A PROBLEM IN AMERICAN ANTHROrOLOGY. 479 



inasmuch as it is only very recently that we are approach in o-ji. standard 

 ot uniformity in these expressions. It is now more than ever essential 

 that the anthropologists should agree upon a method of expressing 

 certain observed facts in somatology, so that the conscientious labors 

 of an investigator who has had a special opportunity for working upon 

 one group of man may be made available for comparison by investi- 

 gators of other groups. 



rro])ably the old method, still largely in vogue, of stating averages 

 is responsible for many wrong deductions. If we take 100 or more 

 skulls of any people, we shall find that the two extremes of the series 

 ditl'er to a considerable extent from those which naturally fall into 

 the center of the series. These extremes, in the hands of a zoologist, 

 would be considered the su])varieties of the central group or variety. 

 So in anthropology we should take the central group of the series as 

 furnishing the true characters of the particular variety or grouj) of 

 man under consideration, and should regard the extremes as those 

 which have been modified by various causes. It may be said that this 

 central group is defined by stating the mean of all the characters, l)ut 

 this is hardly the case, for by giving the mean of all Ave include such 

 extraneous chai'acters as ma}^ have ])een derived by admixture or from 

 abnormal conditions. 



The many difi'ering characteristics exhibited in a large collection of 

 crania ])rought together from various portions of America, North 

 and South, it seems to me, are reducible to several great groups. 

 These may be generally classed as the Eskimo type, the northern and 

 central or so-called Indian type, the northwestern brachycephalic type, 

 the southwestern dolichocephalic type, the Toltecan brachycephalic 

 type, and the Antillean type, with probably the ancient Brazilian, the 

 Fuegian, and the pre-Inca types of South America. Each of these 

 types is found in its purity in a certain limited region, while in otiier 

 regions it is more or less modified by admixture. Thus the Toltecan, 

 or ancient Mexican, type (which, united with the Peruvian, was sepa- 

 rated as the Toltecan family even by Morton) occurs, more or less 

 modified by admixture, in the ancient and modern pueblos and in the 

 ancient earthworks of our central and southern valleys. In Peru, 

 more in modern than in ancient times, there is an admixture of two 

 principal types. At the north of the continent we again find certain 

 traits that possibly indicate a mixture of the Eskimo with the early 

 coast peoples both on the Pacific and on the Atlantic sides of the con- 

 tinent. The north-central Indian type seems to have extended acro>s 

 the continent and to have Irranched in all directions, while a similar 

 ))ut not so extensive branching, northeast and south, seems to have 

 been the course of the Toltecan type. 



This is not theorizing upon the same facts from whicli Mortoiidivw 

 the conclusion that all these types were really one and the same. 



