TOOTH-GENESIS IN THE CANID^. 4G1 



(ii.) Behind the last upper molar the dental lamina is con- 

 tinued backwards for some distance ; it gets considerably dis- 

 torted and broken up ; but one part of it is more enlarged 

 than the rest. The facts do not allow one to speak with any 

 certainty, but I think it is possible to recognize in it the vestigial 

 remains o£ a third upper molar, since I can find no trace of such 

 remains in the corresponding position in the lower jaw. 



Numerical Variation of the Teeth of the Carnioora. 



The number of teeth present in the permanent dentition in 

 living Mammalia varies greatly, and though this variation is 

 somewhat narrower among the Carnivora still it is far from being 

 uniform. This is shown in the accompanying Table (p. 462), which 

 1 have compiled from llower and Lydekker's ' Mammalia,' and 

 certain points, which are worthy of note, may be readily seen. 



(i.) Mluroidea. The maximum number of teeth present among 

 the members of this group is 40, the Felidse and Proteleidae 

 falling as low as 30. 



(ii.) Cynoidea. The teeth A^ary from 40 in Cijon to 46 or 48 in 

 Otocyon ; the Canidge possessing 42. 



(iii.) Arctoidea. The Mustelidse have the smallest number (38), 

 while the Ursidre have 42. 



From this it will be seen that the maximum numerical varia- 

 tion is attained among the JEluroidea, the minimum by the 

 Arctoidea, while the Cynoidea occupy an intermediate positi<3n ; 

 and, moreover, by far the greater number of its members have 

 the same number of teeth, 42. This, I think, justifies the well- 

 established deduction that the ancestral form had 42 or more 

 teeth. It was also probably Pentadactyloid and Plantigrade. 

 These three characteristics are present among the living Ursidae. 



To eifect the numerical variations one of two things must 

 have happened — (i.) either teeth must have, in some instances, 

 been superadded, causing an increase in the number ; or (ii.) 

 some teeth must have become suppressed. I think the balance 

 of evidence is decidedly in favour of the latter, for the following 

 reasons : — 



(i.) Supernumerary teeth, of which examples are given by 

 Bateson (1) in his book 'Materials for the Study of Variation,' 

 are very rare ; and the number is never in excess of that found 

 among fossil forms, and may be regarded as "reversions to a 



