1186 Riuan Thsang* s Itinerary . [Dec. 



read the word Kofadeyan, Kieii poutan. — By dropping the final <j n ; 

 and the j r, as usual to the Chinese authors — we have Tsao for Soorun. 

 De la 300 li ouest, 



1 1 . Kiou chouang ni kia, ou Koue'i chouang no. t^s: 3 * 



Kiou Chouangni, I conclude to be a reading of Khojund. 

 De la 200 li ouest. 



12. Ko han, Toung'an. &8js± 



Kohan is clearly Kookhan or Khokan. 

 De la 400 li ouest. 



13. Pou ho (Boukhara) Tchoung' an. \j*s£ 



Pou ho, Bokhara ; is a fair example of the theory T am induced to 

 propose, — Pouho is no identification of the sound, Bokhara,— the 

 stress of the sound lies in the penultimate kh ; that the Chinese can 

 receive and transmit this sound, we have proof in the two preceding 

 places, — Khojund and Khokan ; of disappearance or change in initial 

 or final letters we are constantly made aware — hence we need not 

 pause on the usual change of p for b. But, if we suppose a Chinese 

 author to be making up a geographical work, and consulting an Ara- 

 bic or Persian book, for his guide — let the work be a Shukustah 

 copy without diacritical points, and we shall be at no loss to under- 

 stand his reading^ for b — and h for kh — in Ijke — and hence ignorant 

 of the real sounds — to have transcribed letter for letter and inserted the 

 name as Pou ho. 



The reading h for kh runs through the whole work. It may be 

 remarked, that in reading from manuscript proper names, with no 

 assistance from the meaning of the terms — that it will be invariably 

 found, that one formation of letters being once in the beginning ac- 

 cepted for certain letters and certain sounds, the same, right or wrong, 

 will be carried through the whole work ; thus we shall find h substi- 

 tuted for kh — and^a for wa — or p or b or/ for w. 



De la 400 li ouest. 



14. Fa ti Si' an. ^UekAj 

 Fati Sian — is clearly Budukhshan— the kh being read as h ; as an 



aspirate it has disappeared. 

 De la 500 li sud-ouest, 



