58 THE PRINCIPLES OF HEREDITY 



96. All the organs and tissues of which an individual is 

 compounded possess the power of independent variation. 

 Even single cells may possess this power. One hand may 

 vary in a particular way, but not the other. A foot may 

 vary, but not the hand. One or more nails out of the twenty 

 may vary but the others not. All evolution in shape and 

 function depends on this power of independent variation. 

 Without it descendants might become larger or smaller than 

 ancestors, but otherwise would remain exact copies. In 

 studying heredity our attention is apt to be distracted by 

 the multitude of variations, and the consequent complexity 

 of the subject, from the main fact of development, viz., the 

 recapitulation of the parent by the child, and, consequently, 

 of the life-history of the race by the individual. In every 

 normal individual in every individual who survives the 

 variations from the parent are as but a drop in the ocean 

 compared to the likenesses. Were it otherwise no organism 

 would reproduce its kind. 



97. In order to get rid of as much as possible of the con- 

 fusion we may begin by imagining a case of extreme simplicity. 

 We may think again of a structure which began as a variation 

 in an individual, whom we name A. ; a structure which was 

 evolved by means of an uninterrupted series of favourable 

 variations in a line of descendants whom we may name 

 B., C., D. . . . L., M. Suppose M. has a child named N. 

 Then, as regards the structure we are speaking of, N. exactly 

 resembles M. ; or he does not. In the former case he has 

 recapitulated M.'s development without additions or sub- 

 tractions. In the latter case he has made additions or 

 subtractions or both : in other words he has varied. The 

 additions are termed progressive variations ; the subtractions 

 are termed regressive variations. Variations may occur during 

 any stage of the development ; that is, N. may vary from M. 

 as embryo, as foetus, or as adult. 



98. Consider the regressive variations first. Obviously a 

 regressive variation is a failure to recapitulate the whole of M.'s 

 development. In other words it is a failure to recapitulate 

 the whole of the life-history as presented by M. N. stops 

 short at a point reached by some ancestor say H. That is, 

 he reverts to H. He recapitulates the development without 

 the additions made by the descendants of H. If his descend- 

 ants recapitulate his variation, it is evident that I., J., K., L., 

 and M. disappear from the series, which becomes A., B. . . . 

 H., N., or rather, since N.'s development is the same as that 

 of H., it becomes A., B. . . . H. The fact that regression is 



