1906.] TEETH OF CREODONTS. 47 



and the patterns produced vary in different families of the 

 Marsupials. 



Were there no more than this to be said, we should be pro- 

 vided with a criterion of marsupial affinity both certain and easy 

 of application. But unfortunately the case cannot be fully stated 

 quite so simply. Whilst it remains quite true that all marsupial 

 enamels present this chaiucter of penetration by the dentinal 

 tubes, the converse is not quite true. Thus Hyrax has an enamel 

 so richly penetrated by dentinal tubes that it might be easily 

 taken to be a marsupial enamel, though in this respect it stands 

 quite alone among placental mammals. But traces of this 

 peculiarity are to be found in much reduced degree in certain 

 Insectivora, notably in the Shrews ; this occurrence in Insectivora 

 may possibly be interj)reted as a survival from some mai'supial 

 form of ancestor. But this explanation is not available for all 

 cases : in very reduced degree the character has been found in the 

 Jerboa, in some Carnivora, and even in Man, though in Man the 

 rarity of the occurrence and its irregularity when it does occur 

 suggest that it is pathological, or at least that it is a reversion 

 towards something which has disappeared long ago. And investi- 

 gations of my own (8) into the development of enamel, and 

 especially of marsupial enamel, distinctly point to this penetration 

 of the epiblastic enamel by tubes continuous with those of the 

 mesoblastic dentine being a primitive character, to which some 

 slight tendency to revert has not been quite lost by placental 

 mammals. 



Hence, in the interpretation of the occurrence of this character 

 a different value appears to attach to negative and positive 

 results : if we find no tubes at all in the enamel, we shall, I think, 

 be quite justified in saying that no near affinity with the Marsupials 

 can exist. On the other hand, if we find rudimentary traces of this 

 penetration, we shall not be justified in attaching great impoi-tance 

 to it as an evidence of marsupial affinity, though if we find an 

 abundant penetration we shall have a character which, so far as 

 is known, is peculiar to Marsupials and to Hyrax. 



Having thus cleared the ground as to the value of the evidence, 

 it remains to describe in slightly greater detail what is met with 

 in the Marsupials, in Carnivora, and in Creodonts. 



Marsupial Enamel. 



A general character of marsupial enamels is the simplicity of 

 the course pursued by the enamel prisms ; each prism pursues, as 

 a rule, an almost straight course from the dentine to the enamel 

 surface, and where marked curvatures do occur, all of the 

 contiguous prisms pursue the same curve, so that no pattei'ns are 

 produced by neighbouring prisms crossing one another. Where, 

 however, the tubes are very abundant, the enamel prisms can 

 hardly be seen at all, and we have to take the tubes as indicative 

 of their course. 



