We are obliged, indeed, to give up the attempt to keep to a line of communicating rivers throughout the whole 24 days. Nor do I see how it is possible to adhere to that condition literally without taking more material liberties with the text. [Illustration: MARCO POLO’S ITINERARIES Nᵒ. V. Indo Chinese Regions (Book II, Chapˢ. 44–59)] My theory of Polo’s actual journey would be that he returned from Yun-nan fu to Ch’êng-tu fu through some part of the province of Kwei-chau, perhaps only its western extremity, but that he spoke of Caugigu, and probably of Anin, as he did of Bangala, from report only. And, in recapitulation, I would identify provisionally the localities spoken of in this difficult itinerary as follows: _Caugigu_ with Kiang Hung; _Anin_ with Homi; _Coloman_ with the country about Wei-ning in Western Kwei-chau; _Fungul_ or Sinugul with Siu-chau. [This itinerary is difficult, as Sir Henry Yule says. It takes Marco Polo 24 days to go from Coloman or Toloman to Ch’êng-tu. The land route is 22 days from Yun-nan fu to Swi-fu, _viâ_ Tung-ch’wan and Chao-t’ung. (_J. China B. R. A. S._ XXVIII. 74–75.) From the Toloman province, which I place about Lin-ngan and Cheng-kiang, south of Yun-nan fu, Polo must have passed a second time through this city, which is indeed at the end of all the routes of this part of South-Western China. He might go back to Sze-ch’wan by the western route, _viâ_ Tung-ch’wan and Chao-t’ung to Swi-fu, or, by the eastern, easier and shorter route by Siuen-wei chau, crossing a corner of the Kwei-chau province (Wei-ning), and passing by Yun-ning hien to the Kiang; this is the route followed by Mr. A. Hosie in 1883 and by Mr. F. S. A. Bourne in 1885, and with great likelihood by Marco Polo; he may have taken the Yun-ning River to the district city of Na-ch’i hien, which lies on the right bank both of this river and of the Kiang; the Kiang up to Swi-fu and thence to Ch’êng-tu. I do not attempt to explain the difficulty about Fungul. I fully agree with Sir H. Yule when he says that Polo spoke of Caugigu and of Bangala, probably of Anin, from report only. However, I believe that Caugigu is the _Kiao-Chi kwé_ of the Chinese, that Ani_n_ must be read Ani_u_, that Aniu is but a transcription of _Nan-yué_, that both Nan-yué and Kiao-Chi represent Northern Annam, _i.e._ the portion of Annam which we call Tung-king. Regarding the tattooed inhabitants of Caugigu, let it be remembered that tattooing existed in Annam till it was prohibited by the Chinese during the occupation of Tung-king at the beginning of the 15th century.—H. C.] NOTE 7.—Here the traveller gets back to the road-bifurcation near Juju, _i.e._ Chochau (_ante_ p. 11), and thence commences to travel southward.