62 BOARD OF AGRICULTURE. [Jan., 



One of the most important methods of eliminating such errors 

 is a repetition of the experiments under the same conditions 

 for a length of time; then a varying of the conditions, chang- 

 ing only a single condition. A great deal has been said in 

 regard to animal foods, relative values, etc., and it would be 

 apparently a very simple matter to ascertain the relative 

 value of foods by making a test with the animals themselves ; 

 but when we get to the work itself we find, that no two ani- 

 mals have precisely the same digestive apparatus ; we find 

 that the condition of the animal as to age, as to the amount 

 of flesh it carries, has a very great influence. Peculiarities 

 of breed, and the manner in which the animal has previously 

 been treated, all have their influence; so that it becomes ex- 

 ceedingly difficult to make an exact experiment that will 

 result in the advancement of our knowledge of agriculture. 

 The experiments should be well planned, and then carried 

 out for a long series of years. As Dr. Sturtevant has just 

 remarked in regard to the experiments at Rothamstead, they 

 have been conducted for a very long time, and there are per- 

 haps no others in the world that have been conducted upon 

 the same plan and for the same length of time; and the re- 

 sults which are of the most value are just now being obtained. 

 I place very little dependence upon experiments conducted 

 only from year to year in different localities. An experi- 

 ment for a year upon one piece of land, and another year upon 

 another piece of land, in another locality, will throw but very 

 little light upon the laws or principles that are to guide us in 

 farm practice. The peculiarities of season may have a very 

 great influence upon the results, so that a field experiment 

 must be conducted long enough to eliminate the errors aris- 

 ing from the variiitions in the seasons themselves. Mr. 

 Lawes, in looking over his experiments for forty years, is 

 enabled to pick out certain bad seasons and certain good 

 ones, put them together, and see what the results are, and 

 in that way he is arriving at conclusions which are very dif- 

 ferent from those that he would have drawn from the experi- 

 ments thirty years ago. As he remarked to me, when he 

 had experimented ten years, he thought he could generalize 



