1

marks of towns, and other objects of consequence. Two oblong squares are introduced-one of Sierra Leone, and the other of Lower Egypt; but the first is of little consequence, and might well be taken out, to make room for a new title; while the fragment of Brazil, and all beyond the longitude of the Cape Verd Islands, should be sacrificed.

This map, as we have already mentioned, having few pretensions to superior merit, we shall be contented with a brief review of some objects that excite remark, beginning with the states of Barbary, and proceeding towards the east and south. Morocco is laid down by d'Anville at too great a distance from the sea, as is evident from the journeys of Lempriere and Hoest; yet in this error he has been followed by major Rennell and the constructor of the present map. The chain and direction of the celebrated Mount Atlas are also wholly erroneous, though not equally so with d'Anville's delineation; the chief defect of whose maps is the vague and arbitrary representation of those great features of nature, which have had more influence on history, and even the characters of nations, than any other feature of natural geography, and ought of course to be traced with the greatest care. In general, this part of the map is copied from d'Anville, but is not so rich in names. In Tripoli there are several improvements; yet the geography of that part of northern Africa remains very imperfect; the country not having been pervaded by any scientific traveler. Egypt affords trivial room for observation; and in Nubia little has been added to the geography of d'Anville. Abyssinia has become a trifling theme; and Bernier's Map of the Sources of the Nile, 1666, contains the most essential features.

From Adel to Mozambic, the African coasts are obscure; the Portuguese, who have settlements in that quarter, being far behind the other European nations in every branch of science. The constructor of this map has rather too boldly laid down some chains of mountains in southern Africa: and his Ethiopia is an arbitrary denomination-the Ethiopia of the ancients being chiefly Abyssinia; while the modern extension of the appellation is alike vague and useless, as it conveys no determinate idea. His other denominations in this quarter are extended much beyond the bounds of real knowledge, and are rather calculated to fill up the map than to convey the just representation of great deficiencies.

In the copy before us, the bank of Aiguilhas, to the south of the Cape of Good Hope, is ludicrously coloured as a sandy desert; so that an ignorant inspector would be led to believe the bays are visited by camels, and not by ships. In the regions round the Cape, too much confidence has been placed in the authority of Vaillant, whose pretensions have been detected by Mr. Barrow. The course of the Orange River, in

particular, is wholly erroneous; and the Fish River (a name which indeed means nothing) may be a fabrication of the French adventurer. We believe the Burning Desert does not, burn more than the others; but this map will certainly not burn from the concentrated irradiations of science. Congo, &c. are copied from d'Anville; yet there are many omissions; and the publisher has been sparing in the expense of letter engraving. There are also some mistakes which evince that the constructor is not versed in the original authorities; but it would be degra-, ding to criticise the present map as a learned production; and we shall reserve such a review for some future capital map of Africa. Suffice it to remark, as a solitary instance, that the learned constructor has placed the country Gabon far to the north of the town of Majumba; while far to the north of Gabon is marked the country of Majombo. The plain fact is, that Majumba, or Manijumba, only means the King of Yomba—a small district round the town, here called Majumba.

The idea here expressed, that the kingdom of Dahomy is the Dauma of Leo, may deserve attention; but that the people of Lamlem are Jews seems a futile and unwarrantable notion. We need not pursue the well-known coast of Guinea, nor the rivers Gambia and Senegal; after which we arrive at the Great Desert, and the kingdoms of Barbary, whence our progress began.

In the central parts, the constructor has chiefly followed the opinions of major Rennell, or rather endeavoured to combine them with those of d'Anville; to whose delineation the major has been obliged to recur, after having expressed, in his map prefixed to the papers of the African Society, 1790, the Niger running from the east towards the west. The constructor of the present map is certainly little enabled to improve upon d'Anville, in countries where more recent discoveries have not been made: but, to display some appearance of learning, he has introduced some references to Ptolemy, which only serve to show his unacquaintance with ancient geography. He gravely supposes, for instance, that the Coloe Palus of that ancient geographer, which every school-boy knows is the lake of Dembea, and which a bare inspection of the maps in Ptolemy's Geography would suffice to identify, gives source to the river Kulla, fifteen degrees to the west; while the river Kulla itself must forsooth be called Coloe! A more puerile supposition, arising from a supposed resemblance of names, and a more gross and radical error, certainly never disgraced any geographical production. We have already expressed our opinion that the river Kulla of Browne is the only stream yet detected in that quarter which can at all correspond with the Gir of Ptolemy: but our constructor supposes the Gir to be the Bahr Misselad. This last seems indicated by Ptolemy in the river

which runs into his Nuba Palus, rightly delineated to the N.E. of the Gir, though the direction of the stream be westerly instead of northerly. So much for the only points in which the constructor seems to aspire to original learning, or indeed the only point; for that relative to the Gir may be faintly traced in d'Anville, who supposes that the Gir or Nile of the Negroes of Edrisi runs north into the lake of Kauga. But the course assigned by d'Anville, and the present constructor, is far too short for the great river Gir of Ptolemy, which seems obstinately rejected by modern geographers. In a recent German map we have seen the river Kulla of Browne delineated as entering the sea at Calabar. But Mr. Browne expresses his opi nion that this river pursues a north-westerly course; and the bare inspection of his map might instantly have satisfied any scientific constructor that the river Kulla alone could correspond with the Gir of Ptolemy.

The chief merit of the present map is, that it assembles in one point of view all the recent discoveries concerning Africa; but the interior parts are laid down with a deceitful appearance of certainty; and the constructor is assuredly not sufficiently imbued with geographical knowledge for the difficult task which he has undertaken.

Smith's New English Atlas, being a complete Set of County Maps, &c. Sold by the Author.

WE have already mentioned this Atlas, and the terms on which it is to be published. There are six numbers now before us, containing eighteen maps, executed with tolerable care and neatness. But the waste of paper around each county seems sufficiently to proclaim that the best mode of executing an English Atlas would be to publish it in sheets, filled to the edges, which, with a proper index map, might be as easily consulted, as if divided into counties. This last plan has also many conveniences well known to every practical geographer, without one advantage to balance them; and we would wish to see it finally dismissed from geography both at home and abroad, when an Atlas of a whole kingdom is intended. Particular surveys of distinct counties or provinces will still retain their claim of applause, when executed by able hands.

It may not be unnecessary to repeat, that the price of each number of Mr. Smith's Atlas, containing three maps, is 65. 6d. coloured, and 8s. stained. It is expected to be completed this

autumn,

Cary's New English Atlas, in a complete Set of County Maps, &c. No. I. Price 75. outlined, and 8s. full coloured. Sold by the

Author.

ON comparing this with Mr. Smith's, we sometimes find more care displayed in Mr. Cary's; as, for instance, in Charnwood Forest in Leicestershire. Mr. Cary's map of Bedfordshire is also far superior to Mr. Smith's in the delineation of the hills. In strict impartiality, therefore, we must prefer Mr. Cary's, however inclined we may be to favour Mr. Smith's spirit and industry. But an equal objection of waste of paper, and want of concatenation of parts, arises against both; and it is singular that neither should have alighted on the only legitimate plan of executing an Atlas of a kingdom.

Chart of the Baltic Straits; namely, the Great Belt, the Little Belt, and the Sound; including the South Part of the Kattegat, with the Western Part of the Baltic Sea, and all the Danish Isles from the Surveys made by the Swedish Admiral Nordenanker, and Professor C. Lous, Director of Navigation at Copenhagen. 75. 6d. Faden.

THE survey by admiral Nordenanker is universally esteemed the most exact which has yet appeared; and the present excellent chart is constructed with great care upon that survey. There can be no doubt that the shores, shoals, &c. &c. are laid down with the utmost accuracy; and it is singularly rich in the observation of soundings.

Chart of the Sleeve, or Gulf of Jutland, and of the North Part of the Kattegat; from the Surveys of the Swedish Admiral Nordenanker, aad those made in Denmark and Norway by Professor Lous, Director of Navigation, and by the Royal Engineers. 75. 6d. Faden.

THIS sheet is connected with the former, being the northern part of the same chart. It includes a supplement of the Sound of Christiana. The praise bestowed on the preceding may be extended to this.

Chart of the Coasts of South America, from Rio de la Plata to Cape Horn, and from Cape Horn to Valparayso, including the Isles of Juan Fernandes. Two Sheets. 15s. Faden.

To the title of this chart is annexed the following infor

mation:

The Spanish chart from which almost the whole of the present is a faithful copy has been geometrically surveyed by royal order in the years 1789, 90, 94, and 95; and presented to his Catholic majesty in 1798 by Don Juan de Langara, minister of marine. That excellent original extends only from the parallel of 36° 30' of south latitude to Cape Horn, and goes not beyond the meridian of 76° 42' west of Greenwich. A greater extension given to this copy in latitude and longitude has enabled us to insert the Isles of Juan Fernandes, which are of some interest to our navigators in the South Sea. We have added, besides, various plans of harbours and roads, appearances of land, tracks of ships, with their soundings, &c. and several other particulars. The improvements made on the south coast of Tierra del Fuego, and omitted by the Spanish hydrographer, might be equally mentioned, could their small importance make them remarkable in a series of discoveries which so eminently increase our geographical knowledge.'

The Spanish survey here mentioned must be that by Malespina, whose name, we understand, is, on account of some slight difference or jealousy, affectedly passed in silence by the Spanish council of marine. The present chart is extremely interesting, as it is founded on the first correct delineation of a great part of the shores of the new continent; and it varies considerably from the charts and maps before published. In the first sheet there are plans and views of Juan Fernandes, &c. with a small chart of the north part of Saint George's Bay, and others of Sea-Bear Bay and Port Saint Elena. The southern sheet also contains several views of land and plans of ports. A remarkable feature of the southern part is the large island of Campana, lat. 48°, and the Campana channel, which divides it from the main land of New Chili.

Cork Harbour; surveyed by John Knight, Esq. Rear-Admiral of the Blue. 5s. Steel.

THIS chart seems to be laid down with attention and accuracy. It is accompanied with short sailing directions, and with six views of the entrance of Cork harbour and adjacent lands.