en any of them die, the bodies are burnt, and then they take the bones
and put them in little chests.

These are carried high up the mountains, and placed in great caverns,
where they are hung up in such wise that neither man nor beast can come at
them.

A good deal of gold is found in the country, and for petty traffic they
use porcelain shells such as I have told you of before. All these
provinces that I have been speaking of, to wit Bangala and Caugigu and
Anin, employ for currency porcelain shells and gold. There are merchants
in this country who are very rich and dispose of large quantities of
goods. The people live on flesh and rice and milk, and brew their wine
from rice and excellent spices.

NOTE 1.--The only MSS. that afford the reading _Coloman_ or _Choloman_
instead of _Toloman_ or _Tholoman_, are the Bern MS., which has _Coloman_
in the initial word of the chapter, Paris MS. 5649 (Pauthier's C) which
has _Coloman_ in the Table of Chapters, but not in the text, the Bodleian,
and the Brandenburg MS. quoted in the last note. These variations in
themselves have little weight. But the confusion between _c_ and _t_ in
mediaeval MSS., when dealing with strange names, is so constant that I
have ventured to make the correction, in strong conviction that it is the
right reading. M. Pauthier indeed, after speaking of tribes called _Lo_ on
the south-west of China, adds, "on les nommait _To-lo-man_ ('les nombreux
Barbares Lo')." Were this latter statement founded on actual evidence we
might retain that form which is the usual reading. But I apprehend from
the manner in which M. Pauthier produces it, without corroborative
quotation, that he is rather hazarding a conjecture than speaking with
authority. Be that as it may, it is impossible that Polo's Toloman or
Coloman should have been in the south of Kwangsi, where Pauthier locates
it.

On the other hand, we find tribes of both _Kolo_ and _Kihlau_ Barbarians
(i.e. _Man_, whence KOLO-MAN or _Kihlau-man_) very numerous on the
frontier of Kweichau. (See _Bridgman's transl. of Tract on Meautsze_, pp.
265, 269, 270, 272, 273, 274, 275, 278, 279, 280.) Among these the _Kolo_,
described as No. 38 in that Tract, appear to me from various particulars
to be the most probable representatives of the Coloman of Polo,
notwithstanding the sentence with which the description opens: "_Kolo_
originally called _Luluh_; the modern designation _Kolo_ is incorrect."[1]
They are at prese