In towing against the ftream, the steep fide is alfo generally preferred, on account of the depth of water; although the current runs so much stronger there, than on the oppofite fide. On these occafions, one ought to be provided with a very long track-rope, as well to avoid the falling pieces of the fteep bank on the one fide, as the fhallow water on the other, when it becomes neceffary to change fides, through the badnefs of the tracking ground. The anchor should always be kept ready for dropping, in case the track-rope breaks.

;

Seventeen, to twenty miles a day, according to the ground, and the number of impediments, is the greatest distance that a large budgerow can be towed against the ftream, during the fair feafon and to accomplish this, the boat must be drawn at the rate of four miles and a half per hour, through the water, for 12 hours. When the waters are high, a greater progrefs will be made, notwithstanding the increased velocity of the current; because the filling of the river-bed gives many opportunities of cutting off angles and turnings; and fometimes even large windings, by going through creeks. And as the wind at this feason, blows upwards in most of the rivers, opportunities of using the fail frequently occur.

In the very fingular navigation across the feels, or inundation, between Dacca and Nattore, &c. in which 100 miles or more, are filed on nearly a ftraight courfe, leaving the villages and groves to the right and left; little difficulty occurs, unless the wind fhould fail for while it continues to blow, it is always fair, during the season of the inundation. The current prefents only a trifling obftacle; fince its motion (which is nearly parallel to the course of the Ganges) is feldom half a mile per hour:

The season of the northwesters, is, above all others, that which requires the most attention and care. Should one of thofe fqualls approach, and no creek or inlet offer for fhelter, when in the wide

taken

rivers; the steep bank, if not in a crumbling state*, fhould always be preferred to the flat one, whether it lie to windward or leeward. If the bank be in a crumbling state, a retreat to a firm part of it, (which is most likely to be found in the ftraightest parts of the river) should be attempted. But if this cannot be done, the flat fide muft be up with; and if it be a lee-fore, the anchor should be thrown out to prevent driving on it. In these cases the mast is always fupposed to be struck; and provided this be done, and the cargo judiciously disposed, it is probable that a well-constructed budgerow will be in no danger of overfetting by the mere force of the wind alone although by an unfortunate, or an ill-chofen fituation, it may be fo much expofed to the waves, as to be filled and sunk by them. At this feason, every traveller fhould be particularly attentive to the nature of the river-bank, as well as to the appearance of the horizon, during the last hours of the afternoon; and if he finds a place of shelter, he should ftop for the night: and not hesitate about losing time, which may be retrieved the next morning, by fetting out fo much earlier. The boatmen work with much more alacrity on this plan; because they have day-light before them to secure their boat, provide fewel, and dress and eat their provifions.

As the water is always either rifing or falling within the beds of the rivers, it is impoffible for a map to aflign precisely where a place of shelter fhall be found, at any given time. Thus much, however, may be concluded, that in a place where the junction of two confiderable channels is effected when the rivers are up, there will be an inlet, or deep bay, throughout the dry season, although one of the channels fhould be dried up. The waters (as we have faid before) are rifing from the latter end of April, to the middle of Auguft and falling during the rest of the year.

The navigation through THE WOODS, or SUNDERBUNDS, is effected chiefly by means of the tide. In the large rivers, or those

See page 207, and the fecond note in the fame page.

that

that communicate immediately with the fea, the circumftances of the tide are more analagous to the ordinary courfe of it, than in the fmall lateral channels which ferves to connect the great rivers together; the motion of the tide in those small channels, being regulated by the pofitions of their openings into the rivers. For if two rivers of equal bed and parallel courfe, are united by a lateral or crofs canal, the flood tide will enter that opening of the canal which lies nearest to the fea, and run through it into the other river; and the ebb tide vice verfa. But as the arrival of the tide depends on the capacity and formation of the rivers, as well as on the abfolute distance it has to run, it will not be an eafy matter to determine its direction at any given time, even with the help of a

map.

There are two diftinct paffages through the Sunderbunds, the one named the fouthern or SUNDERBUND PASSAGE, the other the BALIAGOT PASSAGE. The first is the fartheft about, and leads through the deepest and wideft rivers; and is of course, the moft expofed during tempeftuous weather. It opens into the Calcutta river, thro' Channel-creek*, about 65 miles below the town. The Baliagot Paffage opens into a lake on the eaft fide of Calcutta ; from whence, within a very few years, a fmall canal has been cut to join the lake with the river.

These paffages present to the imagination both a grand and a curious fpectacle: a navigation of more than 200 miles through a foreft, divided into numberlefs iflands by a continued labyrinth of channels, fo various in point of width that a vessel has at one time her mafts almost entangled in the trees and at another, fails uninterruptedly on a capacious river, beautifully skirted with woods, and affording a vista of many miles each way. The water is every where falt; and the whole extent of the forest abandoned to wild beats: fo that the shore is feldom vifited but in cafes of neceffity; except

*A part of this Creek forms the place known of late by the name of New-HARBOUR.

by the wood-cutters and falt-makers whofe "dreadful trade" is exercised at the conftant peril of their lives: for the tygers not only appear on the margin in queft of prey, but often, in the night time, fwim to the boats that lie at anchor in the middle of the river.

These paffages are used during the whole year, by those who go to and from the lower parts of the Ganges and Calcutta, &c: and during the feafon when the western branch of the Ganges is almoft dried up, the whole trade of Bengal (the western provinces excepted) paffes either by Channel-creek, or Baliagot, but chiefly by the former; fome articles of the Company's cargoes being brought more than 900 miles by water, at this feafon.

EXPLANATION of PLATE, N°. I.

A. A. A. Steep Banks, corroded by the current; the fragments of which are depofited, and form the Banks B. B. B.

C. A fand Bank, accumulating to an Ifland. This once joined to D; till infulated by a breach of the river.

E. An Island, formed and inhabited. This alfo was a fand-bank, thrown up round the point F.

G. An Island accumulating in the midst of the Channel.

H. The line of the strongest current.

I. A Gulf occafioned by the force of the current, from the oppofite fide, striking against the Bank: the origin of a future branch iffuing from the great river. In a course of time, the first reach `of it becomes retrogade to the course of the river (see note, page 263) as at K and L.

N. B. The fection of the Branch of the Ganges is exactly fimilar to that of the Ganges itself, except in the article of width.

CORRECTION of the GEOGRAPHY of the INDUS, and its DELTA, &c.

INCE the Memoir was printed, fome better information. re

SING

specting the country of Sindy, and the river Sinde* (or Indus) than what appears in page 80, has been moft obligingly communicated by a perfon of character, who refided fome time in that country, in the service of the East India Company. The delta, and course of the river, have in confequence, been corrected in the map but the form of the coaft has undergone no change, although the position of it, has; for Ritchel and Cape Monze are removed several miles further to the fouth, while their former diftance from Jigat Point is preserved. Cape Monze now stands in Iat. 24° 55', lon. 65° 46′: Ritchel, in lat. 24° 14′ (it was 24° 12′ by Capt. Scott's obfervations) and the mouth of Larry-Bunder river, which was the principal channel of the Indus, during the last century, and early in the present one, is in lat.. 24° 44′; being within one minute of the parallel affigned it, by the India pilot.

The city of Tatta, the capital of the province of Sindy, and fuppofed to be near the fite of the Pattala† of the ancients, is fituated, according to the idea of the abovementioned gentleman, about 38 G. miles to the north of Ritchel, and 50 to the cast of

• Mr. Wilkins makes the proper name of this river to be Stendhoo. Heetopades, page 333, + It is impoffible to fix the exact fite of Pattala, as there are properly two deltas, a fuperior and an inferior one; exclufive of the many iflands formed by the Indus when it approaches the fea. Tatta is near the head of the inferior delta; and the ancient accounts mention only one great delta, having Pattala at the upper angle of it. In Ptolemy's map (Afiæ Tab. XX) Pattala is placed very far below the place, where the Indus firft begins to feparate into branches.