24 
ANCIENT GEOGRAPHY OF KAS'MIR. [Extra No. 2, 
of the Jehlam ami Kisangahga. The easiest route to Kasmir from 
the west leads through the open central portion of Hazara (Urasa) to 
Mansahra; hence across the Kunhar and Kisangahga rivers to Muzaf- 
farabad, and then up by the right side of the Jehlam Valley to 
Baramula. 1 * * In Kusnarl it is easy to recognize with Prof. Sachau the 
present Kunhar River which falls info the Jehlam a few miles below its 
great bend at Mnzaffarabad. a The Mahwi is evidently meant to designate 
the Kisangahga. 8 If thus interpreted the oidy error in Alb§runi’s de¬ 
scription is that it makes the Kunhar join the Kisanganga whereas 
in reality it falls into the Jehlam after the latter’s junction with the 
K isanganga. 
I have shown in my note on Rajat. v. 215 that the route here indi¬ 
cated, which was a favorite one until the modern “Jehlam Valley 
Tonga Road ” was constructed, is distinctly referred to already in 
Kalhana’s account of S'amkaravarman’s marcli to and from Urasa. The 
distance of 8 farsakli corresponds according to Alberuni’s reckoning to 
about 39 English miles. 4 * Referring to the map and the modern route 
measurements 6 * this distance carries us to a point between Mansahra 
and the next stage Abbottabad, i.e., exactly into the neighbourhood 
where according to the evidence given in the above-quoted note the old 
capital of Urasa must be located. ‘ Babrahan ’ which cannot be identi¬ 
fied at present, is perhaps intended to represent the name of this old 
town which could fairly be described as situated midway between the 
Indus and Jehlam. 
From Muzaffarabad onwards, — where there is still a bridge over the 
Kisangahga just as at the time (1783) when Forster crossed here on his 
way from Kasmir to Attock, 6 and as, if our explanation is right, in the 
time of Alberuni,— we can follow the route quite plainly. Alberiini 
counts five days of march “ to the beginning of the ravine whence the 
l This route is described, e.g., by Drew, Jummoo, p. 528, ‘ as the easiest route 
from the Panjab to Kasmir.’ 
8 Kunhar represents the regular phonetic derivative of a Skr. * Kusndri, medial 
s becoming always h under a phonetic law common to Kasmir! and the related 
dialects ; for the change hn > nh compare Grierson, Phonology of Indo-Aryan 
Vernaculars, Z. D. M. G., 1896, p. 33. 
8 I am unable to account for the name Mahwi. Could it be the corruption of 
an Apabhramsa derivative of Madhumatl ? This name, though properly applied 
to an affluent of the Kisangahga, is used in a Mahatmya also for the latter river 
itself ; see Note B, Rajat. i. 37, § 16. 
4 Compare Prof. Sacha,u’s note, India, ii. p. 316. Alberuni values his farsaM. 
at 4 Arabian miles or approximately 4x2186 yards. Hence 1 farsakh = 
English miles. 
& See Drew, loc. ext. 
8 See G. Forster, Journey from Bengal to England , 1808, ii; p. 46. 
