NOTE 4.—This sheet of water is the celebrated SI-HU, or “Western Lake,” the fame of which had reached Abulfeda, and which has raised the enthusiasm even of modern travellers, such as Barrow and Van Braam. The latter speaks of _three_ islands (and this the Chinese maps confirm), on each of which were several villas, and of causeways across the lake, paved and bordered with trees, and provided with numerous bridges for the passage of boats. Barrow gives a bright description of the lake, with its thousands of gay, gilt, and painted pleasure boats, its margins studded with light and fanciful buildings, its gardens of choice flowering shrubs, its monuments, and beautiful variety of scenery. None surpasses that of Martini, whom it is always pleasant to quote, but here he is too lengthy. The most recent description that I have met with is that of Mr. C. Gardner, and it is as enthusiastic as any. It concludes: “Even to us foreigners ... the spot is one of peculiar attraction, but to the Chinese it is as a paradise.” The Emperor K’ien Lung had erected a palace on one of the islands in the lake; it was ruined by the T’ai-P’ings. Many of the constructions about the lake date from the flourishing days of the T’ang Dynasty, the 7th and 8th centuries. Polo’s ascription of a circumference of 30 miles to the lake, corroborates the supposition that in the compass of the city a confusion had been made between miles and _li_, for Semedo gives the circuit of the lake really as 30 _li_. Probably the document to which Marco refers at the beginning of the chapter was seen by him in a Persian translation, in which _li_ had been rendered by _míl_. A Persian work of the same age, quoted by Quatremère (the _Nuzhát al-Ḳulúb_), gives the circuit of the lake as six parasangs, or some 24 miles, a statement which probably had a like origin. Polo says the lake was _within_ the city. This might be merely a loose way of speaking, but it may on the other hand be a further indication of the former existence of an extensive outer wall. The Persian author just quoted also speaks of the lake as within the city. (_Barrow’s Autobiog._, p. 104; _V. Braam_, II. 154; _Gardner_ in _Proc. of the R. Geog. Soc._, vol. xiii. p. 178; _Q. Rashid_, p. lxxxviii.) Mr. Moule states that popular oral tradition does enclose the lake within the walls, but he can find no trace of this in the Topographies.