National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

related in his 嶺表錄異 Ling-piao lu-i (c. 900 A. D.; the passage is quoted in T'ai-p'ing kuang-
chi, 483, 1; in the Wu-ying-tien ed. of the Ling-piao lu-i). Travelling back in a boat from Shan-
tung (?) to Fu-chien, Chou Yü and his companions were blown five days by a gale, and passed six
kingdoms, the first of which was the « Kingdom of Dogs » (Kou-kuo), as they were told by a Hsin-lo
(Corean) man on board : they actually saw « something like men » (ju jên), naked, and carrying
dogs in their arms. Afterwards they passed the « Kingdom of Hairy Men » (Mao-jên kuo), small
sized, and covered with hair like mastiffs. They then arrived at the « Kingdom of Yeh-ch'a »
(Yakṣa, but here, as often, with the value of Rākṣasa, « Devils »), where their boat was stranded
and they had a narrow escape at the hands of the inhabitants. The next kingdom was the « Kingdom
of Giants » (Ta-jên kuo); then the kingdom of 流虯 Liu-ch'iu (another form of the more usual
琉球 Liu-ch'iu, probably here Formosa), the language of which was partly understood by the Corean
passenger, and where they were asked iron in exchange for food; and finally the « Kingdom of Dwarfs »
(Hsiao-jên kuo), in which the naked inhabitants were of the size of children five or six years old.
They passed a last island in which they caught wild sheep, but it was uninhabited. Although Liu
Hsün commends Chou Yü as being a vegetarian, it seems difficult to give an unqualified credence
to what looks in the main like a sailor's yarn. With the exception of Liu-ch'iu, all the names belong
to the stock-in-trade of fantastic geography. But this is the only mention I remember of a Kingdom
of Dogs in the Eastern Sea. Moreover, it has the special feature of not being, as usual, a kingdom
in which people were dogs or looked like dogs, since they merely carried dogs in their arms.
There are many tales of dog or wolf ancestry in southern China with the Man descendants of
the dog 槃瓠 P'an-hu, as well as in Turco-Mongolian traditions where the wolf was a sort of totem-
ancestor (cf. TP, 1931, 463-470). LAUFER (in Y, III, 110) has insisted that « a clear distinction must
be made between dog-headed people and the motive of descent from a dog ancestor, — two entirely
different conceptions ». This is true, but only to a certain point, and the two traditions are some-
times combined. I brought back from Tun-huang the Tibetan translation of the report made to
the Uighur King by five envoys he had sent to inquire about his neighbours; the original text may
be dated about 800 A. D. We hear there of two dogs, one red and one black, ancestors of the tribes
Ge-zir Gu-šu and Ga-ra Gu-šu, who carried off a Dru-gu (= Turkish) girl who had children by
them, the males being dogs, and the females, women (cf. R. STEIN, in TP, 1939, 24). Ge-zir renders
Turk. qïzïl, « red », and Ga-ra, Turk. qara, « black »; the original of Gu-šu (? Turk. qoš) is still
doubtful.
The tale of the tribes of South China which were the descendants of the dog P'an-hu cannot
be traced back earlier than the time of the Later Han, at the beginning of our era. But the connec-
tion of « Barbarians » with dogs is much more ancient; it underlies the use of the « dog » radical
for writing the name of so many aboriginal tribes in the west and south and is more precisely
expressed in the old term 犬戎 Ch'üan-jung, « Dog Barbarians » (cf. DE GROOT, Religious System
of China, IV, 263-269). But such names may generally be due merely to depreciatory comparisons.
The popular belief in a real Kingdom of Dogs is attested, however, in the pseudo-historical 晏子
春秋 Yen-tzŭ ch'un-ch'iu, which cannot be dated later than the first Han dynasty. Yen-tzŭ,
or 晏嬰 Yen Ying, was a minister of Ch'i, who died in 500 B. C. (not 493 as in MAYERS and GILES);
cf. Shih chi, 62, 2; GILES, Biogr. Dict., No. 2483; MARGOULIÈES, Le Kou-wen chinois, 180-183; on