A note from Mr. C. Gardner says:  “There are numerous public baths at Hang-chau, as at every Chinese city I have ever been in.  In my experience natives always take hot baths.  But only the poorer classes go to the public baths; the tradespeople and middle classes are generally supplied by the bath-houses with hot water at a moderate charge.”

NOTE 9.—­The estuary of the Ts’ien T’ang, or river of Hang-chau, has undergone great changes since Polo’s day.  The sea now comes up much nearer the city; and the upper part of the Bay of Hang-chau is believed to cover what was once the site of the port and town of KANP’U, the Ganpu of the text.  A modern representative of the name still subsists, a walled town, and one of the depots for the salt which is so extensively manufactured on this coast; but the present port of Hang-chau, and till recently the sole seat of Chinese trade with Japan, is at Chapu, some 20 miles further seaward.

It is supposed by Klaproth that KANP’U was the port frequented by the early Arab voyagers, and of which they speak under the name of Khanfu, confounding in their details Hang-chau itself with the port.  Neumann dissents from this, maintaining that the Khanfu of the Arabs was certainly Canton.  Abulfeda, however, states expressly that Khanfu was known in his day as Khansa (i.e.  Kinsay), and he speaks of its lake of fresh water called Sikhu (Si-hu). [Abulfeda has in fact two Khanqu (Khanfu):  Khansa with the lake which is Kinsay, and one Khanfu which is probably Canton.  (See Guyard’s transl., II., ii., 122-124.)—­H.C.] There seems to be an indication in Chinese records that a southern branch of the Great Kiang once entered the sea at Kanp’u; the closing of it is assigned to the 7th century, or a little later.

[Dr. F. Hirth writes (Jour.  Roy.  As.  Soc., 1896, pp. 68-69):  “For centuries Canton must have been the only channel through which foreign trade was permitted; for it is not before the year 999 that we read of the appointment of Inspectors of Trade at Hang-chou and Ming-chou.  The latter name is identified with Ning-po.”  Dr. Hirth adds in a note:  “This is in my opinion the principal reason why the port of Khanfu, mentioned by the earliest Muhammadan travellers, or authors (Soleiman, Abu Zeid, and Macoudi), cannot be identified with Hang-chou.  The report of Soleiman, who first speaks of Khanfu, was written in 851, and in those days Canton was apparently the only port open to foreign trade.  Marco Polo’s Ganfu is a different port altogether, viz. Kan-fu, or Kan-pu, near Hang-chou, and should not be confounded with Khanfu.”—­H.C.]