OBSERVATIONS ON THE RODENTIA. 9T 



There are, however, certain points in which the greater por- 

 tion of the species agree, — these, which may be called the 

 typical characters of the group, in my opinion should not be 

 selected, with the departures from such typical characters, 

 for the establishment of primary divisions. The skulls of 

 rodents, for instance, generally possess a large glenoid cavity, 

 longitudinal in its direction ; but in the genus Lepus this ca- 

 vity is remarkably small and narrow : — thus rodents in gene- 

 ral might be arranged in one family, and the genus Lepus in 

 another. 



Or if we take the perfect or imperfect state of the clavicles 

 as our guide, the present order would also be divided into 

 two families or sections. Rodents are typically claviculated 

 animals, those with imperfect clavicles being exceptions in 

 this respect to that structure which is most commonly found 

 in the species. The claviculated or unclaviculated groups 

 cannot, I think, be of equal value. 



My principles will be more fully illustrated by the mode in 

 which I have proceeded in the investigation of the affinities 

 of the animals under consideration. 



After a careful comparison of part with part throughout 

 the whole series of rodent skulls which I have had an oppor- 

 tunity of examining, it appeared to me that the zygomatic 

 arch and ant-orbital foramen afforded the most constant cha- 

 racters. In the Sciuridce, Muridae, and Arvicolidae for in- 

 stance, we find the zygomatic process of the maxillary bone to 

 consist of a large thin plate of bone, which is oblique in its 

 position, and has the lower edge emarginated so as to throw 

 the anterior portion of the zygomatic arch above the plane of 

 the palate : here the anterior outlet of the sub-ovbital foramen 

 is much contracted. On the other hand in the Hystrices, the 

 genera Echimys, Myopotamus, Dasyprocta, and in the Chin- 

 chillida and Caviida, the ant-orbital foramen is very large, 

 enclosed by two meeting branches from the maxillary bone, 

 the lower branch being thrown out from the level of the palate. 



In the hares and rabbits (Lepus), taking the same charac- 

 ter, we find quite a new type of form, indeed these animals 

 appear to be in many respects isolated. In the very imperfect 

 state of the palate, however, and in some other characters 

 which will be hereafter mentioned, there appears to be an ap- 

 proach in the Chinchillida and Caviida. 



Thus we have nearly all the principal genera of rodents 

 thrown into three great sections, which may be easily distin- 

 guished ; and upon arranging the various species of the ge- 

 nera or families above mentioned in these sections, we find 



