40 BOARD OF AGRICULTURE. 



The degree of reliance upon experiments in agriculture, 

 however carefully they may be conducted, cannot be exactly 

 determined, on account of the peculiar nature of the experi- 

 ments themselves, and the impossibility of arriving at a full 

 knowledge of all the conditions which modify or influence 

 them. In this respect, agriculture difiers from most of the 

 exact sciences, since every experiment made is subject to many 

 sources of error, which the chemist in the laboratorv, or the 

 philosopher with his nicely adjusted apparatus, may guard 

 against, but which are entirely beyond the control of the experi- 

 menter on the field. And hence, while these sources of 

 error exist in all agricultural experiments, and while no human 

 knowledge can entirely guard against them, the results of ex- 

 periments tried, so far as known, under similar conditions, may 

 be entirely opposite, and yet both may be correct, — some con 

 ditions and influences being hidden and wholly inexplicable. 



This uncertainty and apparent contradiction attending 

 experiments in farming has greatly discouraged many who 

 entered unon them with enthusiasm, and who find that few 

 farmers agree upon questions of vital importance to them. 



In experiments with manures, for instance, no two results 

 may wholly agree ; but to conclude that they were not, on this 

 account, of any value, would be premature and unjust. The 

 conditions under which they were made may not be wholly 

 understood, the soil or its preparation may differ, or the atmos- 

 pheric influences, the climate or the exposure, may come in to 

 modify or vitiate the result, or the previous cultivation and 

 manuring of the soil may lead to a conclusion entirely unex- 

 pected. It is obviously necessary that all these conditions 

 should be understood, and that the utmost care should be taken 

 to learn and comprehend all others which may even remotely 

 influence the result. 



Many isolated experiments, made by individuals, are reported 

 without sufiicient detail to make them of any value in a scien- 

 tific point of view. Moreover, the experiments most frequently 

 made by the farmer are designed to show the eflfect of com- 

 pound si\bstances ; and these experiments may be needed and 

 answer a good purpose, as far as they go ; but they throw no 

 light upon the exact influence of each constituent of which the 

 compost is made up. What is now wanted is to go behind and 



