In his own bearings, at least on land journeys, he usually carries us along a great general traverse line, without much caring about small changes of direction. Thus on the great outward journey from the frontier of Persia to that of China the line runs almost continuously “_entre Levant et Grec_” or E.N.E. In his journey from Cambaluc or Peking to Mien or Burma, it is always _Ponent_ or W.; and in that from Peking to Zayton in Fo-kien, the port of embarkation for India, it is _Sceloc_ or S.E. The line of bearings in which he deviates most widely from truth is that of the cities on the Arabian Coast from Aden to Hormuz, which he makes to run steadily _vers Maistre_ or N.W., a conception which it has not been very easy to realise on the map.[9] [Sidenote: Singular omissions of Polo in regard to China; Historical inaccuracies.] 71. In the early part of the Book we are told that Marco acquired several of the languages current in the Mongol Empire, and no less than four written characters. We have discussed what these are likely to have been (i. pp. 28–29), and have given a decided opinion that Chinese was not one of them. Besides intrinsic improbability, and positive indications of Marco’s ignorance of Chinese, in no respect is his book so defective as in regard to Chinese manners and peculiarities. The Great Wall is never mentioned, though we have shown reason for believing that it was in his mind when one passage of his book was dictated.[10] The use of Tea, though he travelled through the Tea districts of Fo-kien, is never mentioned;[11] the compressed feet of the women and the employment of the fishing cormorant (both mentioned by Friar Odoric, the contemporary of his later years), artificial egg-hatching, printing of books (though the notice of this art seems positively challenged in his account of paper-money), besides a score of remarkable arts and customs which one would have expected to recur to his memory, are never alluded to. Neither does he speak of the great characteristic of the Chinese writing. It is difficult to account for these omissions, especially considering the comparative fulness with which he treats the manners of the Tartars and of the Southern Hindoos; but the impression remains that his associations in China were chiefly with foreigners. Wherever the place he speaks of had a Tartar or Persian name he uses that rather than the Chinese one. Thus _Cathay_, _Cambaluc_, _Pulisanghin_, _Tangut_, _Chagannor_, _Saianfu_, _Kenjanfu_, _Tenduc_, _Acbalec_, _Carajan_, _Zardandan_, _Zayton_, _Kemenfu_, _Brius_, _Caramoran_, _Chorcha_, _Juju_, are all Mongol, Turki, or Persian forms, though all have Chinese equivalents.[12]