I reviewed some of my works on distribution studies in Microtus montebelli, Eothenomys smithi, Apodemus speciosus, and A. argenteus. (1) Tokuda (1941) in his famous book, ""Biogeography in Japan"", concluded that the occurrence of small field rodents, except A. argenteus, in each of the islands in Japan can be explained primarily by the geological time when the islands separated from each other. According to his conclusion, rodent fauna and its distribution in Shikoku must be the same as those in Honshu and Kyushu, because these three islands separated at the same time. M. montebelli, however, was not collected in cultivated fields in Shikoku, and A. speciosus was dominant in the fields. Because A. speciousus was captured after the removal of M. montebelli from a young tree plantation in Honshu, I thought that there would be a competitive relationship between M. montebelli and A. speciousus (Kaneko, 1979a). (2) The habitat of E. smithi has been described as wooded hills and mountains. But I collected it in cultivated fields at the foot of the Sanuki Mountains in Shikoku, where M. montebelli was absent as mentioned above (Kaneko, 1979b). Furthermore, E. smithi is distributed in cultivated fields in some localities of Honshu where M. montebelli is not collected (Kaneko, 1980a and unpublished data). I think that there would be a competition also between M. montebelli and E. smithi. From my unpublished data, however, E. smithi was absent in some localities at the foot of northern mountains in Shikoku, which are separeted from the Sanuki and Ishizuchi Mountains by alluvial plains (Fig. 5). At present I cannot explain the findings. Summing up some reports, the same distribution pattern of E. smithi is found in the hills of Boso, Miura, and Tama in Honshu, which are not contiguous to the Kanto Mountains: this species is absent in the former three localities, while it is present in the latter. (3) I showed that the occurrence of A. speciousus and A. argenteus in small islands of Honshu, Kyushu, Shikoku, and Hokkaido ...