FiEivD Experiments with Fertii^izers. 143 



The man who follows this second way, realizes that it may not be 

 necessary to supply every crop on every soil with all three of the 

 plant- foods mentioned above. He knows that different crops, 

 while needing all these foods do not need the same quantities of 

 all of them ; he may suppose that this soil or that soil on 

 his farm will yield an ample supply of one or two of these foods 

 for as big a crop of some one kind as he wants to get, and he 

 realizes that it may be profitable for him to find out what the real 

 state of the case is. 



The only way to get this information is the way by which 

 pretty much all that we know about agriculture has been learned — 

 by experiment. The crops that it is most profitable for him to raise 

 must be fed, on the soil upon which he wants to raise them, with 

 these three plant-foods, one by one, and with mixtures of them in 

 the four combinations only that are possible. This is no easy 

 road to learning ; but for an intelligent farmer who has any dis- 

 position at all to study his business it will be pleasant work, and 

 will put some variety into his otherwise too monotonous occupa- 

 tion. 



I propose, then, to give full and simple directions for making 

 these experiments in as reliable a manner as possible, and with 

 the least possible expense and labor. 



The selection of the field. — The field should be as uniform in 

 character throughout as possible, and should have borne the same 

 crop all over, at least in the preceding year, and should have been 

 manured alike all over for that crop. If the field has been pretty 

 well exhausted by a three or four-year rotation, more decisive re- 

 sults will probably be obtained than if it is in good condition. 

 If not level or nearly so the slope should be as uniform as prac- 

 ticable. The underdrainage should be at least fairly efiicient and 

 uniform. The field upon which it may be most desired to make 

 the experiment may not meet all these requirements ; but it is 

 probable that a fair approach to it can be made in most parts of 

 this state. 



The size^ shape and arrangement of the plats. — The size should 

 depend on the uniformity of the soil ; for a very large field, and 

 not fairly alike in the character of the soil and previous manuring 



