246 Heredity. 



by the addition of newer variations above them. It can 

 also be shown that from time to time the peculiarities 

 at the other end of the series, the oldest hereditary fea- 

 tures, are crow^ded out of the life of the organism, and 

 dropped, so that an animal which is high in the scale of 

 evolution does not repeat, in its own development, all of 

 the early steps through which its most remote ancestors 

 have passed. The series of hereditary characteristics, 

 thus growing at one end and fading away at the other, 

 gradually raises the organism to new and higher stages 

 of specialization, and its evolution by variation and he- 

 redity may be compared with the growth of a glacier. 



The slight individual differences are represented by 

 the new layers of snow added by the storms to the de- 

 posit which fills the valley in which the glacier arises. 

 The snows which are soon blown away are those varia- 

 tions which, being of no use, soon disappear; while the 

 snow which remains in the valle}^, and is gradually con- 

 verted into ice, represents those individual differences 

 which are seized upon by natural selection, and gradu- 

 ally rendered hereditary and constant. The long stream 

 of ice stretching down to lower regions, and made up of 

 the snows of thousands of Avinters, receiving new addi- 

 tions at its upper end, and at the same time melting 

 away at its lower, is no bad representation of the long 

 series of hereditary features, once variations, which form 

 so large a part of every organism. If the glacier were 

 not in motion, but stationary, so that the melting of 

 the oldest portion and the additions to its upper end 

 should gradually carry the body of ice up to higher and 

 higher levels, we should have a very perfect pMiallel to 

 the evolution of an organism by variation and heredity. 



The steps in this progress are embodied in a long se- 

 ries of individuals, each of which is, either immediately 



