CHIMAEEOGALE. 149 



tipped teeth and a differently clad tail, so that it is impossible to regard them as 

 generically identical. An examination of the skull proves that although by the 

 number and white character of its teeth it is allied to Crocidura, it is nevertheless 

 remarkably distinct from these non-aquatic shrews. The skull, like that of Crosso- 

 pus and Soriculus, is light and not marked by the strong ridges which characterise 

 the skull of Crocidura. The brain-case is relatively broader than in Crocidura, 

 and, like the skulls of Crossopus and Soriculus, and especially Nectogale, has a 

 much shorter temporal fossa than Crocidura. The occipital region, as in Crosso- 

 pus, Soriculus, and Nectogale, is considerably uptilted, and the frontal region, as in 

 the two first, is considerably depressed. The marked double condyle is a character 

 in which it is resembled to a certain extent by Crocidura, in which, however, the 

 separation of the condyle into two facets is carried to only a very limited extent, 

 whereas in Soriculus and Anurosorex it is as strongly developed as in this water- 

 shrew, and it seems also that this remarkable modification is quite as well marked in 

 Nectogale. The foramen ovale (?) occupies the same position in this shrew as in 

 Soriculus. It is separated from Crossopus by the character of its teeth, which in 

 their colour and number resemble those of Nectogale, except that the inner side of 

 each of the first upper incisors is furnished with a talon. Erom Crossopus it is also 

 separated by the peculiar character of its tail, which in Crossopus is almost mouse- 

 Hke, whilst in the shrew it is densely clad with flattened, coarse and rigid hairs. 

 The ciliation of the feet also of this form approaches more to what occurs in Necto- 

 gale than in Crossopus. From Nectogale it is distinguished by the different 

 arrangement of the hair on the tail and by its unwebbed hind feet and longer 

 snout ; and as it is thus impossible to classify this and the Japan water-shrew 

 with any existing genus, I have created the genus Chimarrogale^ for their 

 reception. 



In its general form and structure it is more closely allied to Nectogale than to 

 any other genus of Insectivora, and it thus appears to me that these two genera, 

 Nectogale and CMmarrogale, are entitled to rank as a distinct sub-Eamily, 

 NectogalincB. 



In connection with these observations on the structure of this remarkable water- 

 shrew, I embrace the opportunity of here describing and figuring on the same plate 

 with CMmarrogale, by way of comparison, an equally remarkable, but terrestrial, 

 bmTOwing shrew exhibiting certain affinities with Talpa, from the neighbouring 

 region of Northern Assam, which has a fauna intimately allied to that of Yunnan, 

 viz.i Anurosorex assamensis, Andr. 



* X ijxappos (mountain-torrent), and y«A^ (Weasel). 



