CHAP, x.] TEETH : HOMOLOGIES. 273 



pp. 259, &c.), in which the members of the series seem to have been 

 so far remodelled that the supposed individuality of the members 

 is superseded. In the Selachians several such cases were given, 

 but in Mammals the most manifest examples were seen in the 

 Phalangers and Ateles marginatus No. 200 (q.v.). In the latter 

 specimen there were four premolars on each side in the upper jaw, 

 and there was nothing to indicate that any one of them was super- 

 numerary rather than any other. In such a case I submit that the 

 four premolars must be regarded as collectively equivalent to the 

 three premolars of the normal. The epithelium which normally 

 gives rise to three tooth-germs has here given rise to four, and I 

 believe it is as impossible to analyze the four teeth and to apportion 

 them out among the three teeth as it would be to homologize the 

 sides of a triangle with the sides of a square of the same peripheral 

 measurement. 



Such a case at once suggests this question : if the four premo- 

 lars of this varying Ateles cannot be analyzed into correspondence 

 with the three premolars of the typical Ateles, can the three pre- 

 molars of this type be made to correspond individually with the 

 two premolars of Old World Primates ? 



In the case of Rhinoptera No. 396, for the reason given in 

 describing the specimen, there is plainly no correspondence be- 

 tween the rows of plates of the variety and those of the type, and 

 the rows are, in fact, not individual, but divisible. 



Though cases so remarkable as that of Ateles marginatus are 

 rare, there are many examples of supernumerary teeth, in the 

 region of the anterior premolars of the Dog or Cat for instance, 

 which cannot be clearly removed from this category. As indicated 

 in the fourth section of this Chapter, it is impossible to distinguish 

 cases of division of particular teeth from cases of the formation of 

 a new number of teeth in the series. Finally, on the analogy of 

 what may be seen in the case of Meristic Series having a wholly 

 indefinite number of members, it is likely that the attempt thus to 

 attribute individuality to members of series having normally a 

 definite number of members should not be made. 



B. 



18 



