862 ON THE EXTINCT GlANT DOEMOUSE OF MALTA, [DeC. 3, 



maxilla, situate considerably in advance of the aiitei'ior zygomatic 

 root. Now in this resjjeet the fossil apparently differs from the 

 Dormice and resembles the Squirrels \ 



As regards the clieek-teeth, these pi'esent a considerable superficial 

 resemblance to those of such Dormice as have complicated enamel- 

 folds on the crown. Closer examination shows, however, a marked 

 difference, well displayed in the accompanying figures (p. 861). 

 In the simpler type of upper molars in the Dormice (Eliomys) the 

 folds form ridges, of which two unite to form columns on the 

 outer side of the crown, so as to give a somewhat tritubercular form 

 to the whole tooth ; and where the ridges are more complex 

 (Mjjoxus), and form more distinct plates, these curve together in 

 pairs on the outer side of the crown. On the other hand, in the 

 Maltese animal the four or five blunt ridges diverge from the 

 inner or postero-internal portion of the crown in a comb-like 

 manner, without any tendency to approximate on the outer border, 

 one or two of these ridges being shorter tlian the rest and arising 

 by a splitting of the enamel-folds. Both Eliomys and Myoxus have 

 distinct outer columns to the upper molars, while Muscardinus 

 differs markedly from the fossil in that the upper premolar is very 

 small and the first molar much longer than the second. The 

 foldings, too, on the second molar are much finer and of a diff"erent 



Comparing this type of dentition with that of the Sciuridce, the 

 nearest approximation is presented by Xerus and Pteromys. In 

 the former two main ridges and two talons proceed outwards from 

 an inner wall ; and in the latter this inner wall becomes thinner, 

 and the two talons are so deve'oped as to be counted as ridges. 

 Although these ridges and the three intervening enamel-folds, or 

 valleys, are much deeper than in the Malta form, yet it seems not 

 impossible that both types of teeth might be derived from a 

 common stock. 



Seeing, therefore, that the Maltese rodent does not belong to the 

 Myoxidce, while it presents certain resemblances to the Sciuridce, I 

 think it may be promionally assigned to the Sciuromorpha, although 

 it is quite probable it may constitute a family (Leithiidce) by itself. 

 As it requires a new generic title, the name Leithia, after its 

 describer, may be suggested ; and the genus may be provisionally 

 defined as specialized SciuroQiorpha with squared upper molars 

 bearing from four to five simple, low, parallel transverse ridges, 

 three of which rise from the postero-interual angle or inner side 

 of the crown, while the remainder are shorter aud are formed by 

 a single or double splitting of the more anterior of the two hind- 



* As pointed out to me by Mr. de Winton, the palatine foramina differ from 

 those of existing Sciuromorpha in extending backwards well into the maxillaj, 

 as in the Myomorpha, instead of being confined to the preniaxillaj. In Arctomys, , 

 however, these foramina intrude somewhat into the maxillse, so that the feature 

 does not seem of much value. Neither am I inclined to attribute very much 

 importance to the distal union of the tibia and fibula which takes place in the 

 fossil form, although such union is unknown in the living Sciuromorpha, while 

 it is universal in the Myomorpha. 



