om the
indications in those extracts it can be gathered, I think, that Birdhul
was not far from the Kaveri (called Kanobari), not far from the sea, and
five or six days' march from Madura. These indications point to Tanjore,
Kombakonam, or some other city in or near the Kaveri Delta.[5] I should
suppose that this Birdhul was the capital of Polo's Sundara Pandi, and
that the port visited was Kaveripattanam. This was a great sea-port at one
of the mouths of the Kaveri, which is said to have been destroyed by an
inundation about the year 1300. According to Mr. Burnell it was the
"_Pattanam_ 'par excellence' of the Coromandel Coast, and the great port
of the Chola kingdom."[6]

[Illustration: Chinese Pagoda (so called) at Negapatam. (From a sketch
taken in 1846 by Sir Walter Elliot.)]

Some corroboration of the supposition that the Tanjore ports were those
frequented by Chinese trade may be found in the fact that a remarkable
Pagoda of uncemented brickwork, about a mile to the north-west of
Negapatam, popularly bears (or bore) the name of _the Chinese Pagoda_. I
do not mean to imply that the building was Chinese, but that the
application of that name to a ruin of strange character pointed to some
tradition of Chinese visitors.[7] Sir Walter Elliot, to whom I am
indebted for the sketch of it given here, states that this building
differed essentially from any type of Hindu architecture with which he was
acquainted, but being without inscription or sculpture it was impossible
to assign to it any authentic origin. Negapatam was, however, celebrated
as a seat of _Buddhist_ worship, and this may have been a remnant of their
work. In 1846 it consisted of three stories divided by cornices of stepped
brickwork. The interior was open to the top, and showed the marks of a
floor about 20 feet from the ground. Its general appearance is shown by
the cut. This interesting building was reported in 1859 to be in too
dilapidated a state for repair, and now exists no longer. Sir W. Elliot
also tells me that collectors employed by him picked up in the sand, at
several stations on this coast, numerous Byzantine and _Chinese_ as well
as Hindu coins.[8] The brickwork of the pagoda, as described by him, very
fine and closely fitted but without cement, corresponds to that of the
Burmese and Ceylonese mediaeval Buddhist buildings. The _architecture_ has
a slight resemblance to that of Pollanarua in Ceylon (see _Fergusson_, II.
p. 512). (_Abulf._ in