IV METHODS AND RESULTS OF ETHNOLOGY 217 



not found in our tongue, very probably exists in our pedigrees. 

 The ethnology of France is still more complicated. Many 

 writers make the Parisian a Roman on the strength of his 

 language ; whilst others make him a Kelt on the strength of 

 certain moral characteristics, combined with the previous 

 Kelticism of the original Gauls. Spanish and Portuguese, as 

 languages, are derivations from the Latin ; Spain and Portugal, as 

 countries, are Iberic, Latin, Gothic, and Arab, in different 

 proportions. Italian is modern Latin all the world over ; yet 

 surely thore must be much Keltic blood in Lombardy, and much 

 Etruscan intermixture in Tuscany. 



" In the ninth century every man between the Elbe and the 

 Niemen spoke some Slavonic dialect ; they now nearly all speak 

 German. Surely the blood is less exclusively Gothic than the 

 speech." 1 



In other words, what philologer, if he had 

 nothing but the vocabulary and grammar of the 

 French and English languages to guide him, 

 would dream of the real causes of the unlikeness 

 of a Norman to a Provencal, of an Orcadian to a 

 Cornishman ? How readily might he be led to 

 suppose that the different climatal conditions to 

 which these speakers of one tongue have so long 

 been exposed, have caused their physical differ- 

 ences ; and how little would he suspect that these 

 are due (as we happen to know they are) to wide 

 differences of blood. 



Few take duly into account the evidence which 

 exists as to the ease with which unlettered 

 savages gain or lose a language. Captain Erskine, 

 in his interesting " Journal of a Cruise among the 

 Islands of the Western Pacific," especially remarks 

 1 Latham, Man and his Migrations, p. 171. 



