296 BOARD OF AGRICULTURE. 



Before the sod was broken for the plots in 1885, the field hiid Iwrne 

 three heavy crops of grass, previous to which the land received a 

 liberal application of manure and unleached ashes. 



In ISB.') the i)lots produced a crop of barley with no manure, the 

 treatment of the plots being uniform tlirougiiout. 



The plots extend east and west, and are arranged in two tiers 

 which run north and soutli, and which are separated ))y a narrow road. 

 The corners of the plots are marked by seasoned maple stakes, which 

 are driven so deep that they will remain permanently. 



It is intended to continue these experiments on the same plots 

 for a series of years. It is onl}' by testing the effect of continuing 

 a method of manuring for a long time that we are able to arrive at 

 conclusions which can be safely followed in farm practice. 



It is proposed, therefore, to devote these plots to a rotation of 

 crops, making an application of fertilizers at stated perio<ls, and 

 keeping a careful record of the crops produced and such other facts 

 as seem to be of interest. 



It is well known that the field experiments of this sort may in- 

 volve errors of considerable magnitude, the main source of error 

 being lack of uniforrait}' in the productiveness of the various plots. 



In these experiments an effort will be made to eliminate errors 

 due to this cause, by (1) having three plots in different parts of the 

 fields treated in the same way, and by (2) continuing the experi- 

 ments for a long time. 



If the average results thus obtained show no marked difference 

 between the effect of two methods of manuring after ten years, it 

 can be safely concluded that so far as immediate profit is concerned 

 it will make verv little difference which method a farmer chooses. 

 It should be recognized, therefore, that the results obtained from 

 the work of any one year have a value somewhat in proporiion to 

 the number of ^ears the experiments have been continued, conse- 

 quently the yields of the plots for 18GG are given only as a report 

 of progress. 



Each tier of eighteen plots is divided into three sets of six plots 

 each. In each set there is one plot which receives no manure, and 

 the other five plots all receive fertilizers of some kind, which are 

 not alike in all respects on any two plots in the same set. Each 

 plot of one set, however, is treated exactly like a corresponding 

 plot in both the other sets, so that each method of manuring is 

 carried out on three plots. The average yield of the three plots 



