166 Mr. M. A. C. Hinton on the 



tooth (" wi ") of at least twenty-one tubercles arranged in 

 three longitudinal rows of seven each ; and in the case of 

 its opponent (" m^ ^^) of at least fifteen tubercles similarly 

 arranged in throe longitudinal series of five each. The fate 

 of these primitive elements has been various in the many 

 subfamilies now recognized. In most groups they have 

 suffered considerable numerical, reduction (greatest perhaps 

 in Gerbillinse, least in Microtinse) ; and this reduction has, 

 in the normal mammalian manner, affected primarily the 

 inner sides of the upper molars and the outer sides of the 

 lower molars. More or less complete transverse fusions 

 between the surviving tubercles of the three prmiitive longi- 

 tudinal rows have taken place in all sections of the Muridse, 

 such fusions being most complete and most difficult to 

 unravel in genera which have become completely hypsodont, 

 tliough still apparent in many which have remained brachy- 

 odont. Far more important, the fate of the three longi- 

 tudinal rows differs from section to section. In Murinse, 

 for example, the tubercles of the median row have grown at 

 the expense of those forming the inner row in upper molars, 

 the outer row in lower molars. In Cricetinae, on the other 

 hand, the tubercles of the median tow are atrophied to a 

 greater or less extent ; sometimes to such an extent that 

 even an experienced eye fails to detect the vestiges that 

 usually remain. Lastly, in Microtinse^ the middle tubercles 

 remain important elements, although fusions with those of 

 the outer or inner rows and the general progress towards 

 hypsodonty, which is so marked a feature in this group, have 

 all but completely masked them as independent structures 

 from our view. But, nevertheless, even in this subfamily, 

 more or less clear traces of these median tubercles as recog- 

 nizable elements of the molar crown are to be found occa- 

 sionally : sometimes upon the unworn summits of the teeth, 

 as in JDicrostonyx ; sometimes regularly persisting in adult 

 stages of wear, as in certain species of Lagurus ; but most 

 frequently cropping up as vestiges of peculiar form in slightly 

 worn teeth, or recurring as atavistic abnormalities in the 

 teeth of quite widely separated genera. In general, widely 

 confluent opposed dentinal spaces (as in the teeth of Evotomys, 

 Eothenoinys, or the well known confluent pair of the " m^ " of 

 Pitymys) may be taken as evidence of a relatively strong 

 development of certain of the tubercles of the median row. 

 Whether my estimate of the number of cusps primitively 

 present be substantially accurate or not, there is, on the 

 whole, abundant evidence that a high degree of complication 

 must have characterized the anterior cheek-tooth in each 



