geography affords.  His description of Arabia is, on the whole, accurate; he has, however, greatly diminished the extent of the Arabian Gulf, and by at the same time increasing the size of the Persian, he has necessarily given an erroneous form to this part of Asia.  The ancient opinion of Herodotus, that the Caspian was a sea by itself, unconnected with any other, which was overlooked or disbelieved by Strabo, Arrian, &c. was adopted by Ptolemy, but he erroneously describes it as if its greatest length was from east to west.  The peninsula to which he gives the name of the Golden Chersonesus, and which is probably Malacca, he describes as stretching from north to south:  to the east of it he places a great bay, and in the most distant part of it the station of Catigara.  Beyond this, he asserts that the earth is utterly unknown, and that the land bends from this to the west, till it joins the promontory of Prasum in Africa, at which place this quarter of the world terminated to the south.  Hence it appears that he did not admit a communication between the Indian and Atlantic oceans, and that he believed the Erythrean sea to be a vast basin, entirely enclosed by the land.

Strabo and Pliny believed that Africa terminated under the torrid zone, and that the Atlantic and Indian oceans joined.  Ptolemy, as we have just seen, rejected this idea, and following the opinion of Hipparchus, that the earth was not surrounded by the ocean, but that the ocean was divided into large basins, separated from each other by intervening land, maintained, that while the eastern coast of Africa at Cape Prasum united with the coast of Asia at the bay of the Golden Chersonesus, the western coast of Africa, after forming a great gulf, which he named Hespericus, extended between the east and south till it joined India.  The promontory of Prasum was undoubtedly the limit of Ptolemy’s knowledge of the east coast of Africa:  the limit of his knowledge of the west coast is not so easily fixed:  some suppose that it did not reach beyond the river Nun; while others, with more reason, extend it to the Gulf of St. Cyprian, because the Fortunate Islands, which he assumed as his first meridian, will carry his knowledge beyond the Nun; and because, at the Gulf of St. Cyprian, the coast turns suddenly and abruptly to the east, in such a manner as may be supposed to have led Ptolemy to believe that it stretched towards and joined the coast of India.

Of some of the interior parts of Africa Ptolemy possessed clear and accurate information; regarding others, he presents us with a mass of confused notions.  He clearly points out the Niger, though he fixes its source in a wrong latitude.  In the cities of Tucabath and Tagana, which he places on its banks, may perhaps be recognized Tombuctoo and Gana.  The most striking defect in his geography of the interior of Africa is, that he does not allow sufficient extent to the great desert of Sahara, while the southern parts are too