1 08 Studies in the Theory of Descent. 



are always only differences of degree and not of 

 kind, as is perhaps most distinctly shown by the 

 very variable A. Prorsa (summer form), in which 

 all the occurring variations differ only by the 

 Levana marking being more or less absent, and, 

 at the same time, by approximating more or less 

 to the pure Prorsa marking ; but changes in a 

 totally different direction never occur. It is like- 

 wise further evident, as has been mentioned above, 

 that allied species and genera, and even entire 

 families (Pieridce\ are changed by similar external 

 inducing causes in the same manner or, better, in 

 the same direction. 



In accordance with these facts the law may be 

 stated, that, in butterflies at least, all the indivi- 

 duals of a species respond to the same external 

 influences by similar changes, and that, conse- 

 quently, the changes brought about by climatic 

 influences take a fixed direction, determined by 

 the physical constitution of the species. When, 

 however, new climatic forms of butterflies, in which 

 natural selection is completely excluded, and the 

 nature of the species itself definitely determines 

 the direction of the changes, nevertheless show 

 variability from the very beginning, we may 

 venture to conclude that every transformation of a 

 species generally begins with a fluctuation of its 

 characters. But when we find the primary forms 

 of butterflies always far more constant, this shows 

 that the continued crossing of the individuals of a 



