iv.] RODENTIA. 39 



each other. Large single hypapophyses are developed from 

 the inferior surface of most of the cervical vertebrae in the 

 Shrews (Sorex) and some of their allies, and in Galeopithecus 

 each vertebra bears at its hinder end a pair of hypapophysial 

 tubercles. 



In the CHIROPTERA all the cervical vertebrae are broad, 

 very short from before backwards, with slender neural arches 

 from which (except in the axis) no distinct spinous processes 

 are developed. In certain forms (as Vesperngo) some of the 

 vertebrae have distinct double hypapophysial spines project- 

 ing backwards. 



In the RODENTIA the atlas has usually broad, moderately 

 long, wing-like transverse processes. The odontoid process 

 is long and slender ; the spinous process of the axis is much 

 developed, while as a rule that of the other cervical vertebrae 

 is exceedingly small. The transverse processes of the fifth 

 and sixth have large inferior lamellae ; that of the seventh 

 is sometimes perforated at the base (as in Lepus], and 

 sometimes imperforate (as in Hydrochczrus}. 



In the Capybara (Hydrocharus] and some others, the side 

 of the arch of the atlas is perforated near its anterior border 

 for the exit of the first spinal (sub-occipital) nerve, and also 

 near its hinder border for the second cervical nerve. 



In the Jerboas (Dipus) a very exceptional condition of 

 the cervical vertebrae occurs. The atlas is free, but all the 

 others are ankylosed together by both bodies and arches, 

 and the bodies are very wide and depressed, as in the 

 Armadillos. 



Among the UNGULATA, the atlas (Fig. 13) in the Pecora is 

 very long, with deep articular cavities for the occipital con- 

 dyles. The transverse processes are not wide, but much 

 extended from before backwards, and flattened from above 

 downwards. Each is perforated by a foramen (s;i) which gives 



