. The town is near the head of the Green River
(the Ts'ien T'ang) which flows in a N.E. direction and falls into the Bay
of Hang-chau. At Chang-shan the stream is no longer navigable even for
small boats. Travellers going west or south-west walk or are carried in
sedan-chairs across country in a westerly direction for about 30 miles to
a town named Yuh-shan. Here there is a river which flows westward ('the
other half goes down'), taking the traveller rapidly in that direction,
and passing _en route_ the towns of Kwansinfu, Hokow or Hokeu, and onward
to the Poyang Lake." From the careful study of Mr. Fortune's published
narrative I had already arrived at the conclusion that this was the
correct explanation of the remarkable expressions about the division of
the waters, which are closely analogous to those used by the traveller in
ch. lxii. of this book when speaking of the watershed of the Great Canal
at Sinjumatu. Paraphrased the words might run: "At Chang-shan you reach
high ground, which interrupts the continuity of the River; from one side
of this ridge it flows up country towards the north, from the other it
flows down towards the south." The expression "The River" will be
elucidated in note 4 to ch. lxxxii. below.

This route by the Ts'ien T'ang and the Chang-shan portage, which turns the
danger involved in the navigation of the Yang-tzu and the Poyang Lake, was
formerly a thoroughfare to the south much followed; though now almost
abandoned through one of the indirect results (as Baron Richthofen points
out) of steam navigation.

The portage from Chang-shan to Yuh-shan was passed by the English and
Dutch embassies in the end of last century, on their journeys from
Hang-chau to Canton, and by Mr. Fortune on his way from Ningpo to the Bohea
country of Fo-kien. It is probable that Polo on some occasion made the
ascent of the Ts'ien T'ang by water, and that this leads him to notice the
interruption of the navigation.

[Mr. Phillips writes (_T. Pao_, I. p. 222): "From Fuyang the next point
reached is Tunglu, also another 100 _li_ distant. Polo calls this city
Ugim, a name bearing no resemblance to Tunglu, but this name and Ta Pin Zu
are so corrupted in all editions that they defy conjecture. One hundred
_li_ further up the river from Tunglu, we come to Yenchau, in which I
think we have Polo's Gengiu of Ramusio's text. Yule's text calls this city
Ghiuju, possibly an error in transcription for Ghinju; Yenchau in ancient