Red-backed Voles of Japan— JAMESON 
595 
as E. smithii.” Lastly, "Anderson’s E. niigatae 
also has been founded upon an adolescent 
animal. . . 
Apparently Hinton attached no importance 
to the occurrence of immature smithi on Kyu- 
shu, Shikoku, and Honshu, and the absence of 
adult smithi everywhere but on Hokkaido. It 
is surprising that he did not consider that these 
forms with rootless molars might actually be 
adults, for on the mainland of Asia and on Tai- 
wan are genera of red-backed voles in which the 
molars are rootless in the adults. Hinton recog- 
nized Eothenomys, Anteliomys, and Aschizomys , 
genera in which the molars are rootless; but 
failed to suggest any similarities between 
smithi and the species of these genera. 
The classification of these voles by Japanese 
mammalogists differs considerably from the re- 
cent presentation of Ellerman (1941) and of 
Ellerman and Morrison-Scott (1951). In the 
latter work the authors were apparently un- 
aware of the very careful study on Japanese 
murid rodents by Tokuda ( 1941 ) . 
In his monograph, Tokuda ( 1941 ) followed 
Oldfield Thomas in general in the classification 
of the Japanese red-backed voles. He consid- 
ered smithi as a separate species in the genus 
Clethrionomys, bedfordiae was retained as a sub- 
species of Clethrionomys rujocanus, and ander- 
soni was judged to be a distinct species but 
closely allied to rujocanus . Tokuda (1941: 51) 
followed Hanaoka in placing Craseomys niiga- 
tae as a synonym of Clethrionomys andersoni. 
Tokuda did not suggest at this time that smithi 
had affinities except in the genus Clethri- 
onomys. 
The first to suggest a different position for 
smithi was Imaizumi (1949), who placed it 
in the genus Eothenomys . Subsequently, Tokuda 
(1955) presented evidence for considering 
smithi a species of Anteliomys. Imaizumi 
(1957), considering Anteliomys a synonym of 
Eothenomys , continued to use the latter name, 
and described a second species, Eothenomys 
kageus, from the northern part of Honshu. 
The most recent classification of the red- 
backed voles in Japan is by Imaizumi (I960). 
In this arrangement there are Eothenomys: 
smithi and kageus; and three species of Cle- 
thrionomys: rutilus, rujocanus, and sikotanensis; 
in contrast to Tokuda, Hinton, and other recent 
students, Imaizumi maintained niigatae as a 
species distinct from andersoni, and placed both 
in Aschizomys. 
During 1952 I spent 10 months in Japan 
and made a small collection of voles. In 1958 
and 1959 I was able to trap additional speci- 
mens of all but one ( Clethrionomys sikotanen- 
sis) of the forms of red-backed voles in Japan, 
and from these specimens and the excellent dis- 
cussions of Tokuda and Imaizumi, it is appar- 
ent that the original descriptions of Oldfield 
Thomas and the recent researches of Tokuda 
and Imaizumi more accurately express the rela- 
tionships of these voles. The work of Hinton 
was executed apparently with little knowledge 
of the geologic history of the Japanese archi- 
pelago and without the advantage of having 
studied these Japanese voles in the field. The 
arrangement given below is, in most respects, 
like that accepted by Japanese mammalogists 
today. Because of the stature of Hinton’s mono- 
graph, his conclusions have been followed by 
European and American zoologists, and reiter- 
ated, in regard to the rodents in question, by 
Ellerman (1941) and Ellerman and Morrison- 
Scott (1951). The environmental distributions 
and economic importance of these voles is dis- 
cussed by Ota and Jameson (in press). 
GENUS Clethrionomys Tilesius 
Originally all the red-backed voles in Japan 
were placed in Clethrionomys. The Japanese C. 
rutilus mikado exists only on the island of Hok- 
kaido. In general appearance and in most details 
rutilus closely resembles the other species of 
Clethrionomys s. str. The skull is light with 
rounded contours, the palate is abnormal (dif- 
fering from Microtus ) , the cheek teeth are rel- 
atively light and are rooted in the adult. The 
mammae are in four pairs. The species in this 
genus are characterized by other features but 
the above are sufficient to separate rutilus from 
the other species of microtine rodents in Japan. 
The species rujocanus has long been known 
as a rather aberrant species of Clethrionomys. 
Miller (1898^: 360) treated in some detail the 
distinctive characters of this vole: 
So divergent is the animal that it may well be 
questioned whether it is to be regarded as a true 
