The countries best known to Edrisi are those of Northern Africa, Spain, and Arabia, which are de scribed with minute topographical exactness. As we accounts of them, however, equally d , and more recent, the chief value is at Cached to that part of his work where he treats of the interior parts of Africa; regions into which, notwithstanding immense efforts, no European tra veller has yet been able to penetrate. In con sequence of the colonies and settlements formed there by several Arabian dynasties, Edrisi possess ed opportunities of information, of which religious estrangement has deprived Christian Europe. His description, however, cannot be -considered as throughout correct; or rather, it is so only'for a cer tain space; after which it deviates into the realms of imagination and conjecture. The leading feature delineated by him is that of the great central river, called the Nile V the Negroes, the Niger of Ptolemy and of the modems, which he represents to rise from the same fountain as the Egyptian Nile, and to flow westwards till it falls into the sea. phis statement has not been confirmed by modern discovery, but it is doubtful how far the modems may not have erred as to the regions designated by Ednsi, and even the precise import of the terms employed by him. From him, however, have been delineated, in all our mo dern maps, Wangara, surrounded by the branches of the Niger, which form in it two large lakes ; with the cities of Gana, Tocrur, Berissa, and Sala, situated along its banks. The eastern coast of Africa is de. scribed accurately as to the countries contained in it, but with a singular error of direction, being glade to run from west to east, parallel to the southern coast of Asia. The idea was probably suggested by the great unknown continent of Ptolemy, which was made to follow the same direction ; and it was ren dered almost necessary by our author's hypothesis, that no country could be habitable to the south of the line ; since such an extent of coast produced in a southerly direction must necessarily have passed that limit. Thus, however, Sofala is placed upon the sea of China, while Madagascar, under the name of Al Comr, or the Island of the Moon, forms part of the oriental archipelago. In Asia, Edrisi shows con siderable acquaintance with the central and southern regions, though he has not described them so fully as Abulfeda and other writers, who resided in Syria and Bagdad. The northern part is occupied by the fabulous abodes of Yagiouje and Magiouje, or Gog and Magog. Their territories were defended by a tremendous castle or rampart of iron, the idea of which was probably derived from some fortified pass in the mountains which stretch across central Asia.
In Europe, with the exception of Spain and Sicily, the information of Edrisi is neither accurate nor ex tensive. Germany is particularly ill delineated ; England he describes as a great Island, in its form resembling an ostrich, in which are seen " cities, lofty mountains, flowing rivers, and a level land, but where perpetual winter reigns." He names its capital and principal sea-ports. To the north is the island of Scotland, some features of which he men tions, though in a manner difficult to identify. He mentions also Ireland, and appears even to have a faint knowledge, of Orkney and Shetland.
The only valuable unpublished manuscripts of Edrisi which pow exist in Europe, are two that are preserved in the Bodleian Library. The first, which was brought over by Greaves from Egypt, is written in the Arabic character peculiar to northern Africa. It is illustrated by a map of the known world, and by thirty-three maps, containing each part of a cli mate, so that there are maps only for the three first climates. The second manuscript, brought by Po cocke from Syria, is written in the Arabic character, as used in that country, and bears the date of 906. Heg. (A. D. 1500). It consists of 320 leaves, and is illustrated by onegeneral and seventy-seven particular maps, the last consequently including all the parts of every climate. The general map has been published by Dr. Vincent in his Periplus of the Egyptian Sea. The writer of this article has examined these and found them to throw considerable light upon the geographical system adopted by the Arabian writers.
There is a manuscript (Cod. DLXXX) in the Royal Library at Paris, which professes to be the production of Edrisi ; but D'Herbelot, it appears, has not made use of it as such ; and De Guignes ex presses positive disbelief on the subject. Hartmann, however, found it to coincide, in many particulars, with the geography of Edrisi. A copy of our au thor's work was contained, at one time, in the library of the Escurial, but it was destroyed by a great fire in 1671.
In 1619, two oriental scholars, Gabriel Sionita and John Hezronita, published at Paris a Latin tran slation of Edrisi's work, bearing the title, as already observed, of Geographia Nubiensis. It is not exe cuted with all that care and accuracy which might have been expected from these learned personages. They have been particularly careless as to the proper names, which are given sometimes in Latin, some times in Arabic, sometimes in neither, nor in any intelligible language. In consequence of the Arabic original being occasionally appended, it has been supposed to have been reprinted along with it; but this is a mistake, there being no edition ever printed except the Roman one. George Hieronymus Vet schius, an eminent German scholar, had prepared a copy of the Arabic original, with a Latin translation, which he intended to have illustrated with notes; but death prevented the execution of this design ; and his manuscript remains deposited in the library of the University of Jena. Casiri (Bibliotheca Arab. Ilispan. II. 18.) mentions that, at the request of many friends, he had determined to re-edit this work, but he never appears to have executed this in tention. The part relating to Africa, pre-eminent, certainly, in point of importance, has been very ably edited by Hartmann, Professor of Oriental Lan guages at Marburg (Edrisi, Africa, Gottingen, 1796). Instead of following the awkward division into cli mates, he has collected together all the notices relating to each particular country ; and has an nexed the statements of the countrymen and con temporaries of Edrisi, so that his work forms nearly a complete body of Arabian geography, so far as relates to Africa. It does not appear, how ever, that he had an opportunity of consulting the Oxford manuscripts, or any others of importance, by which the deficiencies of the printed edition could be supplied. A complete edition of this work is, therefore, still wanting ; and the learned world may naturally expect it from the University of Oxford, which possesses, in Mr, Nicol, Junior Librarian of the Bodleian, a scholar well qualified for the task, and who promises to support the long-established fame of the University in oriental literature.