NATURAL SELECTION 99 



possessed by their progenitor means that the character 

 has died out of the whole breed. 



Let us, for a moment, consider not a breed, but a 

 species in the free life of its native haunts. There 

 is ample evidence that in the derived species generic 

 characters tend to die out very slowly, when once 

 they have been overlaid by the specific modifications 

 that have taken their place. The inherited tendency 

 from the old form shows itself sufficiently potent to 

 break forth through the overlying specific characters, 

 so that characters belonging to the genus that have 

 ceased generally to appear in the derived species are 

 sporadically, now in this, now in that, found in 

 individuals. 



But while this tendency to reappear asserts itself, 

 and continues, no generation passes without several 

 cases of their reappearance in individuals. As all 

 the individuals of all the derived species possess this 

 tendency by reason of their inheritance from the 

 genus, the generic characters that are liable to re- 

 appear may, in a manner, be said to belong to the 

 variability of the species, though when they reappear 

 they cannot properly be called individual variations. 

 But this tendency of generic characters, which in the 

 lapse of time have become faint and weak as compared 

 with their vigour in the old form, to reappear con- 

 tinues to be ever on the wane and to become weaker 

 as the ages pass, and the day shall surely come when 

 it will cease altogether. This day has assuredly 

 come for that species which has lost for all its indi- 



