324 THE INDUCTIONS OP BIOLOGY. 



maintains its type with great persistence, the comparatively 

 heterogeneous domestic race frequently produces individuals 

 more unlike the average type than the parents are. 



Though unlikeness among progenitors is one antecedent of 

 variation, it is by no means the sole antecedent. Were it 

 so, the young dnes successively born to the same parents would 

 be alike. If any peculiarity in a new organism were a direct 

 resultant of the structural differences between the two organ- 

 isms which produced it; then all subsequent new organisms 

 produced by these two would show the same peculiarity. But 

 we know that the successive offspring have different peculiari- 

 ties : no two of them are ever exactly alike. 



One cause of such structural variation in progeny, is func- 

 tional variation in parents. Proof of this is given by the fact 

 that, among progeny of the same parents, there is more differ- 

 ence between those begotten under different constitutional 

 states than between those begotten under the same constitu- 

 tional state. It is notorious that twins are more nearly alike 

 than children borne in succession. The functional conditions 

 of the parents being the same for twins, but not the same for 

 their brothers and sisters (all other antecedents being constant), 

 we have no choice but to admit that variations in the func- 

 tional conditions of the parents, are the antecedents of those 

 greater unlikenesses which their brothers and sisters exhibit. 



Some other antecedent remains, however. The parents 

 being the same, and their constitutional states the same, vari- 

 ation, more or less marked, still manifests itself. Plants 

 grown from seeds out of one pod, or animals produced at one 

 birth, are not alike. Sometimes they differ considerably. 

 In a. litter of pigs or of kittens, we rarely see uniformity of 

 markings; and occasionally there are important structural 

 contrasts. I have myself recently been shown a litter of New- 

 foundland puppies, some of which had four digits to their 

 feet, while in others there was present, on each hind-foot, 

 what is called the " dew-claw " a rudimentary fifth digit. 



Thus, induction points to three causes of variation, all in 



