MEMOIR

OF A

MAP OF HINDOOSTAN, &c.

S

great an extent of country is contained in this map, and the quality of the materials is fo various in different parts, that it became neceffary, in order to prevent confufion, to divide the account of its conftruction into separate fections, agreeable to the natural divifion of the country; and, in fome measure, to the nature of the materials. It is accordingly divided into seven sections: The first contains the fea coafts and islands.

The second, the furveyed tract on the fide of Bengal; or that occupied by the Ganges and its principal branches, as far west as the city of Agra.

The third, the tract occupied by the Indus and its branches.

The fourth contains the tract between the Kiftna river and the countries traversed by the Ganges and Indus; that is to fay, the middle parts of India.

The fifth contains the peninfula fouth of the Kistna.

The fixth, the countries fituated between Hindooftan and China; namely, Thibet, Bootan, Affam, Pegu, Aracan, Ava, and part of Siam: and

The feventh, and laft, contains Tables of diftances between the principal cities, &c.

But, before I proceed to the particulars of the conftruction of the map, it will be neceffary to explain the itinerary measure adopted

in places where no furveys have been taken. The ufual measure of this kind in Hindooftan, is the cofs, or crores, commonly estimated at two British ftatute miles. I have not been able to get the true length of the cofs, as fixed by Acbar, and other Emperors; and, even if I had, it would be of no ufe in the prefent enquiry, as all my Hindooftanny itineraries and tables are in computed coffes.

It may reasonably be expected, that in a country of half the extent of Europe, the estimated length of the itinerary measures, although of the fame denomination, must vary in different parts of it. It is no more than what happens in different provinces of the fame kingdom, in Europe. But as far as we have any data for making a just comparison, the cofs does not vary fo much as one-fixth part over the whole country; and between the northern and fouthern extreme of India, (that is, in an extent of about 1700 miles) the difference is not more than one-fixteenth part. The miles vary much more in their proportions, in the different parts of Europe.

Taking the medium of the cofs throughout Hindoostan, and the Deccan, there will be about 40 of them to a degree of a great circle on the globe: that is, each cofs is about a geographical mile and half. But this is to be understood of horizontal measure; in which the windings and inflections of the roads are allowed: for the estimated routes could not be applied to geographical purposes, by any other rule. The cofs, in road measure, is about one statute mile and nine-tenths; or at the rate of 190 British miles to 100 coffes; one part in feven, being allowed for winding, when the line of distance is extenfive. Or, feven miles of road measure, are allowed to produce fix miles horizontally, or in a direct line.

In Malwa and its neighbourhood, the coffes are larger than any where else, and are about 1,7 geographical miles, or 35 to a degree. And on the road from Baglana to Mafulipatam, they are fo short, that 46 are required to make a degree. But having only one example for the latter proportion, I shall found no rule on it. The proportions that I have adopted for Hindooftan, Malwa, and the

Carnatic,

Carnatic, from a great number of examples, are refpectively 1,43; 1,71; and 1,6 of geographical miles to a horizontal cofs; or 42, 35, and 37 to a degree of a great circle. The cofs of Hindooftan proper, is therefore fhorter than any other, and prevails throughout the greatest extent of country. There is again in Nagpour (the ancient Goondwaneh) a Goondy cofs, which by the medium of all the accounts I could get, is about 2,76 geographical miles, reduced to horizontal distance; or 21,9, or 22 to a degree. This meafure appears to be in ufe by the natives, throughout Mundilla and Boggilcund, as well as in Nagpour; and fometimes occafions great confufion in the reports of the coffids, or couriers: however, they have a computation of Hindcoftanny coffes alfo, in the fame country; and the proportions agree in general remarkably well with that fcale, between the Bengal Provinces and Aurungabad; and between Mundilla and Hydrabad.

Having mentioned the windings of the roads, it may not be improper to give the refult of my enquiries on this head, for the benefit of those who may have itineraries, kept in eftimated distances, to work up. One in feven is allowed as above: and is what will be found to take place in large diftances, in fuch countries as are interfected by deep rivers, or watercourfes or in fuch as have no artificial roads; and where thofe on the natural level, have obstacles to furmount, The degree of winding of roads, in different countrics, is, (cæteris paribus) according to the ftate of improvement, in which the roads are. In India, the roads are at beft, little better than paths, and whenever deep rivers, (which in that country are frequent, and without bridges) moraffes, chains of mountains, or other obftacles, oppose themselves to the line of direction of the road, it is carried round, fo as to effect the easiest paffage; and for this reason the roads there, have a degree of crookedness, much beyond what we meet with in European countries, where bridges are laid over every confiderable watercourfe, and where hills are either levelled, or reduced to a convenient degree of acclivity; and after

all,

1

all, expences faved in many cafes, by the difference of labour between the fmoothing of the direct road, and the forming of a road on the natural level. But the proportions, muft of course vary with circumstances; and may be only one in ten, in a dry, open, country, and one that has a tolerably even furface: but this happens too rarely to found any general rule on. As the line of diftance increases, a greater degree of winding will take place; or, a fhort distance will always be on a ftraighter line than a long one: for in countries where the management of the roads is not arrived at a high degree of perfection, the road through a kingdom will be made up of portions, confifting of the particular roads leading from one city, or principal town, to another; although they may not lie in the general line of direction; and then there will be a general winding, added to the particular one: and the above proportion of 1 in 7, is applied to this compound winding. And, added to this, in very long distances, fome natural obftacle, will, very probably, oppose itself: an arm of the fea; a river of difficult paffage; a morafs; or an impaffable ridge of mountains; and change totally the direction of the road: whilst the parts, on each fide of the obftacle, might have but an ordinary degree of winding: and it is seldom, but that one or other of these, occurs in the space of 100, or 150 miles. Probably 1 in 8* may be a pretty juft general proportion for distances of about a 100 miles: that is, 8 miles by the road, will be feven direct; or what is commonly termed birdflight and where the extent is from 200 to 300 miles, 1 in 7.

Measured distances in Hindooftan, do not often occur, where, at the fame time, the true horizontal diftance is given, except in Bengal and that is a country too full of deep rivers, lakes, and moraffes, to serve as a general standard. In the Carnatic, a dryer country, the medium of winding, in distances of about 100 miles, is 1 in 9. In England, as far as we can trust the maps (which may

This is M. D'Anville's idea, p. 45 and 46 of his Confidérations Géographiques.

be

1

be done, where the distance confists chiefly of difference of latitude) I in 11 is the proportion, in diftances of about 100 miles; and in very great distances, fuch as Edinburgh, 1 in 7.

It may happen that the direct route may lie through a defert or an ill-governed country; in which cafe, travellers will avoid the way, in which famine, or robbery, threatens them, and by thefe means be carried out of the true line of direction: but it is obvious that no rule can be given for fuch cafes. Upon the whole, the degree of winding, as far as depends on natural causes, must be estimated by the compound ratio of the length of the line of distance, and of the nature of the country, as to evenness, dryness, and openness. And of course, fome local knowledge of it will be required, in order to correct the distances in a just degree *.

The term cofs is of high antiquity; and that of coffid, or courier, appears to be derived from it. It seems that the measure of the coss, established by the different Emperors of Hindoostan, has varied confiderably at different times; and has always been longer than the computed one. That fixed by Acbar appears to have been about 2 British miles, and a fixteenth. But of this, I have no farther proof, than what refults from the comparison of the different measures of the road between Patna and Moorfhedabad; being a portion of the great road from Delhi to Bengal, measured by order of one of the Emperors.

M. D'Anville concludes his enquiry into the length of the coss, by determining the number in a degree, on a medium, to be 37; but it must be observed, that he had no measured lines with which he could compare his estimated distances. On the other hand, in the refpective distances of Candahar, Cabul, and Attock, as described by him, each degree contains 47 of Tavernier's coffes.

Those who wish for a general rule for changing horizontal distance into road distance, in their common references to maps; may break the line of diftance, (if very long) into portions of not inore than 100 or 150 miles; and then add to the whole fum of the distances, fo obtained, one eighth part. Thefe portions fhould be contrived, fo as feverally to include the spaces, between the points, that diverge moft from the general line of direction of the whole road. By this means, the errors arifing from the compound winding, will be avoided.

+ Eclairciffemens, p. 14.

SECT.