( xc ) 



ing investigation by the only legitimate method, viz., by 

 experiment and observation. If, owing to the community of 

 the physiological activities of different species not very closely 

 related from the systematic point of view, a physiological 

 variability gives rise to the appearance of the same type of 

 variation (Darwin's " parallel or analogous variation "), such, 

 for example, as melanism, or a darker suffusion of pigment, 

 it is quite intelligible that such a type of variation, if of use, 

 would be seized upon and perpetuated by natural selection. 

 Thus, in accordance with Lord Walsingham's ingenious 

 theory of the dark forms of Lepidoptera from high latitudes, 

 an advantage is conferred upon such forms by their superior 

 absorbing power for solar radiation. Natural selection has 

 here again utilised physiological variability ; it happens that 

 it is a variability of a type common to widely divergent forms, 

 by virtue of its being connected with internal processes which 

 ai;fe common to such divergent forms. The result is that 

 different species of butterflies and moths are alike affected, 

 not by the direct action of low temperature, but by the selec- 

 tion of darker forms — by the survival of certain special types 

 of physiological activity. But if divergent forms can be made 

 by such means to display a similarity in external useful 

 characters, it is conceivable that a similarity of external non- 

 significant characters may also result in groups of different 

 species by the selection of the physiologically fit — the adapta- 

 tion in such cases having reference to latent characters. In 

 this light " the direct action of the environment " disappears 

 as a distinct factor of organic development. 



The bearings of the suggestion which I have attempted to 

 put into a somewhat more definite form in this address will, 

 I hope, be perceived by practical workers. It will be realised 

 perhaps more clearly that there is a very human element in 

 the term " specific character," and that the differential 

 characters wdiich the systematist has learnt to detect with 

 such consummate skill bear no proportional relationship to the 

 actual differences between the forms described. It will also 

 appear that j;)/(?/si'caZ external conditions, as factors in the 



