t hitherto of nothing but sand, covered with _Saccharum_
_spontaneum_. _Andropogon muricatus_ is scarcely to be met with.

_26th_.--We left Mymensing this morning, and proceeded down the
Burrampooter, the banks of which still present for the most part nothing
but a succession of sandy banks covered with _Saccharum spontaneum_.
The stream is not very rapid, and the river, owing to the numerous
islands and banks, does not present so imposing an appearance as the
Ganges. For the last week strong easterly winds have prevailed.

_27th_.--We entered the mouth of the Soormah, or, as the natives seem to
call it, the Barak. The water of this river or portion of the Megna? is
remarkably clear, compared with that of the Ganges; as indeed is that of
the Burrampooter.

_30th_.--Some time after we entered the Soormah we apparently left its
channel, and up to this morning we have passed through a tract of jheels
with a few clear and very deep channels. The villages are built on small
eminences, and are entirely surrounded with water; they have the usual
form, and those houses adjoining the water have fences of an _Arundo_,
which they tell us are intended to keep out the grass. We have since
entering these jheels passed through and between immense beds of
vegetation, formed principally of _Oplismenus_ (Panicum) _stagninus_,
_Leersia_? _aristata_, which by-the-bye is a distinct genus. _Villarsia_
_cristata_, _Nymphaealotus_, Potamogeton, _Azolla Salvinii_, etc. etc.
The only novel things we have met with are _Ischaemum cuspidatum_,
Roxb. (sui generis,) and a small grass intermediate between Panicum and
Chamaeraphis. The wild form of _Oryza sativa_, _Panicum interruptum_
and _Leersia_? _ciliaris_, Roxb. also occur; the two former in abundance.
On the more dry tracts, that occasionally though very rarely occur,
_Andropogon muricatus_ appears. No _Saccharum_ presented itself since
the 28th. High ground was visible yesterday evening, apparently at a
great distance.

_October 1st_.--We have continued to pass through immense jheels: about
6 A.M. we arrived at Hubbe-gunge, a large native town, situated on the
Barak, which does not deserve the name of a river. The actual distance
from this place to Chattuc is about 42 miles, and the high land in that
direction was faintly visible for about 2 hours in the morning. The
ground to the Eastward is losing the "Jheel" character, and appears
densely wooded, and to the S.E. rather high hil