AGRICULTURAL EXPERIMENTS. 297 



rious. It is as much so wi£h animals as with men. I suppose if 

 the giant in the old story had tried to fatten his men in solitary 

 confinement, he might not have succeeded so well. . I remember 

 separating two horses that had been together a long time, and 

 continuing one on the same feed ; but he fell off all the time until 

 his mate returned. I do not know how you will feed them so as 

 to get accurate results. Sometimes, if you put animals together 

 who do not agree, as will often be the case, there will be little 

 parties of them that will naturally run together. If you could get 

 those parties together and feed them, that would be better than if 

 they were mixed with those for whom they have no affinity, as 

 you may say. It would be difficult to manage this matter; I 

 would rather regard the social principle, for we know it is a gen- 

 eral law, than to disregard it. 



I want to call attention to one thing— the question as to experi- 

 ments in feeding. Experiments made in Maine and in Louisiana 

 are almost worthless, unless you take corn from Maine to Louisi- 

 ana, and from Louisiana to Maine, so as to feed them on the same 

 corn ; for the corn in Maine is worth much more a pound than that 

 raised in the South. They can not raise the corn we do, and we 

 can not raise theirs'; and the comparison of experiments made in 

 such localities would be of little value, although not so much as 

 those in Europe. What we want our colleges to do, is to investigate 

 the principles applicable to its own locality. You may prove, for 

 instance, in Michigan, that in your sandy soil a certain kind of 

 manure is the best for corn and wheat, and we try it off in Mis- 

 souri, and it would be a failure. We know it beforehand. We 

 place your barn-yard manure on the ground for wheat, and it will 

 injure it ; but on the sandy soil in Michigan it will be of great 

 value ; on some other soils animal manure will lessen the crop 

 rather thair increase it. There are certain things which you know 

 and r/o don't know. It seems to me we can do more by having a 

 certain class of colleges whose relations are very hear to each 

 other as to soil, take up one series of experiments, and another 

 class of colleges, having another kind of soil and climate from 

 those already related, take another class of experiments — that is, 

 have each take up a certain class of experiments which would 

 come within range of its climate, soil and stock. I do not see, 

 especially in the feeding of animals, *that any experiments in 

 Maine would be of much value to us in Missouri — the climate is 

 so different. I do not see, on the other hand, that their feeding 



