318 
H. Beveridge — The Site of Kama Suvarna. 
[No. 4, 
makes Hiuen Tsiang diverge to the north-east,* or Paundra Vardhana, 
and also causes him to describe two sides of a nearly equilateral triangle, 
between Tamluk and Orissa. This may be seen from M. Saint Martin’s 
map where, however, the route is made still more awkward by his sup- 
position that Paundra Vardhana is Burdwan. This it cannot be, for the 
itinerary places it on the east of the Ganges. f It is rightly placed there 
in the Chino-Japanese map of 1710, of which M. Julien has given a re- 
duction. It seems very unlikely, too, that Hiuen Tsiang would turn 
inland and to the N. W. after arriving at Tamluk. Presumably he went 
there in order to embark for Ceylon, as his predecessor Fa-Hian had done. 
The biography, at all events, tells us J that he designed when at Tam- 
luk, to sail to Ceylon and that he was dissuaded from doing so by a monk 
from southern India. This man advised him not to attempt so long and 
dangerous a navigation, but to sail from the S. W. point of India, whence 
he could make the journey in three days. This would give him an 
opportunity, the monk added, of visiting the sacred places of Orissa 
and other kingdoms, Hiuen Tsiang took this advice and started for the 
S. W. and arrived at Orissa. This is all straightforward ; whereas the 
going to Karna Suvarna from Tamluk involved a detour of at least 140 
miles. 
For these reasons I am disposed to prefer the route given in the 
biography. I am not sure, however, if this is to the advantage of my 
contention that Karna Suvarna is Rangamati. Neither route is dis- 
cor-dant with the identification, but the Si-yu-ki one is more detailed. 
Rahgamati § is nearly due north of Tamluk and 120 or 130 miles off, 
and the borders of Orissa are about an equal distance to the S. W. of 
Rahgamati. We must not press Hiuen Tsiang’s measurements closely, 
for we do not know the exact length of the li, nor do we always know to 
what points he refers. He generally speaks only of countries, not of 
towns, and it may be that the distances are those to and from the con- 
fines of kingdoms. 
* It describes the direction as easterly, but Koch Bihar and Kamrfip lie N. 
E. from Pandua. 
f It seems a happy suggestion of Mr. Westmacott’s that the name Paundra 
is preserved in Abfi’l Fail’s “ Sarkar of Panjra.” The chief objection to the identi- 
fication of Pandua with Paundra Vardhana seems to be that the central or home- 
farm pargana of Sarkar Panjra, viz., Haveli Panjra, lies N. E. of Dinajpur and far 
from Pandua which apparently is in Shashhazari. [Ain, III, XV ; Vol. II, p. 136 
of Col. Jarrett’s translation where it is called Sarkar Pinjarah. Ed.] 
X I. 183. 
§ There are several Rahgamatis, and the best known, perhaps, is that in Lower 
Assam. But the one we have to do with is in Central Bengal and on the Bhagirathi. 
Sir H. Yule suggested that it might be the Kartasina of Ptolemy. 
