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 Indus system is nearly exactly the same as Ptolemy's. That the main
Indus is shown as coming down from the mountains which border Pamir on the
south may be regarded as an improvement. The Zaradrus, Satlej, has not changed its
appearance since Ptolemy; the only difference is that, on Mercator's map, it comes from
a meridional range separating the catchment area of the Indus from that of the Ganges.

The most extraordinary thing is, however, the way in which Mercator has
drawn the Ganges. He identifies Golfo de Bengala with Ptolemy's Sinus argaricus,
and makes two rivers join in the Ganges delta. The one from the west is called
Guenga flu., and corresponds to a certain extent with Gastaldi's Ganga f. But on
the latter is Orissa and it is therefore the same river as the present Mahanadi which
indeed flows through Orissa and has its own independent mouth in the Gulf of
Bengal. The lower course of Mercator's Guenga, on the other hand, represents, as
it were, a last fragment of the classical Ganges. The second river comes from the
north with two branches, one from a mountain range south of Mien regnum, the other
from Vindius mons; this river, after the junction, is called Chaberis, and identical
with Ptolemy's Chaberus, which falls in the Sinus argaricus. As the Chaberis flows
through Aracam, a country which Gastaldi had placed correctly, Mercator has taken
a river from the west and a country from the east and let them meet in the neigh-
bourhood of Bengal. In 1569 the coast of Arakan was very well known by all
Portuguese sailors, so one cannot understand why Mercator has placed this coast-
land in the interior of India.

The worst is, however, that he removes the whole Ganges system far to the
east, places it in southern China and confounds the Ganges with the Hsi-chiang.
Therefore Ptolemy's India intra Gangem Fluvium has to follow on the way eastwards,
and is to be found in upper Burma, Yunnan, and Kwang-si where we read these
words: Indostan que veteribus India intra Gangem. The river itself he calls Cantan
flu. olim Ganges, and the city Cantan olim Gange, which is situated in the Hsi-chiang
delta exactly at the place of Canton. The feeders of the Ganges are exactly the
same as on Ptolemy's map: Diamuna, Ganges, Sarabis and Bepyrus. On the upper
Diamuna, Jumna, is a town Cabol; but on a western tributary of the Indus there are,
also correctly, both Cabul and Chesimur. The mountains from which the feeders of
the Ganges come belong to the Imaus, and there is a legend: Formicæ hic aurum
effodientes homines sunt, the old story which first was told by Herodotus.

Dr. Dahlgren has explained the cause of this singular mistake. On Mercator's
map of the world we read, as an addition to the name of Sumatra: olim Taprobana.
Instead of identifying Ptolemy's Taprobana insula with Ceylon, he believes it to be
Sumatra. And as Ptolemy's Ganges was situated east of Taprobana, the real Ganges
must come out to the sea east of Sumatra. He finds even a resemblance between
the words Canton and Ganges.¹ He confuses two islands south of Asia, and there-