136. CHEMEINFU 239
election, the princes of the old Mongol party elected his brother Arïq-bögä in a diet held near the *Altan river, west of Qara-qorum, towards the beginning of June 1260. On August 18, 1260, Qubilai left K'ai-p'ing-fu and marched against Arïq-bögä; while staying at *JoIdürgä, Qubilai solemnly denounced Arïq-bögä as a traitor (October 8, 1260). Finally, leaving the leadership of the campaign to his lieutenants, Qubilai, coming back from Qara-qorum, « stopped
temporarily » ( chu-pi) at Yen-thing, i. e. Peking, on January 14, 1261; he took the field
again in the course of the year. It seems that this « temporary stay » marks actually the change of the capital. From that time, the central Imperial administration was established in Peking, although this former capital of the Chin and its Imperial palace were then partly in a state of ruin as a result of the Mongol conquest of 1215 (see « Cambaluc »). Qubilai always spent part of the year at K'ai-p'ing-fu, a greater part so long as his new palace in Peking was not completed; but K'ai-p'ing-fu was never his administrative capital, no more than Windsor was of England, or even Versailles of France.
The YS, 58, 3 a, says that the name of K'ai-p'ing-fu was only given to Qubilai's summer residence in 1260, and that would leave us in the dark as to its name between 1256 and 1260; but this is an error, as has been shown by WANG Hui-tsu2, 8, 1 b (to which other mentions of K'ai-p'ing-fu prior to 1260 may be added, as YS, 4, 2 b, at the end of 1259). The name must have been K'ai-p'ing-fu already in 1256.
It is generally said that the name of K'ai-p'ing-fu was changed to Shang-tu (see « Ciandu ») in 1264; this is due to a misinterpretation of a sentence in YS, 58, 1 b, and to an error in YS, 58, 3 a, where the change is attributed to the 5th year chung-t'ung, i. e. to January-August 1264. But K'ai-p'ing-fu had become Shang-tu before then, on June 16, 1263 in fact (cf. YS, 5, 6 a, and the k'ao-chêng at the end of the ch. ; also WANG Hui-tsu2, 8, 1 b).
The Mongol tradition (cf. SCHMIDT, Gesch. der Ost-Mongolen, 113, 137) speaks of K'aip'ing-fu under the name of Sangdu-Käibüng-Kürtü-balyasun, or simply 5angdu-Käibüng. Sangdu is of course Shang-tu, and Käibüng is K'ai-p'ing (on the pronunciation of p'ing with ü, see « Pianfu »). As to Kürtü-balyasun, it is given in Ta-Ch'ing i-t'ung chih (409, II, 2 a) as the local name of the ,f 4.11 Huan-chou of the Yüan, a little to the south-west of K'ai-p'ing-fu; the Mongol chronicler thus seems to have connected wrongly that third name with K'ai-p'ing-fu. SHIRATORI, followed by YANAI, 633, has explained kürtii by Mong. qoriya, qor) an, etc. ; this is impossible, owing to the difference of « class ». The kürdü, « circle », « wheel », of KLAPROTH and of BLOCHET (BI, II, 387, 462, and App. 42) would be preferable phonetically, but does not make much sense. The name might represent kiirtü -_ kä'ürtü, [« place] with corpses »; then Kürtü-balyasun would mean « City of the Dead »; but in such a case, we should expect *Kä'ürtü in the Mongol chronicle. For the geographical identification of K'ai-p'ing-fu, see « Ciandu ».
It is somewhat surprising that in the narrative of the voyage, Polo should speak of « Chemeinfu », and of « Ciandu » in his description of the country and of its customs, without stating anywhere that both names are equivalent. But nowhere does he say either that his Sea of Sarai (see « Saray ») and his Sea of « Gel or Chelan » (see « Gel or Chelan ») are the same as his Sea of Baku (see « Bachu »). Perhaps, although he knew the identical value of these various names, he failed to explain them to Rustichello, who naturally could not go beyond
IIIFカラー高解像度
