138 THE EVOLUTION THEORY 



in Man may perhaps be explained on this principle, and I had already 

 referred it to the more rapid growth and duplication of certain 

 determinants of the germ even before formulating the idea of 

 germinal selection. In this I was at one w^ith the pathologist 

 Ernst Ziegler, who had designated polydactylism as a germ-variation, 

 and in contrast to others had not interpreted it in an atavistic 

 sense, as a reversion to unknown six-fingered ancestors. All excessive 

 or defective hereditary malformations may be referred to germinal 

 selection alone, that is, to the long-continued progressive or 

 regressive variation of particular determinant-groups in a majority 

 of ids. 



The fact that, as far as our experience goes, superfluous fingers 

 are never inherited for more than five generations may be simply 

 explained, for there has been no reason for the intervention of 

 personal selection, either in the negative sense, for the six-fingered 

 state does not threaten life, nor in the positive, since it is not of 

 advantage. The deformity depends on spontaneous germinal varia- 

 tion, which must have taken lAsice in a majority of ids or it would 

 not have become manifest. But such a majority of ' polydactylous ' 

 ids is liable to become scattered again in every new descendant, 

 and to be reduced again into a minority which can no longer make 

 itself felt by the chances of reducing division and the admixture 

 of normal ids in amphimixis. A polydactylous race of men could 

 only arise through the assistance of personal selection; in that case 

 there would doubtless be just as much chance of success in breeding 

 a six-fingered race as there was in breeding the crooked-legged 

 Ancon sheep from a single ram which was malformed in this 

 manner. Without a gradual setting aside of the germs with normal 

 ids, that is, without personal selection, such spontaneous deformities, 

 and indeed all spontaneous variations, must fail of attaining to 

 permanent mastery. 



This must frequently be the case in free nature also, but we 

 shall have to investigate later on, in the section devoted to the 

 formation of species, whether external circumstances (inbreeding) 

 may not also occur which make it possible for spontaneous variations 

 to become constant breed-characters, even although they remain 

 neither good nor bad, and are thus not subject to the action of 

 personal selection. 



In general, however, amphigony with its reduction of the ids 

 and its constant mingling of strange ids will form the corrective 

 to the deviations which may arise through the processes of selection 

 within the id, and which lead to excessive or superfluous development 



