THE FACTOES OF EVOLUTION 175 



of life remain the same, merely because in some individuals of 

 the species there has appeared a new favourable characteristic. 

 It seems clear that as long as useful variations appear as isolated 

 phenomena they undoubtedly disappear again at a later stage, 

 and are in consequence without importance to natural selection. 



Wagner's theory of migration, or the theory of the formation 

 of species by geographical isolation, attempts in a certain sense to 

 master these difficulties. The most important condition of the 

 phenomenon that isolated variations are not again lost by 

 repeated crossing with individuals which do not possess them 

 is found in the isolation of the relevant animals from the rest 

 of the species. If, therefore, some individuals of a species are 

 separated from the rest by emigration or geological causes and 

 placed among new conditions of life, this fact may give rise to 

 a new species. An interesting instance is supplied by the rabbits 

 which were liberated in the beginning of the sixteenth century 

 by Portuguese sailors on a small island near Madeira, Porto Santo. 

 They soon became acclimatized and have assumed a number of 

 quite distinct characteristics : they have grown much smaller, are 

 distinctly ' vicious,' have a reddish fur, and are indeed so widely 

 differentiated from their European parents that it is impossible 

 to pair them. 



It is easily conceived that changed external influences which 

 affect in like manner all individuals of one species produce 

 simultaneously in a large number of animals corresponding 

 useful variations. If a large number of variations exist, 

 tending in the same direction, natural selection would, in fact, 

 be able to accelerate the change. Only with this supposition is 

 it, in my opinion, intelligible that in a comparatively short time 

 there can be formed so many adaptations and new characters 

 that natural selection is able to act at all. Chance, of course, 

 becomes thereby limited, and part of the formative power is trans- 

 ferred to the organism. It is the admission that the organism 

 is capable of suitably reacting from inner causes to external influ- 

 ences, or, as Pauly expresses it, ' the awakened need itself produces 

 the means for its satisfaction.' Selection in that case would only 

 occupy the position of a secondary factor in species-formation. 

 The mechanistic school may call this mysticism, but it is far less 

 mystic than the attempt to ascribe every rational performance to 



