560 



PROCEEDINGS OF THE NATIONAL MUSEUM. 



vol. 48. 



of 114 mm., which amounts to somewhat more than 30 per cent of 

 the average length. 



Although these things are true, the indices might nevertheless be 

 relatively stable, but we do not find them to be so. The cephalic 

 index varies to an amount somewhat greater than 16 per cent of the 

 mean ; the craniocephalic index more than IT per cent ; and the f acio- 

 cephalic index more than 11 per cent. No such differences appear 

 in the wild species. It may be contended that these differences are 

 due to domestication; and we do find considerable, but more nar- 

 rowly limited, variations in the domestic ass. However, there are 

 evidences that similar differences existed among the European horses 

 of the latter half of the Pleistocene ; and on these differences has been 

 based the hypothesis that there existed among the Pleistocene a 

 number of races or subspecies of Equus cab alius. 



If now we consider those eight horses in the above list which are 

 placed above the horizontal line we find the following averages : 



Averages of measurements in millimeters and indices of small domestic horses. 



Basilar 

 length. 



Vertex 

 length. 



Frontal 

 width. 



Ce- 

 phalic 

 index. 



Cranial 

 length. 



Cranio- 

 cephalic 

 index. 



Facial 

 length. 



Facio- 



cephalic 



index. 



For. 

 mag. to 

 vomer. 



Vomer 



to 

 palate. 



Average. 



443 



489 



196 



44.2 



159 



36.0 



337 



76.1 



112 



96 



From this table and that from which these averages are obtained 

 it is seen that the cephalic index is high, indicating a broad face. 

 It is seen, too, that the difference between the highest and the lowest 

 expressions amounts to less than 9 per cent of the averages ; that the 

 cranio-cephalic index has a range of variation of only 5 per cent 

 of the average, and that the f acio-cephalic index has a range of only 

 7 per cent of the mean. 



Considering the eight examples which occupy the lower half of 

 the table on page 559, we find as follows : 



Averages of measurements in millimeters and indices of large domestic horses. 



Averages. 



Basilar 

 length. 



558 



Vertex 

 length. 



600 



Frontal 

 width. 



229 



Ce- 

 phalic 

 index. 



41.1 



Cranial 

 length. 



184 



Cranio- 

 cephalic 

 index. 



33.0 



Facial 

 length. 



413 



Facio 



cephalic 



index. 



74.1 



For. 

 mag. to 

 vomer. 



143 



Vomer 



to 

 palate. 



118 



From this table we learn that the cephalic index is rather low, 

 the average, 41.1, being slightly below the dividing line which Nehr- 

 ing drew between the narrow-faced and the broad-faced horses (240 

 according to his method=41.6 according to the method employed 

 here). The fluctuation of this index amounts to 11.4 per cent in 



