282 BOARD OF AGRICULTURE. 



selves, and also with the crops on the same plats in preceding 

 years. 



The question is often asked, what siz^d plats it is desirablegto 

 make in field experiments ? That is a difficult question. to answer. 

 On some accounts it would be desirable to have the plats very 

 small ; on other accounts it would be desirable to have them of 

 considerable size. Dr. Anderson states that he thinks the. smaller 

 plats are desirable ; he would not have plats exceed 1-100 of an 

 acre in area. Mr. J. B. Lawes, who associated with Mr. Gilbert, 

 in experiments conducted at Rothamstead, writes me he has no 

 confidence in the experiments made on plats of less than 1-20 of an 

 acre. These are two high authorities. Dr. Anderson is an able 

 chemist and a very successful experimenter. Lawes and Gilbert 

 are perhaps the best agricultural experimenters we have had. 

 Their experiments are not fully accurate, but, taken on the whole, 

 they are the best that have been made. ~My own experience is, 

 that a plat of from 1-100 to 1-20 of an acre" should be the limits in 

 size. The difficulties of the very small plats are, you are liable, 

 when you are manuring, to have it extend its influence to the 

 adjoining plats, and the roots of the plant will extend some dis- 

 tance into the soil, so that you are liable to be misled. When 

 you come to weigh the produce of the plats, a very slight error 

 will amount to considerable ; when you make the calculation to 

 ascertain the yield per acre, the error in a single plat would be 

 multiplied by 100 in getting the standard for an acre. In a small 

 plat it requires very great accuracy in weighing, great accuracy 

 in the division of the plat, and great care in the management of 

 the experiment throughout. To obviate this objection, that the' 

 roots run from one to another, I think it wonld be well to have 

 spaces between the plats — a space of several feet left between 

 those plats that are to be used for experimental purposes, and 

 these spaces should be kept clean and free from weeds. But here 

 comes another difficulty. Suppose you have plats laid out, and 

 spaces of three feet between them, you must cultivate the spaces, 

 and need to expend the same amount of cultivation on each one of 

 the spaces as on the other, so that the crop adjoining may be 

 equally influenced. Again, the spaces should be weeded carefully 

 at the same time, cultivated at the same time, and managed as 

 carefully as the plats themselves, or you will vary the result in 

 the plat. 



The objection to the large plats I have already mentioned, but 



