1882.] farmers' convention. 51 



expended upon them has been in a certain sense wasted, but the 

 very knowledge which shows us how we might have done better, 

 comes through a long series of fruitless trials. There is hence a 

 negative value, which perhaps in the end has been beneficial. In 

 some few cases, especially in those which partake of the nature of 

 an investigation, much good has been derived, but yet on the 

 whole, the results have served but little purpose as applied by the 

 individual to an actual problem of farming. "Where good has 

 come, it has usually come in the way of verification, rather than 

 of discovery. The time is now a fitting one to make use of the 

 mistakes of the past in order to discover lessons for the future. 



Mr. J. J. Webb, of Hamden, in the chair. 



The Chairman. I hope gentlemen will be free in asking 

 questions. This is a very important subject, and one upon 

 which a great deal of information can be obtained by asking 

 questions. As there seem to be no questions from the farm- 

 ers, Prof. Miles is in the room, and I would be glad, and I 

 have no doubt you would be, to hear from him in regard to 

 scientific experimenting, as well as practical farm experi- 

 menting. 



Prof. Miles, of Houghton Farm, Orange County, N. Y. 

 I did not intend to say anything this morning upon this sub- 

 ject. The paper already presented has gone over the matter 

 and covered most of the ground. There is very little for me 

 to say. I will, however, remark, that this matter of farm 

 experiments is not only one of great importance, but the 

 experiments, as has been suggested in the paper, are exceed- 

 ingly difficult to conduct. Very few are aware of the difficul- 

 ties of conducting exact feeding or field experiments. Now, 

 in the laboratory, we can make investigations and arrive at 

 very definite results at once ; we can control all the condi- 

 tions under which we are experimenting. An experiment 

 in animal feeding, or an experiment in the field, is a very 

 different thing, for tlie simple reason that we can control but 

 very few of the conditions. We are liable to be misled by 

 the influence of conditions that we cannot control. It be- 

 comes important to eliminate the errors arising from these 

 conditions that are beyond our control in these experiments. 



