It is very probable that two different words, Siang and Sing, got confounded by the non-Chinese attaches of the Imperial Court; but it seems to me quite certain that they applied the same word, Sing or Sheng, to both institutions, viz. to the High Council of State, and to the provincial governments.  It also looks as if Marco Polo himself had made that very confusion with which Pauthier charges Neumann.  For whilst here he represents the twelve Barons as forming a Council of State at the capital, we find further on, when speaking of the city of Yangchau, he says:  “Et si siet en ceste cite uns des xii Barons du Grant Kaan; car elle est esleue pour un des xii sieges,” where the last word is probably a mistranscription of Sciengs, or Sings, and in any case the reference is to a distribution of the empire into twelve governments.

To be convinced that Sing was used by foreigners in the double sense that I have said, we have only to proceed with Rashiduddin’s account of the administration.  After what we have already quoted, he goes on:  “The Sing of Khanbaligh is the most eminent, and the building is very large.... Sings do not exist in all the cities, but only in the capitals of great provinces....  In the whole empire of the Kaan there are twelve of these Sings; but that of Khanbaligh is the only one which has Ching-sangs amongst its members.”  Wassaf again, after describing the greatness of Khanzai (Kinsay of Polo) says:  “These circumstances characterize the capital itself, but four hundred cities of note, and embracing ample territories, are dependent on its jurisdiction, insomuch that the most inconsiderable of those cities surpasses Baghdad and Shiraz.  In the number of these cities are Lankinfu and Zaitun, and Chinkalan; for they call Khanzai a Shing, i.e. a great city in which the high and mighty Council of Administration holds its meetings.”  Friar Odoric again says:  “This empire hath been divided by the Lord thereof into twelve parts, each one thereof is termed a Singo.”

Polo, it seems evident to me, knew nothing of Chinese.  His Shieng is no direct attempt to represent any Chinese word, but simply the term that he had been used to employ in talking Persian or Turki, in the way that Rashiduddin and Wassaf employ it.

I find no light as to the thirty-four provinces into which Polo represents the empire as divided, unless it be an enumeration of the provinces and districts which he describes in the second and third parts of Bk.  II., of which it is not difficult to reckon thirty-three or thirty-four, but not worth while to repeat the calculation.