EVOTOMYS 213 



essentially as in normal members of the subfamily — apart from 

 the peculiarities just noticed; m^ has always a short anterior 

 loop and in most s^^ecies, even in very young stages of wear, no 

 more than four outer and five inner salient angles can be recognized. 

 On two occasions, however (Fig. 78, i and 56), I have found 

 in young teeth an ephemeral enamel islet representing the last 

 trace of a fourth outer infold, and in another case (Fig. 78, 2) the 

 fourth outer angle is comj^licated by a prism fold in front of which 

 there is a distinct vestige of a fourth outer valley ; so that we may 

 conclude that Erotomyn has descended from an ancestor with a 

 more complex )Hj than that now normal in the genus. Much 

 more definite evidence pointing in the same direction is afiorded 

 by the m^, which in jiractically all members of the genus is more 

 complex in early stages of wear than it is in ordinary adult 

 speciments. In this tooth, in some species, as many as five salient 

 angles may be traced on each side in early stages of wear ; but 

 the jiosterior elements are more or less evanescent, and the tooth 

 is commonly reduced to one with three outer and three or four 

 inner salient angles. In very old specimens m^ acquires a pattern 

 strongly resembling those found in such genera as AUicola and 

 Hyperaciius, and as explained above in the general introduction 

 this supi^orts the belief that the genera in question are offshoots 

 from ancient forms of Evotomys. 



Superspecific groups. — The numerous sj^ecies of Evotomys now 

 recognized are all very closely related to each other. They may 

 be arranged in four groups respectively typified by E. qlareolus, 

 E. rutilus, E. iiageri, and E. ritfocanus among Old World forms. 

 The earliest remains of the genus yet discovered are those occurring 

 in the Cromerian (Norfolk Forest Bed Series) deposits, in the 

 High Terrace of the Thames, and in early Middle Terrace deposits. 

 These fossils, judging from the teeth which alone are known, 

 must be ranked as early members of the glareolus or nageri groups. 

 They indicate clearly that the genus was fully developed in 

 Cromerian or U^jper Pliocene times and that we must look for 

 the more primitive ancestral forms in considerably older horizons. 



Of the four groups, that of E. glareolus is the most primitive 

 in that it seems to contain forms primitive enough to be regarded 

 as representing the central stock from which the other rather 

 higher groups may have descended. The primitive characters, 

 brachyodonty, greater complexity of cheek-teeth, weakness of jaw 

 muscles, small size, delicacy, and roundness of the skull, moderate 

 inflation of auditory bullae, small bodily size, greater length of 

 ears and tail, nakedness of feet and tail, dark coloration and 

 difEuseness of mantle, are most clearly shown by the members 

 of this group. 



From such a generalized glareolus stock the rutilus and 

 nageri groups seem to have directly arisen and diverged. 



The rutilus group appears to be an offshoot from some of the 

 lowest members of the glareolus group which has become adapted 



