GENESIS, HEREDITY, AND VARIATION. 283 



showing, in accordance with the almost universal belief oi 

 breeders, that with animals and plants a cross between different 

 varieties, or between individuals of the same variety but oi 

 another strain, gives vigour and fertility to the offspring ; and 

 on the other hand, that close interbreeding diminishes vigour 

 and fertility," a conclusion harmonizing with the current 

 belief respecting family-intermarriages in the human race. 

 Have we not here a solution of these facts ? Relations must, on 

 the average of cases, be individuals whose physiological units 

 are more nearly alike than usual. Animals of different 

 varieties must be those whose physiological units are more 

 unlike than usual. In the one case, the unlikeness of the 

 units may frequently be insufficient to produce fertilization ; 

 or, if sufficient to produce fertilization, not sufficient to produce 

 that active molecular change required for vigorous develop- 

 ment. In the other case, both fertilization and vigorous 

 development will be made probable. 



Nor are we without a cause for the irregular manifestation ot 

 these general tendencies. The mixed physiological units com- 

 posing any organism, being, as we have seen, more or less se- 

 gregated in the reproductive centres it throws off; there may 

 arise various results, according to the degrees of difference 

 among the units, and the degrees in which the units are segre- 

 gated. Of two cousins who have married, the common grand- 

 parents may have had either similar or dissimilar constitu- 

 tions; and if their constitutions were dissimilar, the probability 

 that their married grandchildren will have offspring will be 

 greater than if their constitutions were similar. Or the 

 brothers and sisters from whom these cousins descended, in- 

 stead of severally inheriting the constitutions of their parents 

 in tolerably equal degrees, may have severally inherited them 

 in very different degrees : in which last case, intermarriages 

 among the grandchildren will be less likely to prove infertile. 

 Or the brothers and sisters from whom these cousins de- 

 scended, may severally have married persons very like, or 

 very unlike, themselves ; and from this cause there may 



