yS Bulletin American Museum of Natural History. [Vol. XVI, 



are well known; in 1895 Nehring referred to the long- and 

 short-skulled races of dogs associated with early races of 

 man; quite recently also Wortman (1899) has distinguished 

 between short- and long-jawed races of Tertiary dogs, and 

 Matthew (1901) has distinguished between the long- and 

 short-skulled races of Oreodonts, as a basis of classification. 

 These are a few examples, among many which might be 

 found, of attention directed to such facts ; but I am not aware 

 of any general application of dolichoce pkaly and brachycephaly 

 as factors in cranial and dental evolution, and as correlated 

 with the proportions of the limbs and habits of feeding. 



At all events the principle has not found its way into palas- 

 ontological literature, with which I am fairly familiar, and 

 was reached by myself independently and purely inductively 

 while engaged upon the phylogeny of the Rhinoceroses (1900). 

 After accumulating a great number of facts on the evolution 

 of this baffling group the correlation of long limbs with long 

 skulls (dolichopody and dolichocephaly) and short limbs with 

 short skulls (brachypody and brachycephaly) suddenly ap- 

 peared as a key, and was expressed in the following statement : 



"It [the classification adopted] sets aside several homo- 

 plastic parallel characters heretofore employed in Rhi- 

 noceros evolution and attempts to establish a firmer basis in 

 the fundamental proportions of the skull, whether dolichocephalic 

 or brachycephalic, in the correlated proportions of the body, and 

 in the location of the horn cores. These characters are found 

 to be more distinctive of phyla than the pattern of the molar 

 teeth." 



The full bearings of the principle were only partly per- 

 ceived at this time, and singularly enough I turned to the 

 study of the Titanotheres for the Geological Survey Mono- 

 graph without reference to my previous work on the Rhi- 

 noceroses and wholly unbiassed by any theory. Aided by Mr. 

 W. K. Gregory about eighty-five skulls were measured and 

 studied, hundreds of facts were noted which seemed to have 

 no particular significance; finally all these data were put 

 together and the conclusion was reached again, inductively, 

 that dolichocephaly and brachycephaly are among the doniinat- 



