CONTRADICTORY RESULTS TO BE REJECTED. 77 



experiments to bring out the exact influence of each new cause 

 of diversity. 



I have before me numerous sets of experimental results, care- 

 fully tabulated and published by the experimenters. Each 

 experiment, in general, differs from the others, and the entire 

 differences among the field-results are usually ascribed to the 

 differences among the applications to the variation in quantity 

 or in quality of the substances laid on the land. 



But it is forgotten, not only that two repetitions at least of 

 each experiment ought to be made, but that the action of sub- 

 stances, when mixed or applied together, is different from their 

 action when applied singly. They exalt or lessen the action of 

 one another, and thus there may be no real or direct relation 

 between the increase of crop and the quantity of any new 

 substance contained in the mixture laid on. Thus, common 

 salt produces a certain action upon the potato and oat crops 

 under certain circumstances, and nitrate of soda produces a 

 certain different action ; but a mixture of the two, one-half of 

 each, produces a result in general much more favourable than 

 the average of their actions when applied singly. Thus it is 

 not to the nitrate of soda added that the increase of crop 

 is to be ascribed above that which salt alone gives, but to the 

 special action of the two taken together as a mixed appli- 

 cation. 



It is thus clear that, to arrive at theoretical truth in regard 

 to the precise virtue we should ascribe to any two fertilising 

 substances, we must have at least two experiments with each 

 taken singly, and two with the mixture which contains them. 

 From the differences thus obtained, we may be able to draw 

 conclusions as to the influence which each exercises in the given 

 circumstances, either when applied alone or when mixed with 

 the other substance. 



4. When the results obtained in the same circumstances are 

 obviously contradictory among themselves like some of those 

 from the island of Bute above quoted. A careful examination 

 of published tables of experiments will show many such con- 

 tradictions. 



And yet contradictory results, obtained at different times, by 



