YALE SCIENTIFIC SCHOOL 99 



must be trained to follow the trail of truth. It is so, too, with 

 experiment. There is no lack of agricultural experiments. 

 Unfortunately, however, it is mostly difficult, nay impossible 

 to find any good reason why they have been made, they are 

 so barren of useful results. The empiric experiments at a 

 venture without any probability to guide him. His hap- 

 hazard trials often reveal new facts, but he rarely contributes 

 largely to scientific progress, because he makes hap-hazard 

 experiments, because he does not reason. The philosopher 

 experiments with an object in view, and distinctly in view. 

 He does not indulge in small talk with nature, but puts ear- 

 nest questions to her. The course he follows in the investiga- 

 tion of a subject little known is this: He first collects and 

 collates all the facts kno'OTi with regard to it. He then seeks 

 to construct a consistent explanation of these various facts. 

 It may be that he finds it impossible to do this. Then he 

 must verify the facts, perhaps some are false or he sees them 

 from an insufficient point of view, or he must collect more of 

 them by extending his observations, it may be by experi- 

 ment. He shortly is enabled to form a h3T)othesis, to frame 

 a theory which promises to account for the facts. Yet it is 

 not a hypothesis but truth he seeks, and now he begins to 

 test his theory. Every deduction which he can draw from 

 it must prove true, else the theory is false. He therefore 

 unites the conditions which his theory indicates will produce 

 a given prevised result. If the result follow, his theory is 

 confirmed, otherwise it must be rejected and a new one formed 

 and similarly proved. Here is where experiment assumes its 

 chief dignity and value. Here, it must be suggested by reason 

 or it cannot be expected to answer any good purpose. Here, 

 if rationally devised and skilfully executed, it must reveal 

 a truth, and though the truth be negative it is not the less 

 valuable, for every new negative result limits within narrower 

 bounds the space wherein positive truth is to be sought. . . . 

 How little is to be expected from mere farm experiments 

 conducted without especial aid from science, the past abun- 



