given to « every family of the twelve (? read ' fifteen ') relays of the dog-post (IN kou-ti ; YS,
18, 7 a) ». On August 19, 1320, « the myriarchy (wan-hu-fu) of the Nü-chih (= Nü-chên) and the t'o-t'o-ho-sun (' director ') of the dog-relays were abolished » ( YS, 27, 4 a). On November 10, 1330, «since in the circuit (lu) of the Shui Ta-ta (' Water Tatar ') of the hsing-shêng of Liao-yang rain had been falling heavily since the preceding summer, the two rivers Hei-lung (Amur) and Sung-wa (Sungari) had been flooded and the people had had no fish for their food, and it had come to the point that many of the dogs of the fifteen dog-relays of Mo-lu-sun [etc.] had died of starvation ; grain [sufficient] for two months was distributed to [the population] and paper-money was given to replace by purchase the dogs which had died » ( YS, 34, 9 b ; cf. SHIRATORI, Beiträge, II, 347, 396). According to the YS, 101, 1 a, 4 b, the fifteen dog-relays originally comprised 300 families (hu) with 3.000 dogs ; but the numbers had gradually dwindled to 289 families and 218 dogs. The above texts give a clear view of the difficulties which beset the deportation of criminals to Nu-êrh-kan ; and this explains why new regulations had been adopted in the spring of 1320 : only the criminals guilty of a grave offence were henceforth banished to Nu-êrh-kan (but we know of two cases of banishment to Nu-êrh-kan in 1321; cf. YS, 27, 7 a and b) ; those who were sentenced to banishment for a slight offence were to remain at the agricultural settlement (t'un-t'ien) of Chao-chou (Yuan tien chang, hsin-chi, sect. hsing-pu, 1 a-b; cf. SHIRATORI, Beiträge, II, 335). After the fall of the Yüan dynasty, Nu-êrh-kan was again occupied by the Ming dynasty. As early as in 1403 Yung-lo had sent envoys to Nu-êrh-kan (Shu-yii chou-tzû lu, 24, 1 a), and a Commis-
sariat (js to-ssi) was established there in 1409 ; its centre was on the Amur, one stage to the « east »
of the station Man-ching (near the town of Tyr) where the inscriptions of 1413 and 1433, known as the « Tyr inscriptions » or « inscriptions of the Yung-ning-ssü », were discovered (on these inscriptions, cf. PoPov, in ZVOIRAO, xvi [1906], 012-020, 077, and pl. II ; SHIRATORI, Beiträge, II, 563 ; from personal examination of the original monuments, now in Vladivostok, I can say that PoPov's decipherment of the Chinese text can be corrected and completed in many cases ; cf. also IKEUCHI Hiroshi in Man-Sen jiri ... hôk(5, Iv, 321 ; the first of these inscriptions is trilingual, in Chinese, Mongolian, and Jucen ; the Mongolian text has not been published, but was translated by POZDN EV in Lekcii po istorii Mongol'skoi literatury, III, 70-75 ; as far as I know the Jucen text is still unpublished and untranslated).
I have little doubt that Nu-êrh-kan, the « Botany Bay » of the Mongol dynasty, located in a dreadful spot at the extreme north-eastern end of the Empire, is the « desert island » spoken of by Polo. Without being an island, it is in fact a peninsula, between the lower Amur and the sea, and was quite l robably called an « island » in Persian-speaking circles (for analogous extensions of the notion of « island » in Polo, see « Çanghibar » and « Mogedaxo »).
But if the « desert island » be the region of Nu-êrh-kan, we have to account for the name which Polo gives to it. The ancient original form which Nu-êrh-kan transcribes is not known with certainty. I thought at one time that the name of Nurhaci >> Nurhaci (j- 1626), the founder of the Manchu dynasty which in 1644 became that of the « Great Ch'ing », might be an ethnical form derived from Nu-êrh-kan. But for Nu-êrh-kan the chances are in favour of an original form with palatal vowels. Although POZDN EV, in his translation of the «Tyr» inscription of 1613, transcribes the name once as « Nurban », at other times he gives « Nurgan » in Russian, and even, in one passage
IIIF Color HighRes
