FAEMERS' INSTITUTES. 201 



Let us consider the subject farther. Some of these questions can and proba- 

 bly will some time be solved, but most of them, in my opinion, never Avill be 

 solved. Some of them are not of much account. The solution of these ques- 

 tions is not of so much importance to most of our farmers as to become accus- 

 tomed to practice what they already know to be best, and to become familiar 

 with other things which are well settled and then to put them in practice. For 

 example, there is nothing better settled than that cultivated land should be 

 sown to a variety of crops in succession, yet how few there are Avho practice 

 no systematic rotation whatever. 



Most of our farmers need to see and feel the need of better cultivation of the 

 land, and to learn what is good culture ; to give more attention to selling and 

 buying to advantage. They need to keep accounts ; to learn how to drain land 

 properly and then practice it ; to fight our insect foes thoroughly and in the 

 most economical ways. In feeding animals they need to weigli and measure 

 and estimate for themselves at all times to know whether they are making 

 money or losing it. jSTothing that will ever be done can relieve the farmer from 

 thinking and experimenting for himself. 



How Discoveries are Made. 



Some things are discovered by accident or by trial Avithout much forethought, 

 but more are discovered by those who give thought to the subject. Previous 

 notions or theories are not always safe guides as to what may prove valuable. 

 According to our theories, a certain experiment may not seem worth testing, or 

 it may seem too absurd to be tested. 



TJie Difficulties in Making Accurate Experiments 



are much greater than most people suppose. Some of the favorite popular 

 experiments are the testing of different kinds of fertilizers, or different modes 

 of sowing or planting some crop, or for comparing the yield of different varie- 

 ties. The element within our control is a very small one. We do not usually 

 control the amount of moisture, nor the amount of heat and sunshine, nor the 

 sudden changes from one to the other. Besides this, no two seasons exactly 

 repeat Ccxch other. We cannot find two square rods of soil which are alike or 

 which will produce exactly alike under all circumstances. At first nothing 

 seems easier tlnin to treat four rows of corn across a field or garden with some 

 fertilizer or some mode of planting and take four other rows by the side of these 

 with which to compare them. This one experiment may seem very decisive and 

 satisfactory. The same experiment tried in another portion of the field would 

 show very different results. Repeat these experiments the next year and the 

 conclusions arrived at from one experiment the first year may all be reversed. 

 Before trying our fertilizer we ought to know whether the soil is nearly of uni- 

 form quality, or which part of the field is the best. 



Let any one take two rows or four rows across a field, treated as nearly alike 

 as can be, and weigh and measure the crops on each, and he will likely be sur- 

 prised at the variation. It is more difficult to distribute fertilizers equally than 

 many suppose, especially to distribute barnyard manure, which may not be of 

 uniform strength and value. A difference in yield of crops may by no means 

 be entirely attributed to the difference of fertilizers used. 



Before testing a fertilizer the soil should be carefully drained and tested by at 

 least two crops on previous years without manure, and the results carefully noted. 

 The season may be a very dry one or a very wet one, causing great variation. 



26 



