SUCCESSION OF MAMMALIAN TEETH. 447 



the consideration of reverse process seen to be taking place in 

 the cetacea, and partially from palaeontological evidence. He 

 first (26) adduced Triconodon in proof of this view, regard- 

 ing- the molar tooth of that animal as a product of the fusion 

 of three simple reptilian teeth ; but in his later paper (29) he 

 puts forward the view that the multituberculata are the forms 

 which best demonstrate this point, and regards the three 

 longitudinal rows of cusps seen in the multituberculate molar 

 as indicative of the fusion of three distinct tooth generations, 

 each consisting of a number of similar teeth. In the molar 

 of the higher mammals he sees only two tooth generations 

 represented, viz., the milk and replacing set. This last 

 condition is exemplified in the porpoise, in which he believes 

 a fusion takes place between the two dentitions (28, p. 411, 

 fig. 89), but the facts which he there describes appear rather 

 indicative of the formation of a cusp by outgrowth from a 

 simple conical tooth with far less effort than is required to see 

 in them the fusion of two distinct dentitions. His belief that 

 the "anlage" of the would-be successional molar becomes 

 fused with that of the milk dentition will not hold for all mam- 

 mals, for if the lingual continuation of the dental lamina (35, 

 fig. 20, f. 2c, p. iv.) represents the anlage of the replacing 

 teeth that structure can be seen in some mammals to remain 

 quite distinct from the adult molar and in the end to gradu- 

 ally disintegrate, as the growth energy is abstracted from it 

 by the larger and earlier developed tooth. 



We find then that at present sufficient developmental 

 evidence is not forthcoming to justify the conclusion that 

 the mammalian molars are formed by the fusion of distinct 

 teeth. With regard to the palseontological evidence, that 

 is by no means so conclusive as Kilkenthal seems to sug- 

 gest, for the majority of observers in this branch of the 

 study, and notably Osborn (43) and Schlosser (44), main- 

 tain that the tritubercular tooth is the primitive mamma- 

 lian type and that it is to be derived from the simple rep- 

 tilian cone, not by fusion with two other teeth but by the 

 development of accessory cusps upon outgrowths of the 

 pulp of the simple tooth. The steps in this evolution they 

 exemplify by such forms as Dromotherium and Microconodou 



