NOTE 1.-Tarcasci (G. T.). This word is worthy of note as the proper form of what has become in modern French carquois. The former is a transcript of the Persian Tărkăsh; the latter appears to be merely a corruption of it, arising perhaps clerically from the constant confusion of c and t in MSS. (see Defréméry, quoted by Pauthier, in loco).

NOTE 2.-The stores are now outside the walls of the "Prohibited City," corresponding to Polo's Palace-Wall, but within the walls of the "Imperial City" (Middle Kingdom, I. 61). See the cut at p. 365.

NOTE 3. The two gates near the corners apparently do not exist in the Palace now. "On the south side there are three gates to the Palace, both in the inner and the outer walls. The middle one is absolutely reserved for the entrance or exit of the Emperor; all other people pass in and out by the gate to the right or left of it" (Trigautius, B. I. ch. vii.). This custom is not in China peculiar to Royalty. In private houses it is usual to have three doors leading from the court to the guest-rooms, and there is a great exercise of politeness in reference to these; the guest after much pressing is prevailed on to enter the middle door, whilst the host enters by the side (see Deguignes, Voyages, I. 262).

It

NOTE 4.-Ramusio's version here diverges from the old MSS. makes the inner enclosure a mile square; and the second (the city of Taidu) six miles square, as here, but adds, at a mile interval, a third of eight miles square. Now it is remarkable that Mr. A. Wylie, in a letter dated 4th December 1873, speaking of a recent visit to Peking, says: "I found from various inquiries that there are several remains of a very much larger city wall, enclosing the present city; but time would not allow me to follow up the traces."

Pauthier's text (which I have corrected by the G. T.), after describing the outer inclosure to be a mile every way, says that the inner inclosure lay at an interval of a mile within it!

NOTE 5. This description of palace (see opposite cut), an ele vated basement of masonry with a superstructure of timber (in general carved and gilded), is still found in Burma, Siam, and Java, as well as in China. If we had any trace of the palaces of the ancient Asokas and Vikramadityas of India, we should probably find that they were of the same character. It seems to be one of those things that belonged to some ancient Panasiatic fashion, as the palaces of Nineveh were of a somewhat similar construction. In the Audience Halls of the Moguls at Delhi and Agra we can trace the ancient form, though the superstructure has there become an arcade of marble instead of a pavilion on timber columns.

NOTE 6." As all that one sees of these palaces is varnished in those colours, when you catch a distant view of them at sunrise, as I have done many a time, you would think them all made of, or at least

covered with, pure gold enamelled in azure and green, so that the spectacle is at once majestic and charming" (Magaillans, p. 353).

NOTE 7.-" On the west side, where the space is amplest, there is a lake very full of fish. It is in the form of a fiddle, and is an Italian mile and a quarter in length. It is crossed at the narrowest part, which corresponds to gates in the walls, by a handsome bridge, the extremities of which are adorned by two triumphal arches of 3 openings each. . . . The lake is surrounded by palaces and pleasure houses, built partly in the water and partly on shore, and charming boats are provided on it

for the use of the Emperor when he chooses to go a-fishing or to take an airing (Ib. 282-3). The marble bridge, as it now exists, consists of nine arches, and is 600 feet long. (Rennie's Peking, II. 57.)

Ramusio specifies another lake in the city, fed by the same stream before it enters the palace, and used by the public for watering cattle.

NOTE 8. The expression here is in the Geog. Text, "Roze de l'açur," and in Pauthier's, "de rose et de l'asur." Rose Minerale, in the terminology of the alchemists, was a red powder produced in the sublimation of gold and mercury, but I can find no elucidation of the term Rose of Azure. The Crusca Italian has in the same place Terra dello Azzurro. Having ventured to refer the question to the high authority of Mr. C. W. King, he expresses the opinion that Roze here stands for Roche, and that probably the term Roche de l'azur may have been used loosely for blue-stone, i.e. carbonate of copper, which

would assume a green colour through moisture. He adds: "Nero, according to Pliny, actually used chrysocolla, the siliceous carbonate of copper, in powder, for strewing the circus, to give the course the colour of his favourite faction, the prasine (or green). There may be some analogy between this device and that of Kublai Khan." This parallel is a very happy one.

NOTE 9.--Friar Odoric gives a description, short, but closely agreeing in substance with that in the Text, of the Palace, the Park, the Lake, and the Green Mount.

A green mount, answering to the description, and about 160 feet in height, stands immediately in rear of the palace buildings. It is called by the Chinese King-Shan, "Court Mountain," Wan-su-Shan, "Ten Thousand Year Mount," and Mei-Shan, "Coal Mount," the last from the material of which it is traditionally said to be composed (as a provision of fuel in case of siege).* Whether this is Kublai's Green Mount does not seem to be quite certain. Dr. Lockhart tells me that, according to the information he collected when living at Peking, it is not so, but was formed by the Ming Emperors from the excavation of the

* Some years ago, in Calcutta, I learned that a large store of charcoal existed under the soil of Fort William, deposited there I believe in the early days of that fortress.

existing lake on the site which the Mongol Palace had occupied. There is another mount, he adds, adjoining the east shore of the lake, which must be of older date even than Kublai, for a Dagoba standing on it is ascribed to the Kin

In a plate attached to next chapter, I have drawn, on a small scale, the existing cities of Peking, as compared with the Mongol and Chinese cities in the time of Kublai. The plan of the latter has been constructed (1) from existing traces, as exhibited in the Russian Survey republished by our War Office; (2) from information kindly afforded by Dr. Lockhart; and (3) from Polo's description and a few slight notices by Gaubil and others. It will be seen, even on the small scale of these plans, that the general arrangement of the palace, the park, the lakes (including that in the city, which appears in Ramusio's version), the bridge, the mount, &c., in the existing Peking, very closely correspond with Polo's indications; and I think the strong probability is that the Ming really built on the old traces, and that the lake, mount, &c., as they now stand, are substantially those of the Great Mongol, though Chinese policy or patriotism may have spread the belief that the foreign traces were obliterated. Indeed if that belief were true the Mongol Palace must have been very much out of the axis of the City of Kublai, which is in the highest degree improbable. The Bulletin de la Soc. de Géographie for September 1873, contains a paper on Peking by the physician to the French Embassy there. Whatever may be the worth of the meteorological and hygienic details in that paper, I am bound to say that the historical and topographical part is so inaccurate as to be of no value.

NOTE 10. For son, read grandson. But the G. T. actually names the Emperor's son Chingkim, whose death our traveller has himself already mentioned.

CHAPTER XI.

CONCERNING THE CITY OF CAMBALUC.

Now there was on that spot in old times a great and noble city called CAMBALUC, which is as much as to say in our tongue "The City of the Emperor." But the Great Kaan was informed by his Astrologers that this city would prove rebellious, and raise great disorders against his imperial authority. So he caused the present city to be built close beside the old one, with only a river between them. And he caused the people of the old city to be

removed to the new town that he had founded; and this is called TAIDU. [However, he allowed a portion of the people which he did not suspect to remain in the old city, because the new one could not hold the whole of them, big as it is.]

As regards the size of this (new) city you must know that it has a compass of 24 miles, for each side of it hath a length of 6 miles, and it is four-square. And it is all walled round with walls of earth which have a thickness of full ten paces at bottom, and a height of more than 10 paces; but they are not so thick at top, for they diminish in thickness as they rise, so that at top they are only about 3 paces thick. And they are provided throughout with loop-holed battlements, which are all whitewashed.

2

There are 12 gates, and over each gate there is a great and handsome palace, so that there are on each side of the square three gates and five palaces; for (I ought to mention) there is at each angle also a great and handsome palace. In those palaces are vast halls in which are kept the arms of the city garrison.3

The streets are so straight and wide that you can see right along them from end to end and from one gate to the other. And up and down the city there are beautiful palaces, and many great and fine hostelries, and fine houses in great numbers. [All the plots of ground on which the houses of the city are built are four-square, and laid out with straight lines; all the plots being occupied by great and spacious palaces, with courts and gardens of proportionate size. All these plots were assigned to different heads of families. Each square plot is encompassed by handsome streets for traffic; and thus the whole city is arranged in squares just like a chess-board, and disposed in a manner so perfect and masterly that it is impossible to give a description that should do it justice.]+

Moreover, in the middle of the city there is a great