国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

the case of the Brahmaputra I would say that the source is to be found in one great
glacier and that the brooks from smaller glaciers contribute to the volume of the
brook from the principal glacier.
One year later, or August 15th, 1895, I crossed the Wakhjir Pass, though
I did not pay any special attention to the source; but as far as my observations
went I agree completely with Curzon's opinion. ¹
Curzon's examination of the different rivers which have been claimed as the
source is extremely able and does not omit a single case. He says the little river
joining at Bozai Gumbaz, which CAPUS regarded as one of the parent streams has
not a single argument in its favour. The next claimant is the Pamir-river from
Sor-kul or Victoria Lake, discovered by WOOD in 1838. Wood regarded it as the
true parent stream of the Oxus, and Curzon shows that it has taken fifty years to
rectify the mistake. Neither in length, volume, nor any of the requisite characteristics
can it claim to be the parent stream.
The Bartang, Murghab or Aksu, which comes from the Chakmak Lake has
been defended as the parent stream by VENIUKOFF, RAWLINSON and others, an
opinion which has found favour with some who were unacquainted with the full geo-
graphical details. In favour of the Aksu it has been said that it is longer and has
a greater volume than the Panja. The Russians give it a length of 252 miles from
the Chakmak Lake to Kala Wamar, whereas the Panja from its source to the same
point was roughly estimated at 240 miles. Curzon does not regard the length as a
proof of a river being a parent river. And in this case he finds the length of Panja
to be 270 miles.
Regarding the volume, Curzon's opinion should be compared with the method
I have been following when examining whether the Maryum-chu or the Kubi-tsangpo
should be regarded as the source of the Brahmaputra. Curzon says: ² it is obvious
that the sole method of applying this test is . . . by measurement of the volume of
the two streams at the point of junction. Not one of the advocates of the Aksu-
Murghab theory, however, has ever visited Kala Wamar³. NEY ELIAS had found,
in November 1885, that the Panja was much more voluminous than the Aksu. Curzon
also shows that the source of the Panja is situated at a higher elevation. Finally
he points out that the historical proofs are not without weight. His words run: 'The
final argument is that of historical authority and popular acceptance. In the first
place, it is noteworthy that, from as far back as records extend, the identity of the
river below Kala Wamar with the Panja (and not with the Murghab) above has been
assumed by the inhabitants of the district, indicating that, in their opinion, this was
the parent stream. Like the Greek Oxus, so the Persian name Panja is applied to
the river both below and above Kala Wamar . . .' Already in 1815 Lieut. MACARTNEY