552 PROCEEDINGS OF THE XATIOXAL MUSEUM. vol. 48. 



Ewart, 1 too, has furnished some measurements and indices ob- 

 tained from another skull of a Celtic pony. He has not given the 

 basilar length of the specimen ; but a careful estimate shows that this 

 must have been close to 437 mm. Its celphalic index 'is, therefore, 

 very close to 43.7. While the latter falls somewhat below the former, 

 the indices show that both horses belong among those called by 

 Xehring broad-faced. These two horses will be referred to later ; but 

 the writer does not at present see how Ewart ? s Celtic horses and his 

 Forest horses are to be distinguished by means of skull measurements. 



In 1908 2 Duerst discussed the origin of the various races of domes- 

 tic horses. He, too, concluded (pp. 399, 431) that our modern breeds 

 have arisen from three types, to which he applied the names " the type 

 of the steppe,'' " the type of the desert," and " the type of the woods." 

 These names do not, in all cases at least, correspond to those employed 

 by Ewart. Duerst 's steppe horse is represented by Xehring } s Equus 

 caballus germanicus vobustus; his " type of the woods." or u forest 

 type," has as its representative Equus caballus nehj-tngi, from which, 

 in Duerst's opinion, sprang the Celtic pony ; finally, Duerst's type of 

 the desert is his " Equus caballus pumpellii" which is " ancestrally 

 closely related to the Equus pvzewalskii Poljakoff." These views 

 illustrate the confusion of the subject; but, what is more important, 

 they seem to justify the view that nobody knows exactly how forest 

 horses, steppe horses, plateau horses, and desert horses differ one 

 from another. 



Much time and patience and talent have been expended in meas- 

 uring the skulls and other bony structures of the domestic horses, 

 and many 4; indices " have been determined. Equus caballus may be 

 regarded as the product of two or three distinct races or species and 

 as having been greatly affected by domestication and breeding. 

 Under these conditions it is proper to inquire whether any of the 

 indices obtained from the skulls of this group are of any value in 

 distinguishing species ; and if so, which ones ? Also whether they 

 can be employed in determining the ancestors from which the do- 

 mestic horses have come. It seems essential that these indices should 

 first be tested on unmixed and undomesticated species; and there 

 appears to be no reason why this should not now be done. Already 

 many measurements of Equus asinus, E. hemionus, and E. przeval- 

 slcii have been published, besides those of a few species of zebras. In 

 the United States Xational Museum there is now a considerable num- 

 ber of skulls of zebras of two species and of some subspecies of one 

 of these ; and doubtless there are other specimens in other museums. 



1 Trans. Roy. Soc. Edinburgh, vol. 45, p. 586. 



2 Publication 73, Carnegie Inst., Washington, vol. 2, pt. 6. 



