118 



In the same way we would hâve found another skew 

 curve, had we taken the surfaces of the berries as the 

 object of our measurements, or if, after arranging the 

 berries in the order of their magnitude, we had determined, 

 for every one, the number of them contained in a fixed 

 weight. We thus begin to reaUze that skew curves, far 

 from being the exception, must be the rule in nature. 



9. Reconciliation of this resuit with theory» The 

 resuit hère found seems in direct contradiction with the 

 resuit of the theory, which, as we saw before, demon- 

 strates that the efFect of any causes whatever, satisfying 

 the conditions a, b, c, will be the production of a nor- 

 mal curve. 



For it seems évident at first sight that the causes which 

 produce the variable growth of the diameters of certain 

 berries, being identically the same as those determining 

 the variable growth of the volumes, if the former satisfy 

 the conditions a, b, c, the latter must do so of necessity. 



The conclusion, évident as it might seem to be, is false. 



If the effect of the sunshine on May 2nd niay be con- 

 sidered to be absolutely independent of the effect of the 

 rain on May ist for the diameters, then the same effects 

 are not independent for the volumes. 



The reason lies in the fact that: 



any causes whatever, the effect of which is différent 

 (A) l for différent sized individuals, cannot be considered 

 as independent. 



We see at once the truth of this in our example. For 

 the effect of the rain of May Ist has been to make the 

 several berries unequal. Therefore, if the effect of the 

 sunshine is différent for individuals of différent size, this 

 effect on May 2nd will be another than it would hâve 

 been had the rain of May ist not existed and had, in con- 

 séquence thereof, the berries not been unequal. 



