13.5 11.4 4.5 19.2 11.0 10.0 4.3 Hsin-lung-shan, 65 mi. NE Peking, Jehol USNM 219229, ♂ 27.7 13.8 11.4 4.5 19.0 10.9 10.4 4.4 15 mi. S Lanchow, Kansu USNM 155171, ♂ 27.7 13.6 11.7 4.6 19.0 11.3 9.9 4.5

Comparisons.—From Apodemus peninsulae peninsulae (specimens from various localities in central Korea), A. p. sowerbyi differs in: External size smaller throughout, especially hind foot; upper parts, especially in summer pelage, and dorsal aspect of tail paler; skull smaller and less massive; braincase proportionally more inflated; rostrum shorter and noticeably down-curved. From Apodemus peninsulae praetor of Manchuria (holotype and paratypes), A. p. sowerbyi differs in most of the same ways in which it does from peninsulae as well as in having more shallow zygomatic notches, narrower zygomatic plates and smaller, more slender, upper incisors. From Apodemus peninsulae nigritalus of the Altai Mountains of Siberia (holotype and paratypes), A. p. sowerbyi differs in: Smaller size, both external and cranial; paler dorsal coloration; less convex cranial outline in lateral view; smaller auditory bullae.

Remarks.—Apodemus peninsulae sowerbyi is named in honor of the late Arthur de Carle Sowerby whose collections of mammals from North China and Manchuria have added so much to our meager knowledge of that part of the world.

Four specimens from Hsin-lung-shan, 65 mi. NE Peking, here assigned to sowerbyi, are darker dorsally than mice from farther to the west and in this respect may show approach to A. p. praetor. In all other features, however, they closely resemble the new subspecies.

All of the specimens of sowerbyi available to me are from altitudes of 3000 feet or higher. At lower elevations in North China, destruction of wooded habitats owing to intense land-use practices has probably restricted the distribution of sowerbyi primarily to hilly and mountainous areas where brushy, scrub and forest habitats still prevail.

Apodemus peninsulae, then, is known or suspected to occur over much of southeastern Siberia, Manchuria, Korea and North China. The western limits of its geographic range are unknown. Over this vast area only four subspecies, one newly named, can be ascribed with certainty to peninsulae whereas only two other kinds, giliacus of Thomas from Sakhalin and rufulus of Dukelsky from extreme southeastern Siberia are probably conspecific with it, the latter possibly a synonym of praetor. These considerations underscore the preliminary nature of the present paper. The mammalian fauna of northeastern Asia is scarcely better known today than was that of North America in 1885 when