182 MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



of the whole series, this average may be far from the true average capacity of all ; but we have 

 the following reason for thinking otherwise, : That capacity (1,330) which comes nearest to the above 

 average, belongs to skull H. 7, and this it is, that next after the &quot; type &quot; (H. 40) is the most typical 

 skull of the whole series as shown by its various indices. 



The table of Broca, with which we have to compare this average, gives separate figures for the 

 males and the females of eacli race. We have calculated the combined averages and made our 

 comparisons with these. Broca gives 29 series including the most diverse races, but no American 

 Indians. We find but three of his series having a lower cranial capacity than the Saladoans; 

 these are the &quot; Hottentots and Bushmen,&quot; the &quot; Australians,&quot; and &quot; Farias of Allipoor (Calcutta).&quot; 

 Such inferior races as the Negroes of Africa, the Papuans, the New Caledonians, and the Tasma- 

 nians seem to rank in this character above the Saladoans. 



Our Table xxiv presents some small series of average cranial capacities of (lower) races 

 represented in the Army Medical Museum. They are taken from the series of 101 (see Table 

 LXXXI) 2 Navajos and 10 Peruvians being added. All the races, not American Indians, men 

 tioned in this table, viz, Sandwich Islanders, Mongolians, New Zealanders, American Negroes, 

 and Eskimos, it will be seen, have larger brain cases than our Saladoans. The position of the latter 

 with regard to other autochthones of both North and South America is shown in the following 

 extract from the table : 



(1) Siouan tribes 1463 



(2) Pah Utes 1367 



(3) Apaches 1331 



(4) Ancient Californians 1323 



(5) Navajos 1315 



(6) Saladoans 1313 



(7) Peruvians 1295 



It is not in accordance with current theories that a people as advanced in arts and social organ 

 ization as that of the Salado Valley evidently was should have a cranial capacity superior only 

 to such low races as the Hottentots and Australians. It must be borne in miud, too, that the un- 

 cremated remains of the Saladoans probably represent a superior class of this community. Still, 

 small as is their cranial capacity, it is greater than that of the Peruvians, who were a race more 

 advanced than the Saladoans. We have little to suggest in explanation of these facts. Perhaps 

 the subject of cranial capacity in relation to culture may have to be reconsidered. The Saladoaus 

 were a people of low stature and rather slight physique, and the relation which the skull bears to 

 the rest of the skeleton may be a factor in the problem. We have as yet no evidence to show 

 that distortion reduces the capacity of the cranium. 



$ 14. THE TYPE SKULL 



The following method is the one we have adopted for selecting a type skull from the series: 

 First. Let all the sets of indices be arranged in ordination. Second. Subtract the lowest index 

 in one ordination from the highest. Third. Divide the difference by 2, and add the quotient 

 thus obtained to the lowest index. This gives the theoretical mean of variation. Example: 

 Suppose we have a series of skulls with cephalic indices ranging from 80.00 to 90.00. The flrst 

 step, subtraction, gives us 10.00; the second step, division, gives us 5.00, and the third step, 

 addition, gives us 85.00, which is the theoretical mean of variation. The skull, if any, having this 

 index is the type of the series as far as concerns the cephalic index. In practice, however, where 

 we calculate indices to the second decimal place, it is not usual to find any skull with the 

 index exactly expressing the theoretical mean. The skull most nearly expressing it is taken as 

 the; type. 



It follows that if we take many different series of indices upon the same skulls we have to 

 determine what skull stands in the plurality of instances nearest the theoretical mean. Suppose 

 we calculate ten different series of indices upon 9 skulls (ail odd number is easier for the 

 purpose of explanation). If 1 particular skull expresses the theoretical mean of variation in 



