RECAPITULATION 61 



and regression. But, if we think of some new and simple 

 structure, or better still of a single quality (say colour) of 

 this structure, we are able to see clearly their exact signifi- 

 cation. We see, then, that progression implies a divergence 

 from the ancestral type, a complete recapitulation plus an 

 addition ; whereas regression always implies a reversion towards 

 the ancestral type, an incomplete recapitulation. In other 

 words, every congenital deviation from the parent is either a 

 prolongation or an abbreviation of the life-history, or, in the 

 case of complex variations, it may be a combination of both. 



101. All the inborn characters of an individual take origin 

 in the germ-plasm whence he sprang. They arise because 

 the germ-plasm was so constituted that it tended, under fit 

 conditions of shelter and nutrition, to give rise to an organism 

 having those characters. A principal source of confusion in 

 the study of heredity arises from the fact that the germ- 

 plasm may be so constituted that it tends to give origin to 

 two characters, one of which is incompatible with the other. 

 The tendencies are not incompatible, but the characters are. 

 Only one character will then develop. The other becomes 

 " dormant " or " latent " ; but the tendency, to which it is 

 due, persists, and may become active in some subsequent 

 generation. To take an illustration : suppose a blue colour 

 has been evolved in a line of pigeons A. . . . M. Suppose 

 N. varies from M. in such a manner as regards a particular 

 feather that he not only has a tendency to develop the blue 

 colour, but a new and stronger tendency to develop a white 

 colour. Then the blue colour becomes dormant and the 

 white appears. If O. recapitulates his parent N. he will 

 reproduce the white colour in the corresponding feather. 

 Suppose in addition he varies in such a manner as to render 

 the blue colour dormant in other feathers. Suppose this 

 process is repeated in subsequent generations till Y. develops 

 wholly white feathers. Then, though Y. recapitulates all his 

 ancestors, no blue colour will appear in him; it will be 

 dormant. If, however, Z. reverts to M. or a remoter ancestor 

 he will be entirely blue without a trace of white. His 

 descendants also will have no white unless they happen to 

 vary as N. did, thus beginning the process again, or unless 

 the white has disappeared through the occurrence of a fresh 

 progressive variation which has rendered the white dormant 

 in turn. If Z. reverts to an ancestor less remote than M. he 

 will be mingled blue and white, the one colour or the other 

 predominating according to the remoteness of the ancestor. 



102. We have described no merely imaginary process ; 



