THE RACIAL GEOGRAPHY OF EUROPE. 



623 



This is indeed true. Apparent exceptions prove the rule; for 

 where, in the heart of the Basque territory, the broad heads seem 

 to penetrate to the Spanish frontier, there is the ancient pass of 

 Roncesvaux, celebrated in history and literature. The broad- 

 headed type would naturally have invaded here if at all. Every- 

 where else the long-headed type seems to prevail, not only on 

 the Spanish slopes, but clear over to the foothills of the Pyrenees 

 on the other side in France.* 



If these facts be all true, what has become of our Basque phys- 

 ical type ? Where are our philological theories of purity of racial 

 representation ? If the Basques are indeed an unmixed race, 

 there must be one of these two types which is false. At first 

 the anthropologists sought thus to reject one or the other, French 

 or Spanish, for this reason. Then they laid aside their differ- 

 ences ; they abandoned entirely the old theory of purity of descent. 

 The Basque became for them the final complex product of a long 

 series of ethnic crosses. Each of the conflicting characteristics 

 was traced to some people, wherever found it mattered not. The 

 type was compounded by a formula, as a druggist puts up a pre- 

 scription. Bladd wrote in the light of such views. Canon Taylor, 



Spanish Basques. Tolosa, Guipuzcoa. 



in his Origin of the Aryans, holds that the broad-headed French 

 Basque is only a variation of the Alpine type which we have seen 

 prevails in all the southwest of France, with a dash of Lapp 

 blood. For him the Spanish Basque was, on the other hand, a 

 subtype of the long-faced Iberian or Spanish narrow head. The 



* The general limits of the mountainous country are shown by the five-hundred-meter 

 contour line on our subsequent map on page 625. 



