332 The Theory of Gejietic Modes 



their correlations, and in general morphological ques- 

 tions as such. For all these questions deal with the 

 development of new genetic modes. It is, indeed, a valid 

 criticism of much of the work done by the new methods 

 that it assumes that the variations which are tabulated 

 do represent the same modes, or stages of development or 

 evolution. If we accept the view that all characters of 

 the organism are in part epigenetic, — are the outcome 

 of hereditary impulse plus the bionomic conditions under 

 which this impulse develops, — then variations themselves 

 differ at each stage of the individual's development, 

 differ with age, growth, etc. In gathering, tabulating, 

 and treating variations, therefore, only those can be put 

 together which belong to the same stage ; and this is 

 most difficult to determine. Suppose we undertake, for 

 example, to measure the variations in the length of nose 

 of a species, it will not do to take the noses of indi- 

 viduals at different ages, which represent different stages 

 of maturity ; nor will it do, on the other hand, to take 

 noses from different environments, where different reactions 

 have occurred and different amounts of growth in this 

 direction or that have been possible. So as to the deduc- 

 tions which may be made from such measurements for the 

 theory of evolution, there may be very different formulas 

 of variation at different evolutionary stages, grades, or 

 modes. Influences which are very powerful in effecting 

 variation in simple organisms, may be largely ineffective 

 at later periods.^ 



^ I am not sufficiently versed in their results to judge whether these con- 

 ditions are sufficiently allowed for; but it seems to be a legitimate demand to 

 make of the statistician of biological measurements, that all his cases be at the 

 same stage of development, and that they all occur in common environing 

 conditions. Since writing this passage, I have come upon the article by 



