the Vkmity of Paukgan or Pagan. 
the position of the Jo towns, as given by the slave, making 
Thilsen the one farthest north, in place of being the one ^^rthest 
south, as the slave represents ; and it must also be observed, that 
the authority of the slave is supported by that of the heredi- 
tary chief’s kinsman, who placed Thilaen directly west from 
Paukgan, as already mentioned. Neither is the account of the 
towmclerk confirmed by the map which he was drawing, when 
stopped by our sudden departure, while it was unfinished. 
Many of the differences, indeed, may be owing to this circum- 
stance, as he would probably have altered the names and situa-, 
tions of the places, when he came to review and complete his 
work; for I probably mistook his meaning, while I wrote the 
names during the operation. Thus, perhaps, I wrote Jo in 
place of Zho, the words being pronounced nearly as if written 
Yo and Jo in English. Thiloen, it must be observed, is not 
laid down in this draught. Whether he intended to place it on 
the Jowa opposite to Kakhiap or Jo I cannot say. 'This would 
have been agreeable to what he told me on the day before ; but 
I suspect that I then mistook his meaning, as all other authori- 
ties agreed in placing Thilaen in the southern part of the Jo 
territory. The general course of the Erawadi being from north 
to south, we are to consider the term above as signifying north, 
and below as agnifying south; but the map is drawn very pb- 
liquely, the cardinal points being near its corners ; for the re- 
markable hill called Pouppa bears about E.S.E. from Paukgan, 
and, therefore, the meridian will be nearly in the line that® 
have added. 
On the whole, rejecting Taunduaen as a city belonging to the 
dSnghiin and Launsci, which, although partly occupied by Jq, 
yet seems to be a dowery settled on one of the king’s wives, we 
have Jo, Zho, Thilaen, and perhaps Kakhia-p as capitals of the 
petty princes of the Jo nation, each probably possessing a terri- 
tory like one of our shires or counties, that is to say, from 600 
to 800 square miles in extent. 
About the middle of the sixteenth century of the Christian 
era, Paukgan is reported by the natives to have been the seat of 
empire ; and considerable portions of the wall and ditch still rCr 
main, from which we may judge that its fort has been fully as 
large as that of Amarapura. There is now much empty space 
