It is proper in this place to restate and correct the line of distance between Gana and Dongola; which will be found in the Proceed. Afr. Assoc. for 1798, p. 122. There it appears, that Kauga is given by Edrisi, at 30 journies, equal on his scale, to 570 G. miles from Dongola; and it is found by Mr. Browne’s statement, to fall at 578. Again, Mr. Horneman says, that Fittré, (or Kauga,) is 40 journies to the eastward of Kashna, (See above, [p. 138.]) This interval of distance, taking Kashna as it is placed in the Map of 1798, is 653 G. miles; equal to 16⅓ per day; and is perfectly satisfactory. Edrisi allows 36 journies between Gana and Kauga; so that by this account, Gana ought to be nearer to Kauga, by four journies, than Kashna is. In the Map of 1798, Gana is placed eight miles too far to the east. (See the Proceed, p. 121.) This corrected, Gana will be 82 from Kashna; which, however, is still too much for four journies, and agrees better to five.
The interval thus corrected, between Gana and Kauga, is 575 G. miles in direct distance; which allows no more than 15½ per day, for Edrisi’s 36 days; whilst his ordinary scale is 19. If therefore, the report of Edrisi is right, Gana must either lie more to the west, or the course of the Niger, along which his route leads, to within ten journies of Kauga, must form a very deep curve to the south, after passing the dominions of Houssa, (of which Gana makes a part.) That it does decline to the south, Mr. Horneman was repeatedly informed, as may be seen above, pages [115,] [117,] but whether in a degree sufficient to occasion the difference above-mentioned, is not ascertained.
M. D’Anville also had an idea, and so describes it in his Map of Africa, 1749, (possibly from actual information), that the Niger declined to the south beyond Gana; so that the termination of it, in the lake of Semegonda, was 3½ degrees of latitude to the south of Gana; and which, by the bye, would place the lake of Semegonda, about the same parallel with the Bahr Hemad of Mr. Browne. At present, however, we must be content with proportioning the distance between Gana and Kauga; which, admitting a curvature to the south, allows a rate of 16⅓ or 16½; or that of ordinary caravan travelling. It has been shewn that Horneman’s report of the distance between Kashna and Fittré, is very consistent: and there can be no question, that, of the two, we ought to prefer Horneman’s report, from its being less subject to error than the other, which has passed through so many hands, in its way to us.
[58]The Thames, between Maidenhead Bridge and Mortlake, (about 41 B. miles in a straight line,) has a fall of rather more than one foot eight inches in each mile. But this is a smooth part of its course: for although the Thames does not, by any means, spring from very elevated lands, it cannot be supposed to fall less than four feet in each mile, taken on its whole course. On this idea, the Niger should have a fall of more than 8000 feet, in its supposed course to the White River. But even if two feet and a half (which appears very moderate indeed) be allowed, the fall would be no less than 5625 feet, or 115 yards more than an English mile. Is it to be credited, that the bed of the White River, on the south of Darfoor, and at a point not very remote from its source, is a mile lower than the springs of the Joliba, or Niger?
[59]Browne’s Travels, page 473.
[60]Edrisi says, “Ab urbe Ghana ad primos limites terræ Vancáræ est iter octo dierum (orientem versus)—Atque hæc eadem est insula longitudinis trecentorum milliarium, centum et quinquaginta latitudinis, quam Nilus undequaque circundat toto anni tempore. Adveniente verò mense Augusto, et æstu gravescente, Niloque inundante, insula ista vel certè major pars illius aquis obruitur, manetque sepulta aquis quamdiu Nilus terram inundare consuevit.” (Sionita, p. 11 and 12. See also Hartmann’s Edrisi, article Vankara, p. 47 et seq.
[61]Ebn Haukal, an Arabian geographer, of the tenth century, (lately translated by Sir William Ouseley,) gives the following account of the river Heermend, p. 205.
“The most considerable river of Sejestan is the Heermend, which comes from Ghaur to the city of Bost, and from that runs to Sejestan, to the lake Zareh. This lake is very small when the waters of the river are not copious; when the river is full, the lake increases accordingly. The length of this lake is about 30 farsangs, (about 110 B. miles,) and in breadth about one merhilch. (Day’s journey, or 24 B. miles.) Its waters are sweet, wholesome, &c.”
It is well known, that there is no outlet from the lake Zareh.
[62]Begama of Edrisi.