| 
. 
. 
| 
4 
————er ee 

MURINA sol 
is in complete accord with the results obtained in MWzcrotine, 
where in addition to cusp 7 other ancient elements (x, ') are 
sometimes present though they have not yet been found in the 
teeth of living Murine. The results in question fully support 
the theory of a multitubercular origin of the rodent cheek-tooth 
postulated by Forsyth Major, and are in complete conflict with 
the tritubercular theory of Cope and Osborn. 
The lower cheek-teeth are in one way more modified than 
the upper ones ; the outer row of tubercles corresponding to the 
inner row of the upper teeth is so reduced that it has been 
either completely ignored or else treated as a mere cingulum 
by all writers except Tullberg. In Afodemus and many 
exotic genera this third row is comparatively well-developed, 
and consists normally of three tubercles in 7,; it has, on the 
other hand, completely vanished from the teeth of J/us and 
many other genera; throughout the sub-family it shows the 
high degree of variability which is characteristic of vanishing 
structures. From the circumstance that it is a marked feature of 
the teeth of Afodemus and some other J7urine, that it forms an 
important part of the teeth of some MWtcrotine, e.g. Dicrostonyx 
(Pl. XXVIII., Fig. 2), and occurs ephemerally in the young teeth 
of still more remote relatives, e.g. Spalax (Méhely, Speczes Generis 
Spalax, 1913, 305, fig. 10), and from the fact that it comprises such 
ancient and well-known molar elements as the ‘proto-” and 
‘“‘hypoconids” of trituberculy, we may conclude that this third 
or outer row of tubercles was an important feature in the lower 
molars of the ancestral IZurzdw. It has suffered reduction and 
even obliteration in consequence of the hypertrophy of the 
‘intermediate ” cusps of the teeth of these remarkable rodents. 
Two other elements of the lower cheek-teeth deserve notice. 
In wz, and m, a posterior ‘‘accessory” tubercle (Pl. XXVIII, 
Figs. 4-10) is almost constantly present in JMZurine, this 
structure appears to be the homologue of part of the posterior 
transverse loop of the corresponding teeth in MZtcrotine. In 
most species of Afodemus, and in some other murines, 7, has 
an anterior ‘‘accessory” tubercle which appears to represent 
the anterior loop of the mz, of Wzcrotus. 
Because of their forward position in the jaws the greater 
share of the work of mastication falls upon the anterior cheek- 
