

EXTENT OF EXPERIMENTAL PATCHES. 39 



a form are more likely to be uniform in quality, are more easily 

 measured with accuracy, are more compact, and fit more closely 

 into each other. The plants grown upon them are more removed 

 also from external influences. 



The custom of making experiments upon two or more drills 

 or rows of a crop is objectionable and untrustworthy. Among 

 other reasons, it is so because the roots of plants stretch 

 themselves laterally to considerable distances, those of the 

 turnip, for example, to distances of five or six feet. The outer 

 rows or drills, therefore, are by no means beyond the influence 

 of the adjoining soil or crop, or of any variety, superabundance, 

 or scarcity of manure, by which the adjoining soil may be 

 distinguished from that on which the experimental crop is 

 growing. 



2. The extent of an experimental piece of ground is limited 

 chiefly by the labour required to weigh the crop produced, 

 which increases almost in equal proportion with the breadth of 

 land employed. Where comparative experiments are to be 

 made, a quarter of an acre, at least, should if possible be devoted 

 to each. By this means, unobserved or accidental differences in 

 the nature or condition of the soil, or in the quantity or quality 

 of the manure applied to the several parts, which might materi- 

 ally affect the absolute produce of a smaller portion, will be in 

 some measure eliminated, and more trustworthy comparative 

 results obtained. This quantity, of course, may at any time be 

 increased with advantage, and it may be somewhat diminished 

 when the necessities of the case, such as the whole size of the 

 field, or the extent of similar land at the experimenter's 

 disposal, demand it, or the obvious and known uniformity of the 

 soil, &c., seem to render an entire quarter of an acre unneces- 

 sary. 



3. The kind of soil, if it is uniform in quality, is of little 

 comparative consequence, generally speaking. Whether rich 

 or poor, light or heavy, it may be experimented upon, if its 

 agricultural history and its chemical composition are known. 

 The results obtained will of course be specific to the kind of 

 soil. What is called a virgin soil, or one to which no manure 

 has for a long time been applied, if within reach, should be 



