1893.] DR. C. J. FOItSYTH alAJOH UJN" MlOCJiNB SQUIRllELS. 195 



inferior iu number to the non-flying Sciuroniorpha, and have, to all 

 appearances, greatly diminished from Tertiary times up to the 

 present, so that their special means of locomotion do not seem to 

 have proved of moi'e value in the struggle for existence to those 

 of the nou-fljnng. 



(3) One may thirdly suppose that there is no direct connexion 

 whatever between the Sciurojderi and recent tiviuri or iSdurlda'. 

 From Tertiary times up to the present, the species of riying- 

 ►Scjuirrels have been gradually diminishing iu number, their 

 characters having proved inadaptive, whilst the sjiecies of iSciuridse 

 have been increasing. The points of similarity in the grinding- 

 teeth of Sciurus prevosti and tSciuropternsJimhriatus, on the whole, 

 are very slight (and so are those between Sciaropterus volucella 

 and Sciunis hufhonms) ; with a httle practice it is at once possible 

 to distinguish an isolated tooth of the one from that of the 

 other. Their skulls, moreo\'er, ai'e veiy different. 



For my part, I rather incline towards the third supposition, 

 although admitting that the grounds on which it is based may 

 not be convincing. At any rate, the characters of tlie cranium as 

 well as those of the dentition, though greatly varying, gi\e on the 

 whole a family likeness to all the Flying-Squirrels, so that 1 cannot 

 but separate them into a cUstinct subfamily from the Sciurinse. 



The Pteromys, sensu strictisshno, have probably evolved from a 

 single Sciaroj>tems-\\\.Q form, and Eupetaunis is apparently the 

 more specialized descendant of some Pteromys. 



The really important characters in which some Sciari and 

 Spenno/iMH approach the Sciuropteri, as in the restricted 

 iuterorbital region of Culohotis, the general elongation of the 

 frontals of various Sciuridfe, the general shape of the grinding- 

 teeth in Eosciurus, and in the stronger development of the first 

 ridge in the upper molars of the Otospenao/ihiU, are all such that 

 they may be considered as ancient inheritances. 



Therefore we need not admit any recent connexion between the 

 Sciuropteri and the above-mentioned members of Sciurida;. 



Anatomical characters and palseontological evidence point in the 

 same direction, viz. that the Sc'mropteri are the Httle modified 

 remnants of a very old and once widely spread group. There is 

 not sufficient evidence for admitting that they have evolved from 

 forms of non-flying Squirrels identical with, or very closely allied 

 to, those actually living ; their power of flying may not be a com- 

 paratively recent character. They are specialized, no doubt, 

 compared to the remnant of Sciuridse ; but the ancestral non-flying 

 types may neither have been Sciuridse nor even Sciuromorpha. 



It would be more consistent with these vieMS to place the Flying- 

 Squirrels in a distinct family ; but for this it will be time enough 

 when their recent as well as their fossil forms are better known 

 than is the case at present \ 



^ The present paper was compfetelj fiuished when I fii'st became partially 

 acquainted with H. Winge's " Jordfundne og nulevende (3-navere (Eodentia) 



13* 



