52 MR. OLDFIELD THOMAS ON MAMMALS [Feb. 4, 



5. Felis microtis M.-Eclw. 



$ . 1198, 1278. Sasuna, N. Island. 300'. 

 (91.10.14.1, 92.1.11.3, and 96.2.28.1. P. A. Hoist. 1891.) 

 This is an essentially Korean member of the Tsu-shima fauna, 

 no» member of the genus Felis being found in Japan. 



" The Wild Cat is common in Tsu-shima, living near the sea- 

 shore, where it is said to go nightly at low tide to fish. In the 

 stomach of 1278 I found fish-bones and some feathers of the 

 pheasant. The natives consider the flesh of the Cat especially 

 delicate food. Native name : ' Yama-niku ' = Mountain-Cat." — 

 M. P. A. 



6. Martes melampus tsuensis Thos. 



c? . 1261-1277. Sasuna, Tsu-shima. 

 (91.10.14.2-3, 92.1.11.1-2. P. A. Hoist. 1891.) 

 When describing this form in 1897* I had not any Japanese 

 Martens of the hedfordi type to compare it with, that being clearly 

 the animal to which it is most allied, and not the typical melampus. 

 JSTow, however, a comjjarison of Mr. Anderson's skins of the two 

 shows that, in winter pelage, the Tsvi-shima Marten is distin- 

 guishable from that of Southern Hondo by its whitish crown 

 and the absence of the yellowish tuft at the end of the tail. 



" Common in Northern Tsa-shima, and occurring in the South 

 Island. Like the Cat it is most often found near the sea. The 

 stomachs of the two caught were empty, but the animals were 

 excessively fat. They are considered a fine food by the peasantry 

 of the islands, who call them ' Wata-boshi' = Cotton-cap." — 

 M.P.A. 



premolars and only f incisors being evidently wrong, as a comparison of the milk 

 and permanent teeth both of Di/mecodon and UrotricJms shows that the ante- 

 penultimate premolar does not change, and is therefore p^, while distinct traces of 

 the premaxillo-maxillary suture can be seen just behind the third tooth in the upper 

 jaw. Of the incisors it is probably ii that is missing in the lower jaw. 



In the form and relative proportions of the permanent teeth Dymecodon resembles 

 TTrotriohus very closelj', the curious alternation in size of the lower teeth, on which 

 Mr. True based the name of the genus, not existing in the later dentition, and it is 

 in fact only due to the milk pj being, not unnaturally, smaller than the permanent 

 Pi which coexists with it. 



But in the milk stages, which have not previously been compared, the differences 

 are far greater. For while in Dymecodon the two anterior incisors are subequal, 

 similar in form, scarcely overtop the teeth posterior to them, and in Mr. True's 

 words " resemble the teeth of Fhoccena," those of Urotrichus have already pro- 

 gressed a long way towards the specialisation found in the adults of both genera, 

 rail considerably surpassing mi-, and this again being much longer than mi^ ; and 

 below again the most anterior tooth (mio, as I suppose) is decidedly longer than any 

 of the next three teeth. 



Prom this it would appear that Dymecodon is in a more primitive stage of 

 evolution than Urotrichus. 



Mr. Ivanai's specimens of Dymecodon were captured at Shibu, onMt. Yatsugatake, 

 Nagano Ken, N.W. of Yokohama, at an altitude of 5700'. 



Mr. Kanai also obtained at Kamisuwa-machi, in the same Ken, an example of 

 VesfertUio murinus stiperans Thos., a Bat new to the fauna of Japan. 



* Ann. Mag. N. H. (6) xix. p. 161 (1897). / 



