THE RACIAL GEOGRAPHY OF EUROPE. 763 



or eyes, the fallacy of direct comparison between the north and 

 south of Europe again becomes apparent. In the third place, it is 

 not easy to correct for the personal equation of different observ- 

 ers. A seeming brunette in Norway appears as quite blond in 

 Italy because there is no fixed standard by which to judge. The 

 natural impulse is to compare the individual with the general 

 population round about. The precision of measurements upon 

 the head is nowise attainable. 



There are two principal modes of determining the pigmenta- 

 tion of a given population. One is to discover the proportion of 

 so-called pure brunette types that is to say, the percentage of in- 

 dividuals possessed of hotli dark eyes and hair. The other system 

 is to study brunette traits without regard to their association in 

 the same individual. This latter method is no respecter of persons. 

 The population as a whole, and not the individual, is the unit. 

 North of the Alps they have mapped the pigmentation in the 

 main by types ; in France, Norway, Italy, and the British Isles 

 they have chosen to work by dissociated traits. Here again is a 

 stumbling-block in the way of comparisons. The absolute fig- 

 ures for the same population gathered in these two ways will be 

 widely different. Thus in Italy, while only about a quarter of the 

 people are pure brunette types, nearly half of all the eyes and hair 

 in the country are dark. That is to say, a large proportion of 

 brunette traits are to-day found scattered broadcast without asso- 

 ciation one with another. In Europe, as a whole, upward of one 

 half of the population is of a mixed type in this respect. Iq 

 America the equilibrium is still further disturbed. Nor should 

 we expect it to be otherwise. Intermixture, migration, the influ- 

 ences of environment, and chance variation have been long at 

 work in Europe. The result has been to reduce the pure types, 

 either of blonde or brunette, to an absolute minority. Fortu- 

 nately for us, in despair at the prospect of reducing such variant 

 systems to a common base, the results obtained all point in the 

 same direction whichever mode of study is employed. In those 

 populations where there is the greatest frequency of pure dark 

 types there also is generally to be found the largest proportion of 

 brunette traits lying about loose, so to speak. And where there 

 are the highest percentages of these unattached traits, there is also 

 the greatest prevalence of purely neutral tints, which are neither 

 to be classed as blond or brunette. So that, as we have said, in 

 whichever way the pigmentation is studied, the results in general 

 are parallel, certainly at least so far as the deductions in this 

 paper are concerned. Our map is indeed constructed in conform- 

 ity with this assumption.* 



* Dr. Livi, in his atlas to the superb Anthropometria Militare, has shown the parallel- 

 ism very clearly in Charts VI to IX, inclusive. The method employed in reducing the 



