is often a strong prejudice against the use of nitrate, 

 this experiment shows that, when judiciously applied, its 

 properties as a manure are superior to those of ammonia- 

 salts. When used continuously for thirty-two years, its power 

 to produce growth appears to increase rather than to diminish ; 

 as during the last sixteen years the total produce per acre 

 per annum of plot 9a in excess of the produce of plot 7 each 

 receiving the same amount of nitrogen has been 70841b. as 

 against 53131b., a difference of 17711b. in favour of the 

 nitrate. 



