隠岐島

The Islands of Oki lie in the Japan Sea about 30-40 miles off Matsue, situated at the northern coast of south-west Hondo, the main island of Japan. The depth of the sea between the both is under 200 m. The geologists think that Oki Islands had connected to Hondo by the Shimane Peninsula up to diluvial epoch. Dogo Island, the largest of the group, have an area of 216 km2; predominating there the mountainous landscape with scanty alluvial plains. The vegetation of this island shows an appearance of the climax of deciduous forest consisted of chiefly deciduous trees such as oaks. Dozen Islands are the general name for the group of islands which situate at some distance from Dogo. The topography of Dozen is generally flat, and simpler than that of Dogo. Everywhere of these islands are artificially cultivated, and the secondary vegetation mainly covers all the islands. No large-sized endemic mammal is known in Oki Islands, both in Dogo and Dozen ; Lepus brachyurus okiensis from Dogo being the largest one. As to the other mammals, Mogera wogura kobeae, Urotrichus talpoides minutus, Crocidura dsi-nezumi, Apodemus speciosus navigator, Apodemus geisha celatus, Anteliomys smithii okiensis are recorded from Dogo. Moreover, it is sure that Mustela sp. and Glirulus sp. (both may be characteristic subspecies) inhabit in Dogo, though they are not yet examined taxonomically. It is worthy of notice that the majority of endemic mammals from Dogo Island are distinguished from allied species of Hondo as subspecies by the previous systematic studies. To work on the interesting problem of subspeciation, we tried to make clearer the characteristics shown by the former workers as subspecies. We classified by ages all specimens of both Oki and Hondo respectively, and statistically compared both samples in each division. In this report, Apodemus speciosus speciosus of Hondo and Apodemus speciosus navigator from Oki were mainly chosen as the materials for the purpose above mentioned, and we touched a little to the other mammals. in short, we could point out some characteristics to be distinguished from typical subspecies of Hondo by statistical treatment in the specimens of Oki of Apodemus speciosus navigator (I) together with Apodemus geisha celatus (II), Anteliomys smithii okiensis (III) and Urotrichus talpoides minutus (IV). For the external characteristics, that both subspecies (I) and (II) of Oki have shorter tail in comparison with body length than that of typical subspecies of Hondo, that subspecies (IV) have longer tail on the contrary, that there is no recognizable external distinction in subspecies (III) (but being distinguished by the characteristics in the skull) and others became clear. We compared Apodemus speciosus navigator with Apodemus speciosus speciosus in detail basing on the abundant materials. Ages were identified by the pattern of grinding surface of m3. We divided all samples into five classes ; namely stage 1 (juvenile), st. 2 (subadult), st. 3 (adult I), st. 4 (adult II) and st. 5 (senile) by age, and compared both groups in each class of age respectively. By such comparisons, it became clear, in addition to lower tail ratio of navigator than that of speciosus, that on distance between the first upper molars (m1～m1) and length of upper molar series (m1～m3), the former from Dogo was somewhat longer than that of the later, and specimens from Dozen remarkably longer than speciosus. The portion represented by m1～m1 and m1～m3 grows most slowly in the skull. The skull grows anteriorily and posteriorily from that dead centre ; the anterior extension being greater than posterior extension. Judging from the fact that in the forest mouse from Oki this portion is longer than that of Hondo, especially in the animal from Dozen remarkably, we think that the animal from Dozen may be distinguished also from navigator of Dogo as another subspecies. However, we did not deal with taxonomical consideration of the forest mouse from Dozen, because of the insufficiency of materials. To lay the foundation on the studies of subspeciation of the small mammals from Oki Islands, the present investigation was planned to make more exact taxonomical treatment of the materials. We think that such a new trial as the classification by ages and the statistical comparison in each class was an excellent method for the purpose.