The map now before us may be regarded as the best yet published of Italy; as, to the advantages of d'Anville's learning, are superadded a larger scale and great improvements from the authorities mentioned in the advertisement. After this just commendation, we may be permitted to remark a most glaring defect in the delineation of the mountains, which are laid down on the antiquated and inaccurate plan, now reprobated by every skilful geographer, as wholly remote from the natural formsthis map being sprinkled with a number of mole-hills in every direction, intended to represent the grand chains of the Alps and Pyrenees! This adherence to a method now justly exploded must either proceed from want of information in the progress of geology and the modern art of constructing maps, or from a prejudice in favour of the ancient manner-alike deserving of reprehension, as it tends to impede the progress of the science in this country. It is wholly impossible from this map either to judge of the direction of the main ridges, or to estimate their comparative heights. Upon comparing the Alps, as here laid down with those in the maps of Switserland by Weiss or Mechel, the campaigns of Bonaparte by Dalbe, or even the south-eastern parts of France reduced from Cassini by Rochette himself, the vast superiority of the recent plan will be instantly perceived. Nor can the smallness of the scale be urged as an objection; for any engraver of common skill could easily reduce these maps to a small size; in which case the chief and leading ridges would alone appear, thus preserving those very parts which deserve to be prominent. It was doubtless far easier for the constructor to retain the antiquated mole-hills of d'Anville than to lay in the mountains in the just and scientific manner adopted by recent geographers; but the consequence of this want of knowledge or industry may easily be foreseen in the superior merit and success of any map that should add these advantages to others contained in the present publication.

We have neither time nor inclination to enter into a minute criticism on this production. A few mistakes might be pointed out in the spelling, as Gemini for Gemmi, Difentis for Disentis, Macagnana for Macugnaga, &c. &c. &c. The confused manner of laying in the mountains also impedes the facility of consultation. In Sicily, for instance, though generally reputed a plain and fertile country, it is difficult to trace the names of places, or even the courses of rivers; because, while it is ludicrously divided into its three famous Vales, it is so completely covered with strings of beads, in imitation of mountains, that, upon comparing it with the Alps in this very map, it will be found. by far the more Alpine country of the two! This singularity be comes the more apparent on a collation with d'Anville's own Map of Italy, in which Sicily is laid down with far more truth and exactness than it here appears on a scale of double the size. This APP. Vol. 34. 2 R

striking instance may perhaps serve finally to explode the random caterpillars, or bead strings, formerly used to represent mountains, and which any constructor of maps was accustomed to introduce at his own pleasure, like the elephants and lions of ancient geography, while, in the Oriental style, he called on the

mountains to cover his sins.

Lower Egypt and the adjacent Deserts, with a Part of Palestine; to which has been added the Nomenclature of the Roman Age. By L. S. de la Rochette. 75. 6d. Faden. 1802.

THIS map will be found to vary considerably in many important respects, particularly the shape of the lakes at the mouth of the Nile, from the map of d'Anville. No doubt M. de la Rochette has not only used the map of Niebuhr, but also such materials as are supplied in the recent French memoirs. Yet this production would no doubt have received superior advantages if its appearance had been delayed till the drawings taken by the English officers could have been collected and compared. A late work of the French general Reynier, accompanied with a map, might also have been found useful. We are informed that the longitudes here assigned to Suez and Cairo are liable to objections, which will of course affect many of the other positions. The form of the lake of Edko or Etko, and the canal of Alexandria, seem also to be inaccurate; the latter, on an actual survey, having been found to pursue a very winding course. There also seem to be some errors in the French maps concerning the natron lakes, of which Mr. Browne as well as other travelers only discovered two.

We need not examine what M. de la Rochette terms the nomenclature of the Roman age, as on this subject it can hardly be expected that he should have added any thing to the discoveries of d'Anville and other able illustrators. This map, which is delineated with care and neatly engraved, embraces a part of Palestine extending beyond Acca, corruptly called Acre, in the north, and Jerusalem in the east, thus apparently intended to illustrate the marches of Bonaparte, as well as the British transactions in Egypt. But as we expect many new elucidations concerning Lower Egypt from the memoirs and surveys of several British officers, we shall delay till a future period a more ample investigation of the improvements that may be made in the geography of that interesting country.

A New Map of the Island of Trinidad; made by Order of His Excellency Sir Ralph Abercromby, K. B. Lieutenant-General and Commander in Chief of the British Forces in the West-Indies. By F. Mallet, Captain of the Surveying Engineers. Four Sheets. With a short Description and reduced Map. 11. 75. Faden.

1802.

THIS beautiful map has been published in consequence of the cession of that important island by Spain to Great-Britain. It is on a scale of more than half an inch to the English mile, and exhibits all the allotments of the plantations. The island is pervaded by three chains of mountains; one along the northern shore, one along the southern, and another in the middle. They all run from east to west.

The description begins with informing us that the extreme length of Trinidad from east to west is sixty British miles, and its breadth from north to south forty-five miles, containing two thousand four hundred square miles, superficial measure. But, on measuring the map by the scale there laid down, we find the extreme length from east to west to be not less than eighty-five British miles, and from north to south more than sixty. This mistake is wholly unaccountable, and does very little credit to the accuracy of the compiler of the description. In fact, it is idle to compute the length by the long southern promontory. The real length of Trinidad is from north to south about sixty British miles, according to the present map, while the breadth from west to east is about forty; or, including the slender promontories, and taking a medium, about fifty British miles.

This map is said to have been grounded on the Spanish maritime survey made by order of government in 1793; but the topography of the interior was executed by Mr. Mallet in 1797, when the British forces took possession of the island. It is added that it not only contains all the plats of land granted by Spain, but those which remain to be disposed of, divided into squares. It seems not a little remarkable that the proprietors, as appears from the list of names, are mostly French. The monstrous fish mentioned in p. 5 is the hammer-headed shark. The acquisition of this noble island may no doubt lead to great advantages in our West-India trade.

A Map of the United States of North-America, drawn from a Number of critical Researches. By Aaron Arrowsmith, Geographer. Four Sheets. 11. 15. Sold by the Author.

THIS interesting map, having been published in 1796, can only be here noticed for the improvements made in the edition before us, and which extend to 1802. These are chiefly in the

north-west division, in which the Lake Superior assumes a considerable difference of form, and the sources of the Mississippi vary considerably from their former appearance. North-America having particularly engaged our geographer's attention for several years, there can be little doubt that the present map, which is engraven with great clearness, will be found superior to any other which has appeared of the United States.

A Map of Turkey in Europe, drawn from a great Number of accurate Astronomical and Geographical Manuscripts and printed Documents. By Aaron Arrowsmith. Two Sheets. 10s. 6d. Sold by the Author. 1801.

IT is well known that the geography of European Turkey is singularly imperfect, as the ignorance and fanaticism of the Turks, not to mention their jealousy of their European neighbours, have prevented any accurate survey. Of the provinces on the north of the Danube delineations have been taken with some exactness by officers in the Russian and Austrian service; and of many parts of Greece actual surveys were made by the count de Choiseul: but for the other extensive provinces, particularly Albania and those in the west, the materials are extremely defective. From the routes of various travelers, and other authentic sources, Mr. Arrowsmith has however been enabled to draw up a map, which, though it cannot be regarded as in any degree complete, is nevertheless superior in many respects to any general map of European Turkey. The mountains are, with few exceptions, laid in chorographically; and the engraving is very clear and neat. But we must regret that the sea is marked by a mere outline-a circumstance which gives an unfinished appearance. The remedy in such a case would be to tinge the sea with verdigris; but as maps, like books, are intended for distant posterity, it is probable that no tinge would stand a course of years; and it seems desirable that a map should be perfect in itself, without requiring any adventitious aid, which may often be administered by unskilful hands.

The rivers in Moldavia and Walachia are laid down with great minuteness; but the other parts of these provinces do not seem to be equally finished, as there is not the shadow of a hill, or any elevated ground; and even the Carpathian mountains are delineated in a faint and obscure manner. But the great chain of Hamus, and the branch called Rhodope, are well expressed, and throw light on some doubtful parts of ancient geography. In the north-west there is an omission of a considerable part of Bosnia, together with the whole of Croatia, and a fragment of Herzgovina or the duchy of St. Saba. Yet it is

probable that in a very short period of time the present map may be more complete than at present.

As this map is chiefly interesting with regard to ancient history, we wish that the ingenious author had accompanied it with a short memoir, pointing out the sources of his infor mation. If a partition of Turkey be in agitation, as is averredand is indeed to be wished for the sake of humanity, lest those beautiful regions should become the prey of ignorant and cruel bashaws, whose petty wars would spread universal destruction and depopulation-the marches of foreign armies will soon illustrate the geography of this country. If not, it would be desirable that enlightened travelers should visit Albania, and other parts little known, instead of presenting us with a thousand voyages to Constantinople and Smyrna, where nothing remains to be discovered.

A New Map of Africa, including Arabia, the Mediterranean, and Part of the Coast of South-America; compiled from the Observations of the latest Travelers. Four Sheets. 10s. 6d. Wilkinson. IN this map, that of d'Anville has been used with some care; and the latest discoveries are inserted, so as to constitute, upon the whole, a performance not indeed of the first merit, but tolerably decent and exact. The routes introduced are those of Bruce, Browne, and Park, with that of Messrs. Watt and Winterbottom to Teemboo, Vaillant from the Cape of Good Hope, and Dawes from Sierra Leone.

Even in large maps, we have sometimes been surprised at the want of taste and skill in the general square. This defect we had recently occasion to censure in a map of America; and, in the present instance, the absurd admission of a considerable part of Brazil greatly injures the appearance and effect of the wholethe continent of Africa being thrown to one side, in opposition to the rules of just design. Owing to this strange idea, several isles on the N. E. of Madagascar, &c. are omitted, though they strictly belong to the African division of the world. On the other hand, Arabia is laid down for the first time with some care from the description of Niebuhr, though far indeed from perfect accuracy-the constructor having neither talents nor experience sufficient to attain geographical eminence. On a general view of this map, it must also be observed that the mountains are delineated in the old perspective manner; an appearance foreign to nature when they are viewed from above, as must be the supposition in geographical delineation. The sandy deserts, however, are well represented: but, in staining a map, it seems unnecessary to mark the outlines with a deeper colour; nay, it is frequently injurious, as it hides the course of the rivers, the