THE HUMAN SPECIES. 255 



gence, in order, by their aid, to establish a primaeval unity of a 

 genus, how much more important must the same method prove 

 in researches after the aboriginal unity of a sub-typical stem 

 of Man. If there were no such other indications as have 

 already been noticed, by these facts alone we may with confi- 

 dence appeal to the presence of a considerable portion of Cau- 

 casian blood, in the composition of the master race of the 

 Polynesian islands. It is undeniably conspicuous in some of 

 the groups, less so in others, and evident in despite of linguistic 

 considerations, which, to say the least, are still not sufficiently 

 mature to admit the generalizing conclusions of Humboldt. 

 The Maori tongue of New Zealand is an example, which, 

 while it shows the presence of a Semitic element in the com- 

 position, is but feebly tinged with Malay; perhaps, by reason 

 of the great majority of its component words being the offspring 

 of Papua dialects, the basis of the population being originally 

 of Eastern Negro derivation, only by degrees amalgamated or 

 destroyed. Whence these two races came, can now be only 

 conjectured from the reminiscences of the people, that two 

 immigrations originally took place on these islands; they still 

 name the localities, and assert one to have come from the east 

 and the other from the west. To individuals or families of the 

 earliest Polynesian wanderers, the introduction of at least one 

 system of doctrine, in South America, may be ascribed ; and 

 to another, of Caucaso-Mongolians, a second, which appears to 

 have reached the north-west coast, and finally to have estab- 

 lished itself on the plateau of Anahuac. These considerations 

 lead us to the New Continent, before the two historical arche- 

 typical stocks of the Old can be traced out without interrup- 

 tion. 



THE AMERICAN SUB-TYPICAL STEM. 



Though researches on the primitive population of America 

 may be deemed unphilosophical, because the conclusions are 



