278 DK. G. ELLIOT SMITH ON 



In tliis investigation our first aim must be the determination of the extent to which 

 the hrain in the Edentata conforms to the architectural plans which prevail elsewhere in 

 the Mammalia. Tlicn we shall be able to appreciate the modifications of the jirevalent 

 tjpe or types and study the factors M'hicli produce these variations, either to meet direct 

 physiological needs which the mode of life of the individual imposes, or to conform, to 

 some definite morphological plan which is not so directly the expression of functional 

 modification, and therefore is likely to have some definite taxonomic value. 



Before entering into the discussion of these problems, it will be well to review as briefly 

 as possible the state of our knowledge concerning the relationsliips of the Edentata at the 

 present time. 



It \^ ould perhaps be difficult to discover elsewhere am.ong the Mammalia five natural 

 groups of mammals which present such striking contrasts in appearance and habits of 

 life as we find in the five families which include all the existing Edentata. Striking as 

 these contrasts are, they are probably equalled by the marked diff'erences in the 

 anatomical features of the animals. The only reason for uniting these heterogeneous 

 families in one group is the fact that they possess in common certain negative characters 

 which distinguish them from all othei- mammals. The most important of these negative 

 features is the roducticn of the teetli, but even the characters of this reduction, the 

 featiu'es of the few teeth Ihat remain in certain families, and the little we know of their 

 development present most manifest distinctions. 



In spite of the absence of any very decided anatomical links between the different 

 recent families, the evidence of palgeontoiogy many years ago showed that both the 

 Sloths and Ant-eaters, although markedly dissimilar in themselves, exhibit signs of close 

 kinship to the extinct Ground-Sloths. Moreover, it is now very generally admitted that 

 the Armadillos also branched off at an early period from a stock common to the 

 ancestors of the Ant-eaters, Sh)ths, and Ground-Sloths. Hence Gill* has suggested the 

 separation of these three New-World families as a suborder from the Old-World families. 

 To this suborder he gives the name Xena)-thra, because in these American families the 

 zygapojjhvses of the vertebrae possess additional articular facets. 



The Imge collections of extinct Tertiary mammals which have been unearthed within 

 recent years have tlii'()\\ n a considerable amount of light upon the vexed problems of the 

 kinship of the Edentata. 



If we admit that the group of extinct mammals which Wortman has called Ganodontaf 

 represents the ancestors of the American Edentata, we must 1 e prepared to recognize in 

 the latter the degenerate representatives of a group of animals which are imited by 

 the closest ties of kinsliip to the ancestors of the Eodentia, Carnivora, and Ungulata. 

 It is not necessary to do more than mention in passing how vastly this suggestion is at 

 variance vdih the older and as yet more generally accepted view that the Edentata are a 

 very primitive group, which is so distantly related to the ordinary Eutheria that some 



* Gill, ' Standard Katural History,' 1884. 



t S. L. ■\Vortmaii, ' Psittacothcrium, a Member of a New and I'limitive Suborder of tlie Edentata,' Author's 

 edition, copied frcm tlio Hull. Amer. Mus. Nat. Hist. vol. viii. aitiele ivi. p. 259 (189t)). 



