Dental Forwi/I(( of fJie ^hmdiv. 165 



of the first two classes runs comparatively little risk of 

 inverting the sequence in cases of tlie third class. The 

 principle involved in each of the three classes is one and the 

 same. If, for example, we examine an absolutely unworn 

 front lower cheek-tooth of Mitnomys or Arvicola -we find that 

 it possesses a tubcrcuhir cap, reminiscent of the bunodont 

 and brachyodont ancestors of the gionp ; such a cap occurs, 

 too, in the molars of most other rodents with " prismatic " 

 patterns in the adults. In slightly deeper levels of the 

 crown certain tubercles fuse together in such a way that 

 when the tooth is worn down the characteristic prismatic 

 pattern gradually comes into view. In an earlier species of 

 Mimomys (e. g., M. p/locanicus we find that ^' mi " (in the 

 conventional sense) has three re-entrant folds upon its outer 

 border for a considerable portion of the animaPs life ; but as 

 middle age is approached the walls of the foremost fold knit 

 together at a certain point — the inner part of the fold then 

 being converted into an enamel "islet," while the outer part 

 is represented by a vertical groove which channels the outer 

 surface of the tooth. Gradually the ^' islet '''' and the groove 

 die out in deeper levels of the crown, so that in old age, 

 when wear is far advanced, all traces of these vestigial struc- 

 tures are lost. In later species (e. g., M. intermedius) 

 exactly the same process, with exactly the same result, is 

 seen — with a difference ; the process is hastened, and the 

 last traces ('" islet " and external groove) of the third or 

 front outer fold vanish before the animal is fully grown. 

 In Arvicola amphibius and terrestris we have strictly parallel 

 changes taking place upon the molar crowns within the first 

 few weeks of the individual's existence ; in these creatures 

 there is a tubercular cap upon the summit of the tooth ; 

 then, in slightly deeper strata, the beginnings of those 

 " prisms " and folds which form the characteristic pattern of 

 the adult tooth, plus other prisms and folds, which, after 

 lingering for a day or two, the transitory memorials of 

 forgotten ancestors, rapidly disappear. 



By comparing the results obtained from the three lines of 

 inquiry indicated above, it is possible to form an idea of the 

 minimum amount of complexity which the cheek-teeth 

 must have possessed in the most primitive Murickx;, or at 

 least ill those forms directly ancestral to the various living 

 subfamilies. Investigations of this kind, involving the 

 careful examination of many thousands of teeth, closely 

 pursued for more than twenty years, have led me to infer 

 that the anterior cheek-teeth in the ancestral ]\[urida3 were 

 multitubercular structures^ consisting in the case of the lower 



