EXPERIMENTAL STATIONS, 205 



seasons previously under the same conditions throughout, 

 having mapped it out like a chess-board into small portions, 

 whose produce should be determined and used as a basis for 

 qualifying the results of the experiments. This is a process 

 requiring much time and trouble, and is not likely to be waited 

 for patiently by those who are anxious to derive early informa- 

 tion regarding any experimental investigation. 



Owing to the variations of soil occurring in one plot, compara- 

 tively little importance can be attached to the results of cropping 

 on that plot (No. 36), which is divided into portions each the 

 YJ-gth part of an acre. Indeed, that particular part of the 

 Pumpherston station is so unequal in its soil that it must be 

 rejected altogether. This, which seemed at first a great misfor- 

 tune, will in the end be turned to considerable advantage. An 

 attempt will be made to make a homogeneous soil hj removing 

 the soil from the entire plot, and after mixing it thoroughly 

 together and dressing the subsoil, spreading it over the plot 

 again in a uniform manner. This will be done during the sum- 

 mer, when the land is dry and friable, and able to Ije thoroughly 

 riddled together. In order to attain reliable results with plots 

 the 1^2^^^ of an acre, the very smallest disturbing causes must 

 be guarded against. Little inequalities of soil are fatal to them. 

 They are a kind of large open-air pot experiments, and if their 

 soil were as uniform as that employed for pot experiments, the 

 results obtained would be thoroughly reliable. 



Enough has been said regarding the inaccuracies incident to 

 field experiments arising from diversity of soil among the various 

 plots. Although in Pumpherston station these are considerable, 

 they are yet capable of being estimated in great measure, and are 

 not such as to render the results derived from the cropping un- 

 trustworthy. They will diminish year by year under a uniform 

 system of cultivation and manuring, and the chief object con- 

 templated in bringing them so prominently into notice is to 

 show that they are receiving a due amount of attention, and to 

 impress upon experimenters the necessity of exercising the 

 utmost caution in interpreting the results of field experiments, 

 and especially those of a first year, or indeed of a first rotation. 

 Much harm has been done to the progress of scientific agriculture 

 in the past by experunenters being too sanguine, and by those 

 interested in the experunents being too impatient. 



It will be seen from the chart accompanying the diagrams, that 

 the system of cultivation is a four-years' rotation of turnips, 

 barley, grass, and oats. The winter which has just passed has 

 been one of almost unexampled severit}^, and has done great 

 damage to the turnip crop. It is noteworthy, however, that the 

 injury done to the crops on both experimental stations, which 

 were manured only with artificial manures, though it is consider- 



