STUDIES IN ANIMAL LIFE. 113 



sembling themselves, it is not a whit more surpris- 

 ing that the dogs of Victoria should resemble the 

 dogs of Semiramis, than that they should resemble 

 their parents : the chain of four thousand years is 

 made up of many links, each link being a repetition 

 of the other. So long as a single pair of dogs re- 

 sembling each other unite, so long will there be 

 specimens of that species, simply because the chil- 

 dren inherit the characteristics of the parents. So 

 long as negroes marry with negroes, and Jews with 

 Jews, so long must there be a perpetuation of the 

 negro and Jewish types ; but the tenth generation 

 adds nothing to the evidence of the first, nor the 

 ten thousandth to the tenth. 



I believe that this fallacy, which destroys the 

 whole value of the Cuvierian argument, has not 

 before been pointed out; and even now you may 

 perhaps ask if the fixity of species is not proved by 

 the fact that like produces like ? So far from this, 

 that it is only by the aid of such a fact in organic 

 nature that we can imagine new species to have 

 arisen; in other words, those who believe in the 

 variability of species, and the introduction of new 

 forms by means of modification from the old, al- 

 ways invoke the law of hereditary transmission as 

 the means of establishing accidental variations. 

 Thus, let us suppose the Egyptian king to have 

 had one hundred dogs, all of them staghounds, and 

 no other form of dog to have existed at that time 

 in that country ; the dog species would be repre- 



