32° and conclude, of course, that the five cities beyond it were situated along the coast of the province of Sus, and in the bay southward of Cape Nun: but certainly short of Cape Bojador.

Scylax of Caryandra says, that the distance is 12 days' sail from the straits to Cerné: that is, two to the Promontory of Hermæus, three thence to Soloeis; and seven more to Cerné. This requires a rate of 104 miles per day and is nearer to that, which might be expected from a modern ship, than an ancient one. But his Periplus within the straits, gives a rate which is generally not very different from that of other ships of those days; or about 36 G. miles'. It must be allowed that the many examples adduced, ought to have more weight, than those alone, between Carthage and Cerné, even if the general rate of Scylax did not accord with the rest. Besides, the Periplus of Hanno above quoted, furnishes the strongest presumptive proof that the rate of Scylax did not ex

The southern province of the kingdom of Morocco. For instance, he says it is 754 days' sailing from Canopus to the Columns, tracing the sinuosities of the coast; we conclude, according to the usual mode of coasting. This gives a rate of 32 G. miles per day, and Hanno's rate between Cerné and the Gambia is 35, on 14 days' sailing. At the same time it must not be omitted, that Scylax says, that the voyage between Carthage and the Columns may be performed in seven days and nights, with a favourable wind. This requires a rate of 107 for each day and night, and is not very different from the 12 between the Columns and Cerné. But the same authority allows generally 36 only, within the straits, in a variety of instances. How are these accounts to be reconciled?

ceed, but rather fell short, of that Periplus. For, this latter places Thymiaterium at two days' sail from the straits; and Scylax allows the same distance between the straits, and the Hermæum Promontory, which place he describes to be short of Thymiaterium; consequently, by his account, this place must be more than two days' sail from the straits. He then reckons three more days to the Promontory of Soloeis; which distance, according to the same proportion, can only reach to Cape Blanco; but nevertheless, considering the vague nature of his description, Cape Cantin may have been intended. As to the remainder of his chain of distance, it is not worth regarding; as he reckons only seven days between Soloeis and Arguin.

Pliny says, (lib. v. c. 1.) that the river Lixus, (that is, the northernmost of the two, of that name, and the Lucos of the present time; a position well known;) is 57 MP. from Tingi, (or Tangier ;) and Rutubis 313 MP. farther and he adds, that still farther on, is the Promontory of Solis. The 313 will reach to Saffy; allowance being made either for the inflexions of the land route, or those of a coasting voyage. Hence Saffy may be taken for Rutubis, or rather Rusibis Portus, as we find it in Ptolemy, who places it within 10 minutes of the true latitude of Saffy. The Promontory of Solis

In two numbers, 25, and 32. The Itinerary has 54 MP. between Tingi and Lix, which differs little from Pliny, and is * justified by the distance of the Lucos R. from Tangier.

Pliny says that another authority gave 112, which must be a mistake.

then, is by Pliny's account, to be looked for beyond Rusibis, or Saffy; although it is not said how far: but Pliny could not at any rate have had Cape Bojador, which is about 6 degrees to the south of Saffy, in contemplation. And as Ptolemy has a promontory named Solis Mons, at about 70 miles to the southward of Rusibis, we may suspect that it was intended for the same place as the Solis of Pliny; and possibly too, for the Soloeis of Herodotus, of Hanno, and of Scylax, though somewhat misplaced. Nor could Ptolemy have had Bojador in contemplation, because his Solis Mons is placed four parts in five nearer to Atlas Minor, than to Atlas Major; or in other words, to Cape Cantin, than to Cape Bojador.

If we may regard the Solis Mons of Ptolemy, as the Promontory of Soloeis itself, this may be reckoned a positive notice respecting its situation; and indeed, the only one that occurs; although the presumptive evidence of Hanno and Scylax is very strong. But there is some difficulty in supposing that the promontory intended by the above writers, formed any part of the comparatively straight coast, which is found between the Capes of Cantin and Geer, when the characteristic distinction of Soloeis seems to have been prominency, beyond the line of the coast to the northward of it.

There are few parts of Ptolemy's geography, in which the latitudes agree so well with the modern observations, as in the part between the Strait of Gibraltar and C. Bojador. In effect, there is a remarkable coincidence in many points, as will ap

pear by the subjoined table; so that this part of the coast must have been much frequented; but, it

9

• Comparison of certain parallels, in Ptolemy, with the

modern observations, and charts.

* We cannot help regarding the Paa Island of Ptolemy in lat. 32°, as being intended for Madeira. The latitude differs but little, but it is certainly too near to the coast of Africa, by many degrees of longitude. But as the Fortunate Islands were known to Ptolemy (his Erythia, in 20°, must have been intended for one of them, probably Fortaventura), what island so far to the north as 32o, could have been meant, but Madeira? This conjecture, in our idea, is rendered more probable by the description which Diodorus (lib. v. c. 2.) gives, of a large island, fertile, well wooded and watered, and situated at many days' sail to the west

is

very remarkable, that, although the parallels are so generally exact, the bearing is out full four points of the compass; it being nearly S b. E, in Ptolemy, when it is in reality about SW b. S1. And hence it may be collected, that, when the latitudes could not be applied to the correction of the bearings, the ancients formed very erroneous calculations of them 2. But this does not, in the present case, destroy the harmony of the positions, in respect of each other, so far, as to prevent them from being recognised, by means of the general resemblance of the figure of the coast, combined with the parallels.

But to the south of Atlas Major (Bojador) the latitudes are not only in general wrong, but the figure of the coast loses all resemblance to the truth, until we come to Cape Verde: for even Cape Blanco (of the Sahara), which is the most prominent part of the coast, recedes, in the descriptions of Ptolemy, within a direct line drawn from Cape Bojador to Cape Verde. This latter is also about 23 degrees too far south; the mouth of the Gambia 21 degrees; and that of Senegal nearly a whole degree.

By a reference to the comparative table of latitudes (in the note) and to the geography of Ptolemy,

of the coast of Africa. It was said to be discovered by certain Phoenicians, who were blown by a storm into the Atlantic, as they were coasting Africa.

1 Between C. Spartel and Bojador, the diff. lat. is 568, and the departure 410; whence the bearing is about S 36° W : Ptolemy has S 10° E: whence the error is about 46°.

2 Thus, the eastern shore of the Mediterranean lies nearly NE and SE, in Ptolemy; instead of about N b. E, as it ought to be.