GERMINAL SELECTION 151 



as we have already seen, the different mental (jualities of the parents 

 are to a certain extent separately transmitted, that is, since they may 

 appear in the children in the most diverse combinations, we should 

 rather be surprised that pronounced talent in a specific direction can 

 persist in a family for two and a half centuries than that it should 

 do so very rarely. For reducing division is always combining the 

 existing mental qualities anew, and amphimixis is adding fresh ones 

 to them. 



Thus germinal selection, that is, the free, spontaneous, but 

 definitely directed variation of individual groups of determinants, is 

 at the root of those striking individual peculiarities which we call 

 specific talents ; but it can attain to the highest level only rarely and 

 in isolated cases, because these talents are not favoured by personal 

 selection, and therefore the excessively highly developed determinants 

 upon which they depend may be dispersed in the course of genera- 

 tions; they may sink to smaller majorities, or even to minorities, in 

 which case they will no longer manifest themselves in visible mental 

 qualities! 



We deduced the process of germinal selection on the basis of the 

 assumption that the nutrition of all the parts and particles of the 

 body, therefore also of the determinants and biophors of the germ- 

 plasm, is subject to fluctuations. We regarded the resulting variations 

 of these last and smallest units of the germ-plasm as the ultimate 

 source of all hereditary variation, and therefore the basis of all the 

 transformations which the organic world has undergone in the course 

 of ages and is undergoing still. 



We have still to inquire whether we can give any more precise 

 account of the nature of these units of the germ-plasm. If I mistake 

 not, we may say at least so much, that all variations are, in ultimate 

 instance, quantitative, and that they depend on the increase or 

 decrease of the vital particles, or their constituents, the molecules. 

 For this reason I have hitherto always spoken of only two direc- 

 tions of variation — a plus or a minus direction from the average. 

 What appears to us a qualitative variation is, in reality, nothing- 

 more than a greater or a less, a difi'erent mingling of the constituents 

 which make up a higher unit, an unequal increase or decrease of 

 these constituents, the lower units. We speak of the simple growth 

 of a cell when its mass increases without any alteration in its 

 composition, that is, when the proportion of the component parts and 

 chemical comljinations remains unchanged ; but the cell changes its 

 Vonttitution when this proportion is disturbed, when, for instance, 

 the red pigment-granules which were formerly present but scarcely 



