186 



The Rothamsted Grass Experiments. 



half rather less than 30cwt. ; whereas over the whole period it 

 was rather more than 32cwt. Hence there was a reduction in 

 the total produce during the later years. But, in order to 

 estimate the effects of the manure, it is necessary to deduct 

 something for the annual yield of the soil and seasons ; and, 

 as plot 18 had, like plot 3, been unmanured from the com- 



TABLE XLVL AVERAGE PER ACRE PER ANNUM, BY A MIXTURE 

 SUPPLYING THE ASH CONSTITUENTS, AND THE NlTROGEN, OP 1 

 TON OF HAY; PLOT 18. 



Average per Acre per Annum. 



mencement of the trials in 1856, it will be fair to deduct 

 from the total amount that obtained on the unmanured 

 plot. The differences, given in the right-hand column, show 

 an average annual increase of produce, due to the manure, of 

 not quite 12Jcwt. during the first period, about 13|cwt. 

 during the second, and not quite 13cwt. over the whole 



