190(3.] TEETH OF CREODOXTS. 57 



was sent to me by Dr. Matthew as an example of an early true 

 Carnivore. In it the enamel prisms ai-e almost straight and no 

 decussation, or only the faintest trace of decussation, of the prisms 

 of different planes is to be seen. It resembles chiefly the enamel 

 of Dklynictis, and difi;ers in respect of its greater simplicity from 

 that of the other Creodonts examined and from recent Carnivora. 

 My section of Cynodictis embraces the whole tooth, so that there 

 is no question as to greater complexity of pattern existing in any 

 other parts of the tooth. In some of the Creodont enamels, 

 and particularly in Cynodictis, slight indications of a rudimentary 

 penetration of the enamel by dentinal tubes are seen, but in none 

 does it exceed or even attain to the amount seen occasionally in 

 recent Carnivora {cf. text-fig. 17, p. 51). 



Conclusions. 



The nature and the limitations of the evidences of affinity which 

 can be derived from a study of the minute structure of teeth have 

 already been alluded to, and it must not be forgotten that it is 

 unsafe to build too much upon any one single character. 



Bat, so far as the structure of their enamel may be taken as 

 evidence, neither Borhycena, Pachycena, Hycenodon, Sinopa, 

 Mesonyx, Oxycena, Didynictis, nor Cynodictis presents any greater 

 resemblance to Marsupials than do the recent Carnivora. On the 

 other hand, with the exception of Didynictis and Cynodictis, the 

 enamel has reached just that stage of evolution found in the true 

 Carnivora, and the enamel patterns are strikingly similar to those 

 of recent Carnivora. 



The uniformity of the patterns fou:nd in all of the Creodonts 

 examined, excepting again Didynictis and Cynodictis, seems to 

 point to the structure of their enamel having attained to a sort of 

 finality ; that is to say, it was probably not undergoing any rapid 

 evolutionary changes, a conclusion borne out by its close resem- 

 blance to that of their descendants, the recent Carnivoi'a. 



The absence of the peculiar stamp of the marsupial, the tubular 

 enamel, would justify us in saying that they certainly do not stand 

 very near to any marsupial, and that if thei*e be a marsupial 

 ancestor, or an ancestor common to the Maivsupials and to the 

 Creodonts, it must be sought considerably farther back than any 

 of those examined. This is a somewhat disappointing conclusion : 

 when I undertook the investigation I quite expected to find some 

 distinct indication of marsupial relationship ; that is to say, I 

 expected to find that the general resemblance in macroscopic 

 character of the dentitions to those of the polyprotodont 

 Mai'supials would have been accompanied by histological 

 resemblances. 



I have also been sui-prised to find that the enamels of Didynictis 

 and of Cynodictis are actually simpler than those of the other 

 Creodonts, and simpler than most recent Carnivora. As Cynodictis 

 at all events appeal's to be nearer to the true Carnivora than 

 are the Creodonts, the simplicity of its enamel as compared with 



