EHIZOMYS. 315 



There are also other characters, such as the form and internal structure of the 

 stomach and csecnm, by which Rhizomys nearly approaches to Siphneus and Arvi- 

 cola, besides many strong resemblances to each other manifested by the skeletons of 

 the first two, and between their skuUs and that of Armcola. These considerations 

 which will be ampMed in the following sketch of the structural peculiarities of 

 Rhizomys would seem to determine that this genus has its nearest allies in Siphneus 

 and Arvicola, a fact which A. M.-Edwards^ has pointed out and on the strength of 

 which he has instituted a small natural division represented by Arvicola, of which 

 he regards Siphneus as a derivative type, modified by the external circumstances in 

 which it is destined to live. 



The leading feature of the vertebral column of Rhizomys, as in Siphneus, is, 

 the more than usual breadth and strength of the cervical vertebrae, developed doubt- 

 less in relation to the heavy head and burrowing habits of the animal, but differing 

 from that genus in that there is no amalgamation of any of the cervical vertebrae. 

 The atlas has its spinous process reduced to two nodules, but in the axis this process 

 is massive, as in Siphneus, having considerable antero-posterior expansion. Erom the 

 third to the seventh cervical vertebra, Rhizomys, unhke Spalax, Georychus, A^^vicola, 

 and Rathyergus, has the spinous processes well developed and gradually increasing 

 in size from before backwards. The transverse process of the atlas, contrary to that 

 of Siphneus i is flattened and of considerable extent, as in Rathyergus ; its lower or 

 anterior aspect being rather deeply concave, its lower division being prolonged down- 

 wards as a well-marked ridge on to the inferior arch of the vertebra which bears a 

 nodular hypapophysis, as in Siphneus, much more reduced than that which occurs in 

 rats. Erom the axis to the sixth cervical, the two divisions of the transverse process 

 increase in size from before backwards, the lower portion being placed somewhat 

 anterior to the one above it, so that the artery is more protected than if the processes 

 lay directly over each other. The extremity of the sixth process, as in Siphneus, is 

 widely bifurcate, while only the upper division of the seventh process is developed. 

 The spinous process of the first dorsal little exceeds in vertical extent that of the 

 seventh cervical, but this process rapidly increases in length to the fourth dorsal, 

 where it attains a length of half an inch ; and as far as the twelfth dorsal, it is much 

 du'ected backwards, diminishing in height, but gaining in antero-posterior extension 

 to the penultimate lumbar. 



Metapophyses begin to show themselves on the twelfth dorsal, and go on developing 

 in size as they are traced backwards to the last lumbar, being also well developed on 

 the first sacral and traceable tlu^oughout the sacral elements. Anapophyses begin on 

 the thu'teenth dorsal, and attain their maximum size on the second lumbar, disappear- 

 ing on the last lumbar. The transverse process of the first lumbar vertebra exists 

 only as a rudiment, but this process increases in size to the last lumbar, but it does 

 not attain the same proportional development that it assumes in some Muridce. The 

 bodies of all these vertebrae are rounded on their under surfaces, and do not show 

 any of the ridge-like hypapophyses which characterise the murine skeleton; no 



* Eech. des Mammif, p. 78. 



