Where the MSS. in the repetition of a name afforded a choice of forms, I have selected that which came nearest the real name when known.  Thus the G. T. affords Baldasciain, Badascian, Badasciam, Badausiam, Balasian.  I adopt BADASCIAN, or in English spelling BADASHAN, because it is closest to the real name Badakhshan.  Another place appears as COBINAN, Cabanat, Cobian.  I adopt the first because it is the truest expression of the real name Koh-benan.  In chapters 23, 24 of Book I., we have in the G. T. Asisim, Asciscin, Asescin, and in Pauthier’s MSS. Hasisins, Harsisins, etc.  I adopt ASCISCIN, or in English spelling ASHISHIN, for the same reason as before.  So with Creman, Crerman, Crermain, QUERMAN, Anglice KERMAN; Cormos, HORMOS, and many more.[2]

In two or three cases I have adopted a reading which I cannot show literatim in any authority, but because such a form appears to be the just resultant from the variety of readings which are presented; as in surveying one takes the mean of a number of observations when no one can claim an absolute preference.

Polo’s proper names, even in the French Texts, are in the main formed on an Italian fashion of spelling.[3] I see no object in preserving such spelling in an English book, so after selecting the best reading of the name I express it in English spelling, printing Badashan, Pashai, Kerman, instead of Badascian, Pasciai, Querman, and so on.

And when a little trouble has been taken to ascertain the true form and force of Polo’s spelling of Oriental names and technical expressions, it will be found that they are in the main as accurate as Italian lips and orthography will admit, and not justly liable either to those disparaging epithets[4] or to those exegetical distortions which have been too often applied to them.  Thus, for example, Cocacin, Ghel or Ghelan, Tonocain, Cobinan, Ondanique, Barguerlac, Argon, Sensin, Quescican, Toscaol, Bularguci, Zardandan, Anin, Caugigu, Coloman, Gauenispola, Mutfili, Avarian, Choiach, are not, it will be seen, the ignorant blunderings which the interpretations affixed by some commentators would imply them to be, but are, on the contrary, all but perfectly accurate utterances of the names and words intended.

The _-tcheou_ (of French writers), _-choo_, _-chow_, or _-chau_[5] of English writers, which so frequently forms the terminal part in the names of Chinese cities, is almost invariably rendered by Polo as _-giu_.  This has frequently in the MSS., and constantly in the printed editions, been converted into _-gui_, and thence into _-guy_.  This is on the whole the most constant canon of Polo’s geographical orthography, and holds in Caagiu (Ho-chau), Singiu (Sining-chau), Cui-giu (Kwei-chau), Sin-giu (T’sining-chau), Pi-giu (Pei-chau), Coigangiu (Hwaingan-chau), Si-giu (Si-chau), Ti-giu (Tai-chau), Tin-giu (Tung-chau), Yan-giu (Yang-chau), Sin-giu (Chin-chau), Cai-giu (Kwa-chau), Chinghi-giu (Chang-chau), Su-giu (Su-chau), Vu-giu (Wu-chau), and perhaps a few more.  In one or two instances only (as Sinda-ciu, Caiciu) he has _-ciu_ instead of _-giu_.