28 PHYSICAL INVESTIGATION. 



possess but a small degree of constancy. 1 Where mongrels 

 intermix not among themselves but with one of the parent 

 type (e. g. Mulattoes with Negroes), and revert to it, it is no proof 

 for the specific difference of the parent stocks, since it is a well 

 known fact that in the intermixture of races the more numer- 

 ous race absorbs the less numerous. 



This reversion represents only a special application of the 

 general rule, according to which Blumenbach has endeavoured 

 to determine the distinction between species and race, namely, 

 that all differences between individuals or groups of individuals 

 which may be considered as having been produced by external 

 influence, or have arisen in the course of time, do not consti- 

 tute specific differences. 



Though the general validity of this rule may be readily 

 admitted, still its application is often very difficult and uncer- 

 tain. When, as mentioned by Wagner (in Prichard), the same 

 species of animal is met with in Sumatra, Java, Borneo, Timor, 

 Celebes, and even in the Philippines, or in the Asiatic conti- 

 nent, exhibiting in all these countries constant variations, it 

 might be difficult to decide whether in such cases we should 

 assume different species or races. The fact is, that the axiom 

 that all individuals which only differ in variable pecularities, 

 belong to the same species, is only a different version of the 

 notion of species as an assemblage of constant characters, and 

 presents nothing new in relation to the distinction between 

 species and race. This conception of species possesses, never- 

 theless, the advantage of fixing our attention on the point 

 from the investigation of which, the definition of species, in 

 doubtful cases, may be expected, namely, the range of variation 

 for every species. 



It may also be observed with regard to this last mark of 

 distinction, that the inferences drawn therefrom to unity of 

 species have greater value than when applied to difference of 

 species. The former must be considered proved as soon as 

 it is demonstrated that the greatest variations in the respective 

 individuals are still within the limits of those differences which 



1 I. Geoffrey St. Hilaire thinks that the characters of a race are the more 

 constantly transmitted the older the race is. 



