The natural resouc€s]of the soil ; its ability to provide, 

 potash in one case , phosphoric acid in another, etc., constitute 

 a factor of uncertainty in feeding plants which is not met with 

 in animal nutrition, and it is this uncertainty which makes it 

 necessary to make local experiments with fertilizers ; I do not 

 mean that results obtained in one place are of no value in an- 

 other, for this is not the case, there are certain classes of soil 

 which are so similar in their origin and composition that a very 

 certain prediction may be made concerning their requirements, 

 and it is true that the great bulk of the hill lands, or "drift soils" 

 of New Hampshire respond similarly to fertilizers of practically 

 the same composition ; but I do mean that there are ex- 

 ceptions and local tests can point them out, hence the necessity 

 for experiments on soils having unlike characteristics. 



Value of crops from plot experiments. 



The results tabulated on page 14 I have rearranged as 

 above : commencing with the greatest value of crops, for 

 each year and following down to the lowest, also for the 

 total of three years, the last column shows the position as re- 

 gards the best yield, the omission of oats and hay for 12th, 

 13th, 14th, 15th, place are occasioned by the fact that no fertili* 



16 



