the fourth volume of Stuart's Athens, Atlantes are repre- | Antarctic Ocean, is included in the Atlantic, but these consented standing upon a plinth placed on the entablature tiguous parts cannot well be separated in a description of above the pilasters of the cella of the temple, and sup- the Atlantic. porting with their heads and arms the entablature on which the beams of the roof were to have been placed. The Atlantes of this temple were twenty-five feet high, built in courses of stone, corresponding with the walls of the cella, and partly attached to it. The annexed cut, showing [From Temple of Jupiter at Agrigentum.] Though the Atlantic Ocean extends from pole to pole, its breadth is comparatively not great. The two continents which form its shores approach nearest one another between 69° and 71° N. lat., where the coasts of Greenland are only 800 geographical miles from those of Norway, a distance hardly greater than that between Nantes in France and Cape Wrath in Scotland. Between Cape St. Roque in Brazil, about 5° S. lat., and the coast of Sierra Leone in Africa, between 5° and 8° N. lat., the continents are not above 1500 geographical miles from one another, or about as far as the North Cape from the Nore. These are the two parts where the width of the Atlantic Ocean is least. Its greatest breadth is under 30° N. lat., where the peninsula of Florida and the western coast of Marocco in Africa are separated by upwards of 3600 geographical miles, or 60° of latitude. Humboldt compares the form of the Atlantic Ocean to that of a longitudinal valley, whose projecting and retiring angles correspond to one another. He supposes it to be formed by a very violent rush of the waters from the south, which being repulsed by the mountains along the coast of Brazil, took a direction towards the coast of Africa, and the front elevation of the figures, with a profile of one of formed the Gulf of Guinea; here being stopped by the high hem, has been engraved with the permission of the pub-coast of Upper Guinea, and obliged to run again to the west, ishers of Stuart's Athens. (For a more detailed account of the stream gave origin to the Caribbean Sea and the Gulf these figures, see vol. iv. cap. i. of Stuart's Athens, pub- of Mexico, and issuing thence ran between the mountains Ished by Weale, Holborn.) of western Europe and those of North America, until it gradually diminished in velocity and force, and at length subsided. In confirmation of this hypothesis, he observes, that the primitive mountains in the Brazilian provinces of Rio, Minas Geraes, Bahia, and Pernambucco are under the same degree of latitude as those of Congo, and that the immense plain along the banks of the Amazon river corresponds to that traversed by the Quorra: further, that the mountains of Parime in America lie opposite to those of Upper Guinea, and that the great plains, which before this catastrophe, according to his hypothesis, extended to the south of the present mouth of the Mississippi, and by the submersion of which the Caribbean Sea and the Gulf of Mexico were formed, are under the same parallel as the great desert of the Sahara. In the Tepidarium of the baths at Pompeii, Atlantes of baked clay, in high relief, and incrusted with the finest marble stucco, painted to represent life, are ranged at equal distances round the room, to support an entablature from which the arched ceiling springs; in the intervals between the figures, niches are formed for the dress of the bathers. The figures are about two feet high, and stand, like those at Agrigentum, on a plinth. In the annexed cut, from [From Pompeii.] the Society's work on Pompeii, a representation of these figures is given. The South Atlantic Ocean does not offer any other pecu liarity in its formation, but the Northern is distinguished by several. First, we may observe the formation of its islands lying within the polar circle. They are countries of considerable extent, but divided by extremely narrow, long, and winding straits, of very difficult navigation, which is increased by their being only for a few weeks in the year free from ice. For instance, the group of Nova Zembla consists of at least two larger ones; that of Spitzbergen of three larger and many smaller ones; and it is rather more than a supposition, that the extensive country known by the name of Greenland is composed of a number of large islands, divided from one another by narrow, long, and winding straits. This peculiarity in the formation is repeated, though on a less scale, in the islands which skirt the coasts of Norway, where, in some instances, the straits which once divided them from the continent have been filled up by earthy matter, and now resemble exactly Glen More in Scotland. In no other part of the seas has such a dis In the architecture of the modern Italians, the Atlantes are often found supporting the entablature over an entrance to a palace or a garden. At Milan, there is a colossal exam-position of islands been observed, except in those on the ple of the former; and the rustic gate to the Farnese Gar dens at Rome, the design of Vignola, may be adduced as an example of the latter. ATLANTIC OCEAN is the name given to that part of the ocean which separates the old from the new world: it washes the eastern shorcs of the Americas, and the western shores of Europe and Africa. Nature not having fixed any boundary between it and those seas which are adjacent to and communicate with it, we shall suppose that it is divided from the Pacific Ocean by a straight line drawn from Cape Horn, the southern extremity of America, to the antarctic pole, and from the Indian Sea by another drawn from the Cape of Good Hope to the same pole. On the north we may say that its boundaries are fixed by nature, in that continuous and impervious barrier of ice which extends between 80° and 82° N. lat. from the coast of Greenland to the island of Nova Zembla. By fixing these boundaries, a | part of the Northern Polar or Icy Sea, as well as of the No. 139. S.W. coast of America, and probably those on the N.W., about the latitude of Admiralty Island. But a still more remarkable and more important feature of the North Atlantic, is its connexion with mediterranean, or close, seas of great extent. Such are the Baltic Sea and the Mediterranean Sea in the Old Continent, and Hudson's Bay and the Gulf of Mexico, with the Caribbean Sea, in the New World. These seas doubtless form part of the Atlantic Ocean; but they cannot be considered as bays or gulfs, the connexion between them and the Atlantic being effected by narrow straits, and not by an open sea; and, besides, they extend so far into the continents, that some of them, as the Mediterranean Sea, affords by itself a navigation of 3000 geographical miles. Similar seas indeed exist in the Indian Ocean, where the Gulfs of Persia and of Arabia resemble rather the Mediterranean and the Baltic Seas, than the Gulf of Bengal or that of Guinea; but they are of much less extent. This peculiarity in its form brings VOL. III.-E [THE PENNY CYCLOPÆDIA.] the Atlantic Ocean and its appendages into immediate contact with a much greater extent of country than the other seas that wash both continents. We accordingly find that the continental shores of the Atlantic exceed in extent those of the Pacific Ocean and the Indian Sea, the two other great divisions of the Ocean, taken together, though the latter cover at least three times the surface of the former. The continental coasts of Europe from the strait of Waigatz to that of Caffa (the entrance of the sea of Azoff), are about 17,000 geographical miles; those of Asia along the Black Sea, the Sea of Marmora, and the Mediterranean Sea, are nearly 3000 miles; and the coasts of Africa, along the Mediterranean Sea, are upwards of 2000 geographical miles. Add to these the western shores of Africa from the strait of Gibraltar to the Cape of Good Hope, which comprehend about 6000 geographical miles, and the whole eastern shores of the Atlantic Ocean amount to 28,000 geographical miles. In computing its western shores, we shall consider Greenland as a part of the continent, though it probably is not strictly true; and on this supposition we find that the eastern shores of America comprehend about 20,000 geographical miles. Consequently the shores of the Atlantic Ocean have a circuit of about 48,000 geographical miles. The coasts of Asia are upwards of 30,000 geographical miles; but nearly 3000 of them belong to the Mediterranean Sea, and consequently to the Atlantic Ocean. The eastern coast of Africa may be computed at 6000 geographical miles, and the western coast of America at upwards of 11,000. Thus the coasts of the Pacific Ocean and those of the Indian Sea taken together do not amount to much more than 44,000 geographical miles, or nearly 4000 miles less than those of the Atlantic Ocean. We shall observe, that in this calculation the northern shores of Asia along the Polar Sea are included, and as they amount to upwards of 2600 geographical miles, the account is still more in favour of the Atlantic Ocean, if this length is subtracted. We shall not enlarge on the advantages which such a peculiar form of the Atlantic must offer for the progress of civilization. These advantages would extend to a great distance into the interior of both continents, if the number and magnitude of the rivers which flow into the Atlantic were proportionate to the extent of its shores. On the eastern side, the surface, whose drainage falls into the Atlantic, is comparatively limited, and does not comprehend even the whole of Europe: the greatest river of this part of the world, the Volga, carries its waters to the Caspian Sea. No European river of the first or second class flows immediately into the Atlantic Ocean; the largest being probably the Rhine, whose course does not exceed 700 English miles. But three rivers of the second class, the Nile, the Danube, and the Dnieper, enter the Mediterranean Sea or its branches. The boundary line, which marks the region from which the waters run into the Atlantic Ocean on the east, is extremely irregular. On the north it begins with the most northern extremity of the Uralian Mountains, and follows that range to near the sixty-first parallel, where, at the sources of the Kama, it suddenly turns to the south-west and then to the west, in which direction it continues to the sources of the Volga, hardly 150 miles distant from the Gulf of Finland. From this point it runs nearly south to the 55° of lat., from which it extends east-south-east between the tributaries of the Volga on one side, and those of the Dnieper and Don on the other. Having thus attained the 45° of E. long., and nearly the 52° of N. lat., it takes a due southern direction between the rivers Don and Volga, and nearly traversing the middle of the Caucasus, it declines to the south-west, and separates the upper course of the Euphrates from the small rivers which fall into the Black Sea and the Gulf of Scanderoon. It then runs along the coast of Syria at an average distance of less than a hundred miles, and turns round to the Isthmus of Suez. In Africa it encloses the valley of the Nile, the upper part of which is of unknown extent. To the east of this river, the boundary of the Atlantic runs along the shores of the Red Sea, a branch of the Indian Ocean, and at the sources of the Nile it is at least 1600 miles distant from the Mediterranean Sea, and consequently from the Atlantic, the greatest distance which it probably attains in the old world. From near the mouths of the Nile, it runs due west, following generally the thirtieth parallel till it arrives at the shores of the Atlantic, opposite the Canary Islands. To the south of the thirtieth parallel, the boundary of the drainage of the Atlantic Ocean falls in with its shores; the great African desert not being included in it. What parts of Africa south of the Sahara belong to the basin of the Atlantic Ocean, our present geographical knowledge does not enable us to decide with accuracy. Perhaps we shall not much overrate it, in supposing that the drainage of half of its surface flows to the Atlantic. We therefore may suppose that the basin of the Atlantic contains about three millions of square miles in Europe, not half a million in Asia, and about six millions in Africa; which all taken together do not amount to more than nine millions and a half, or about one-fourth of the continent of the antient world; but the new continent belongs almost entirely to its basin. In South America, the water-shed between the Pacific and Atlantic Oceans runs at a distance of from 25 to 200 miles from the shores of the former, except in the very southern extremity of the Andes [see ANDES]; and the extensive plains which cover the greatest part of the surface of that continent send their waters to the Atlantic Ocean. Probably not less than six millions of square miles of the surface of South America belong to the basin of the Atlantic, and only half a million to that of the Pacific Ocean. In North America, the line which separates the waters falling into both oceans lies at a much greater distance from the shores of the Pacific Ocean; but even here the great plains to the east of the Stony Mountains send their rivers to the Atlantic: so that, if we assign to the Pacific Ocean even the northern region traversed by the Mackenzie River, the area drained by the rivers falling into the Atlantic may amount to upwards of six millions of square miles, whilst those falling into the Pacific probably do not drain more than two millions. According to this account, the basin of the Atlantic Ocean comprehends about nineteen millions of square miles on both continents; and the remainder, amounting to about twenty-seven or twenty-eight millions, belongs to the basins of the Pacific and Indian Seas, and to those of a few inland lakes, or to a few deserts which have no water. The Atlantic Ocean being, in the present state of the commercial world, the most frequented high-road of communication, has been examined more completely than the other seas, with respect to its facilities for navigation. The dangers and difficulties produced by numerous and intricate groups of islands are of less frequent occurrence in this sea than in any other for, if we except the chain of islands which separates the Gulf of Mexico and the Caribbean Sea from the Atlantic, and which therefore are to be considered as forming part of the shores of the ocean, it can hardly be said to contain any group of islands between 50° N. lat. and 50° S. lat. The groups of the Azores, Canaries, and Cape de Verde Islands, as well as those of Guinea and the Bermudas, are small, and present few difficulties to navigators. The Canaries, including Madeira, are much resorted to by vessels, from their situation on the verge of the regions in which the elements essential to navigation (viz., the air and the water) undergo a change: for to the south of that group, the winds, as well as the motion of the sea or the currents, are generally much less changeable than in the latitudes nearer to the poles. With respect to the winds, the whole surface of the Atlantic Ocean may be divided into three regions, in one of which the winds maintain a constant course from east to west, and have obtained the name of trade-winds. This region extends to about 30° of lat. on both sides of the equator. The other two regions, to the north and south of the thirtieth parallel in both hemispheres, are subject to a continual change of the winds, and are therefore called the regions of variable winds. It is not here our object to enter into an explanation of the natural causes which produce the phenomenon of the perpetual or trade winds [see TRADE WINDS]; but we shall historically observe the deviations from these general rules, which are found to exist in the Atlantic Sea, and which themselves affect the navigation of it no less than the trade-winds. In the eastern part of the region of the trade-winds, these winds blow, on the north side of the equator, from north east; and on the south, from south or south-west, as we shall see hereafter. If they continued in these directions, they would of course meet one another, but this is not the case both trade-winds are separated from one another by the region of calms. This region is not always of the same extent, and does not occupy the same part of the ocean, though it always extends over the whole of it from the coasts of Africa to those of America. It sometimes occupies not much more than two, and at other seasons up to ten degrees of latitude. It is a remarkable circumstance that it does not extend equally on both sides of the equator, but is rather situated in the northern hemisphere. It rarely extends to the south of the line, and never farther than two and a half degrees of south latitude, whilst, on the north of the equator, it sometimes advances even to the thirteenth degree of latitude. The position of the sun visibly influences the position of the region of calms, as well as the range within which the trade-winds blow. In the summer months, especially in July, August, and September, the calms advance towards the north, and extend sometimes, as we have already observed, to the thirteenth parallel; at this time the southern trade-wind encroaches considerably on the northern hemisphere, being found as far as the fourth or fifth degree of northern latitude. It may even be said that this last-mentioned trade-wind passes the equator all the year round, except in January, when the region of calms extends to two and a half degrees of south latitude. The central line of this region may therefore be placed at about re or five and a half degrees of north latitude, and its mean breadth may extend over five or five and a half degrees of latitude, or from 300 to 330 geographical miles. The calms which reign in this region would oppose an insuperable obstacle to the progress of vessels, if the water was not daily agitated by a squall. At noon, a black and well-defined cloud appears towards the east, which seems to announce a violent thunder-storm; suddenly a wind rises, blows with great violence for a few minutes, sends down a few drops of rain, and immediately the calm returns. It is only by means of these daily squalls that the region of calms can be passed by vessels, but it always proves a very disagreeable navigation. The trade-winds themselves are subject to change in their extent, according to the seasons, and in their direction, according to the different degrees of longitude. They withdraw farther from the equator when the sun is in the hemisphere in which they blow, and they occupy a wider range towards the coasts of America, than at a short distance from the old world. In the seas bordering on the latter they are rarely encountered at thirty degrees of latitude, and often not before reaching the twenty-seventh parallel, so that the mean boundary may be placed at twenty-eight and a half degrees. Towards America, however, they extend somewhat more towards the north, even to thirty-two degrees of latitude, so that the mean boundary may be fixed at the thirtieth parallel. The direction of the trade-winds changes with their progress from east to west. Near the old continent, and north of the equator, the direction is from north-east, or nearly so, but farther off it declines more to the east, so that in the middle of the ocean it is east a quarter north, and on the coasts of America it blows from due east. No navigation is more pleasant than that with the trade-winds. They are rather a breeze than a wind, and their blowing is uniform, constant, and not interrupted by squalls. The waves raised by them are low, and their swell is gentle. All these circumstances induced Humboldt to think that it is less dangerous to pass from the Canaries to America, than to traverse one of the small lakes of Switzerland; and he compares this navigation with the gentle descent on a slow-flowing river, or rather a canal. The Spaniards have called this part of the Atlantic Ocean il golfo de las damas-supposing that even ladies could muster courage enough to navigate it; for here the passage to America may safely be effected in an open boat. To these general observations we shall add a few others, with respect to the difference between the northern and the southern trade-wind. The northern trade-wind declines only from north-east to east in its progress westward, but it is less constant towards its northern boundary than the southern towards the antarctic pole. Between the twentieth and thirtieth degrees of north latitude, sometimes violent north-western winds prevail. For that reason, vessels bound for the West Indies or South America sail along the old continent till they attain the twenty-first parallel, when they turn to the west. The southern trade-wind is more regular, and always preserves its direction, and it is also less boisterous towards its southern boundary. It extends, as we have already observed, so far north, that it is sometimes found in force as far as the fifth or even sixth degree of north latitude. Along the coasts of Southern Africa, it blows from the south-west; but at a distance from the coast it becomes by degrees more southerly, and as it proceeds farther to the west its direction becomes nearly due east. The meridian of twenty degrees west of Greenwich may be considered as a line of separation between the winds which blow from the south and those from the east. To the east of that line, the direction of the winds varies between S.W. and S.S.E., and to the west of that line between S.S.E. and E.S.E. The trade-winds do not begin to blow on the coast of the continents, but only at a considerable distance from them. This is to be attributed to the different degree of temperature of the land and of the sea, which, when the land is extensive, must be considerable, and produce a great change in the wind. Thus, between the northern trade-wind and the African coast, from the Canaries to the Cape de Verde Islands, or, more exactly, between Cape Bojador and the mouth of the Senegal, the wind blows constantly from the west. This phenomenon is to be accounted for by the nature of the great desert called the Sahara, whose surface, composed of loose sand, is heated by the sun to an excessive degree, and, rarefying the superincumbent air, causes it to rise. When this rarefied volume of air comes in contact with the more dense strata of air covering the sea, the latter expands over the deserts, and in this manner the west wind along this coast is produced. But this influence of the Sahara does not extend beyond the Cape Verde Islands. Farther to the south, as far as Cape Mesurado, or more properly between 15° N. lat. and the equator, that is, to the east of the region of calms, a kind of monsoon prevails, blowing in certain places to the distance of seventy leagues off shore (fifty leagues off Sierra Leone), and proceeding from the north-east or north from September to June, but in the remainder of the year from south-west. It is remarkable that along this coast the currents are likewise periodical, and change together with the winds: but they always run in an opposite direction to them. In the Gulf of Guinea or Benin, and along the coast of Southern Africa to 30° S. lat., the wind does not materially differ from the southern trade-wind, blowing constantly from the south-west, or nearly so. The changes which are produced in the direction of the southern trade-wind by the continent of South America are considerable. Along the coast of Brazil a regular monsoon is formed, which between September and March proceeds from between N. E. to N.E. E.; and from March to September from between E. & N. to E.S.E. This monsoon blows commonly with considerable force, and extends to a great distance from the shore, especially in the months of June and July, when it is at its height, at which time it sometimes advances nearly as far as the middle of the Atlantic Ocean. To the north of the equator the tradewind does not undergo any change along the coast of America. This is probably to be attributed to the lowness of the coast of Guiana, and that of the extensive and frequently inundated plains along the lower course of the Orinoco. It is even observed that in this quarter the trade-winds extend so far over the land, that their effect is felt at Angostura, 240 miles from the mouth of that river. [See ANGOSTURA.] These are the winds which blow on both sides of the equator to the thirtieth degree of lat. in the region of the perpetual winds. To the north and south of this region the winds are variable; but it is observed that westerly winds prevail in both hemispheres; and, according to the computation of Major Rennell, the proportion between those that blow from the west to those from the east, is as 9 to 5 in the northern hemisphere. Besides being variable in direction, they vary likewise extremely in the degree of force with which they blow. The currents of the Atlantic are less important than the winds; but still they contribute considerably to accelerate or retard navigation, and on that account deserve the greatest attention. But, as hardly sixty years have passed since they began to attract the attention of the navigator as well as the geographer, and as the subject is involved, from its nature, in many difficulties, the information respecting them is not such as could be wished; still much has been collected, which is both interesting and useful. It would seem that there is hardly any considerable portion of the ocean which always remains still. The tides do not occasion an absolute removal of the water from one place to another, except near the coast; and even there this motion is limited to a comparatively small distance. But The drift-current is, in some measure, observable all over that portion of the Atlantic Ocean which is under the influence of the trade-winds; but as these winds are not very constant to the north of the 23d parallel, and rarely extend to the south of the 9th, the current is constant only between these two boundaries. In the region of the calms it is very weak, and often entirely ceases. But in those regions in which the southern trade-winds blow, it is again perceptible and constant, except along the coasts of Africa, where it has rather a northerly than a westerly motion; the latter, however, becomes by degrees more prevalent in proportion as the wind takes that direction in advancing to the west. The mean velocity of this current is from 9 to 10 miles per day, or, according to the computation of Humboldt, only one-fourth of the velocity with which those rivers in Europe commonly flow on which observations have been made. The drift-current, which in the northern portion of the Atlantic is produced by the prevalent westerly winds, flows in a westerly direction; but it is not perpetual, and is so slow, that, when a ship keeps clear of the Gulf Stream, it only manifests itself generally on the whole course of a voyage from Europe to America and vice versa, retarding the former and forwarding the latter. besides the tides, two kinds of motion are to be distinguished | both sides of the equator, as far as 22° W. long., where it in the sea, which we shall name with Major Rennell the sends off a branch to the north-west. Soon afterwards it drift-currents and the stream-currents. declines somewhat to the south, and runs in this direction The drift-currents owe their origin to the effects produced towards the two capes of St. Augustin and St. Roque, on on the surface of the sea by the perpetual or prevailing the Brazilian coast. At the distance of about 300 sea-miles winds; the former, even where they do not blow with great from these capes, it divides into two currents; the northern, force, by their uninterrupted continuance displace and push running along the shores of Guiana, and hence deriving forward the upper strata of the water, and thus produce a the name of Guiana Current, enters the Caribbean Sea motion towards the region to which they blow. These by the straits which separate the Leeward Islands, lying drift-currents are constant, and run always in the same to the south of Martinique, from each other and from direction and commonly with pretty equal velocity. The the continent of South America; and in some measure drift-currents produced by the prevalent winds are not so in this sea it may be supposed to terminate its course. constant and do not always run in the same direction nor The Brazil Current, or the other branch of the equatorial with the same velocity. In the Atlantic Ocean, the former current, runs to the south-west along the shores of Brazil, kind of drift-current is found only between the tropics, to the mouth of the Plata River, and may even be traced where it is produced by the trade-wind; and the latter to to the Straits of Magalhaens and of Le Maire. The the north and south of 30°, where they are ascribed to the whole length of this current, from St. Thomas to Cape effects of the prevalent winds. St. Roque, amounts to upwards of 2500 nautical miles; and if we add the Guiana current, from the point of division opposite that cape to the strait dividing the island of Trinidad from that of Grenada, its course is increased by 1500 nautical miles more. The breadth of the current is different in different parts. Near the islands of St. Thomas and Anno Bom, it extends not quite over three degrees of latitude, occupying about 160 miles. But, in proportion as it advances to the west, it increases in breadth; opposite Cape Palmas it extends from 1° 45' N. lat. to near 5 S. lat., thus occupying in breadth more than six degrees, or upwards of 360 nautical miles. Farther to the west it enlarges still more, and attains its greatest breadth, extending over 7° or 8° of lat. from 44 or 5° south of the equator to 24° or 3° north of it. Here, therefore, the breadth of the current occupies 450 geographical miles, or not much less than the whole length of Great Britain, from the Lizard to Cape Wrath. But having soon afterwards, between 22° and 23° W. long., sent off a branch to the north-west, it narrows to about 300 nautical miles; and this breadth it probably preserves to the point where it divides opposite the capes of St. Augustin and St. Roque. The velocity of the current is different in different parts, and increases or decreases according to the seasons, it being much greater in summer than in winter. From Anno Bom to 10° W. long. it runs, at an average, from 25 to 30 miles per day; but from 10° to 16° W. long. it is much more rapid—making, in the same time, from 44 to 79 miles at the end of June and the beginning of July. This seems to be the strongest part of the whole stream. But it is only in the months of May, June, July, and August, that it runs with great force; from October to March it is moderate, and sometimes very weak. Between 16° and 23° W. long. lies the common track of the vessels; and here the rapidity of the current rises often to 45, 50, and even 60 miles per day, but its mean velocity may be estimated at about 28 nautical miles; it is strongest near the equator, and stronger to the north than to the south of it. From 23° to the coasts of Brazil, the current becomes rather stronger, and seems to be less affected by the seasons; but its velocity in these parts is not exactly ascertained; it seems, however, to run 30 miles and upwards per day. The temto the seasons and the different parts of its course, but it is always some degrees lower than that of the ocean. The water of the ocean to the north of the current is 80° or 81° Fahrenheit, and to the south, 78° or 79° in summer; but in the current, the thermometer shows, near Anno Bom and St. Thomas, only 75°, and not more to a great distance westwards, where the temperature falls even to 73°, and at this temperature it remains for more than 12° of longitude. Afterwards it rises again to 74°, and by degrees to 76° Fahr. In summer the temperature of the current may be estimated as being, at an average, 5° or 6° under that of the water of the ocean; but in winter it is much less. This current greatly affects the course of vessels which are obliged to cross it, and creates great delays to those who, passing from the north to the south, traverse the equator west of the 23° of long., carrying them forcibly to the west beyond Cape St. Roque, where they are driven towards the northern shores of Brazil, and are not able to regain their course till after weeks, and even months, of toilsome labour. It is a fortunate circumstance that the direction of this current does not coincide with the region of the calms; otherwise, both together would probably form an impenetrable barrier to the progress of vessels navigating these seas. But the southern trade-wind commonly blows in that region It is easy to conceive that the drift-currents, especially the permanent, are very favourable to navigation, by rendering the voyages to some countries more easy, more certain, and less dangerous. But the stream-currents are much less so. Up to the present time they have commonly proved adverse, causing great loss of life and property, and forcing vessels out of their course. Many navigators, running from Madeira to Teneriffe, and expecting to arrive at the latter island, have unexpectedly found their vessels cast upon the shores of Africa, nearly 300 sea miles out of their course. Such errors can only be detected by frequent astronomical observations, and by comparing them with the dead reckoning. If they are not detected in time, shipwrecks sometimes become unavoidable. We cannot compare the stream-currents of the ocean with the rivers of the continents. The stream-currents cover such a portion of the surface of the sea, that were they trans-perature of the water in the current varies also, according ferred to the continents, they would no longer be considered as rivers, but as large branches of the sea. The causes to which they owe their origin are still involved in obscurity; our observations have not yet penetrated into the depth of the sea, they have only slightly investigated its surface, and there are some facts which lead to the opinion that the stream-currents are of great depth, and in many parts, if not in all, extend to the bottom of the sea. This indicates clearly that their origin must not be ascribed to changes which take place on its surface, and cannot affect the lower strata of its waters. The opinions which have been formed on this object may be seen under the article CURRENTS. We shall here only notice the largest of the current streams which belong to the Atlantic Ocean, and indicate their extent, velocity, and temperature, their only properties which, up to this time, have been in some degree ascertained. Two large stream-currents traverse the Atlantic Ocean; the Equatorial Current, running from the coast of Africa to that of South America, and the Gulf Stream, flowing from North America to the shores of Europe. The Equatorial Current, so called from its course lying under or near the Line, may be supposed to be formed between the islands of St. Thomas and Anno Bom, in the bight or bay of Benin. Hence it proceeds to the west on | where the equatorial current runs, at least during those | known of the Atlantic currents, is the Gulf Stream, which months in which it runs with great force. traverses the sea between the parallels of 36° and 44° in the That branch of the equatorial current which separates northern hemisphere. Its origin is in the Gulf of Mexico, from it between 22° and 23° W. long., and at about 24 N. where the warm water which enters it from the Caribbean lat., is rather favourable to navigation, by forwarding the Sea, between Cape Catoche and Cape St. Antonio, by being course of vessels returning from the southern hemisphere. subjected to a nearly circular rotation, and influenced proIts course lying in a north-western direction, it may be bably by other causes still unknown, is raised to a high decalled the North-west Current. Its breadth is consider-gree of temperature, the thermometer indicating 86°, while able, and may be estimated, at the point of separation, at under the same parallel (25° N. lat.) the ocean only shows 78°. 180 or 200 miles; farther northward, even at 300; and at a Two currents, which put in motion perhaps three-fourths of more considerable distance, at 240 nautical miles; but its the waters of the Gulf of Mexico, unite about 60 nautical velocity is not so great as that of the main equatorial current. miles to the westward of Havannah, between the bank of Up to 10° N. lat., however, and even a little farther, it may Isabella on the side of Cuba, and the Tortugas on that of run from 20 to 24 miles per day; but it then slackens, and Florida Reefs; and this union gives rise to the Gulf Stream becomes less perceptible, though it may at all times be In the beginning its course is not rapid, and along the traced to 18° Ñ. lat., and commonly even to 25°. In the shores of Cuba it is weak, and sometimes nearly imperatter part it bends more to the northward, and at last is ceptible: but it soon increases in velocity, and before it lost in the drift-current, to which it seems to give a north- enters the Strait of Florida at the Salt Kays, it runs one westerly direction, which is observable all the way from mile and a half per hour on an average; in the strait itself 35° W. long. to Trinidad. The accelerated motion of the it acquires a formidable swiftness. The Strait of Florida drift-current in these seas may also, in some measure, be begins at the Salt Kays, a reef lying 114 nautical miles to the effect of this north-west current. the north-east of Havannah, and extends thence to the northward, where it terminates between the reefs of Cape Cañaveral and the northern termination of the Lesser Bahama Bank, at about 28° N. lat. After entering this strait, the velocity of the Gulf Stream soon increases to 23, 3, and occasionally 4 miles per hour; but after running at this rate about 90 miles, it arrives at the Narrows, where, between Cape Florida and the Bimini Islands (a small group belonging to the larger group of the Bahamas), the strait is only 44 miles wide, and its water-way, by reefs and shoals, is straitened to 35 miles. Here the current runs, in the month of August, at 5 miles per hour, and seldom below 5 through the remainder of the strait, which towards its northern extremity widens to about 50 miles. In this course the current has traversed about four degrees of lat. northward, but its temperature is not sensibly diminished. The Guiana Current is, properly speaking, the continuation of the equatorial current, and runs from Cape St. Roque in Brazil, to the Island of Trinidad, along the low coast of Guiana, and at no great distance from it. It is of considerable breadth, and of great velocity, running at some places two miles per hour. Here too it is observed that its velocity is much greater in summer than in winter and spring; and it is besides much increased by the waters rushing from the mouth of the Amazon river into the sea; for after this has taken place, the current runs three miles per hour. It is, however, to be observed, that the waters of the Amazon river do not mingle with those of the current: they cut them at right angles, and running with great violence, cause eddies and whirlpools; but at last the river passes the current, and is observable at a distance of 500 miles from its mouth. It is remarkable that the Amazon causes no change in the direction of the current. Farther to the north, the Orinoco enters the current. This river, which pours a prodigious mass of water into the ocean on both sides of the island of Trinidad, enters the current at a very acute angle; and thus soon mixes its water with it, and considerably accelerates its course. Soon afterwards the current enters the Caribbean Sea by the straits lying between Trinidad and the island of Martinique. Between Trinidad and Grenada, it runs from 1 to 13 mile per hour; less between St. Vincent and St. Lucia; and between the latter island and Martinique its course is reduced to 21 miles per day. Farther to the north, and especially at the Virgin Island, the sea-water runs only from 8 to 10 miles per day, and that is the common rate of the drift-current. We find no observations whether the water of this current differs in temperature from that of the ocean. It may be said that the Guiana current is lost in the Caribbean Sea, for in the latter no perpetual current has been observed. The whole course of the Guiana current may be reckoned at about 2500 nautical miles. The third branch of the equatorial current is the Brazil Current, which branches off from the equatorial at 8° S. lat., opposite Cape St. Augustin, at a distance of about 300 miles to the east of it. Between the point where it branches off and 16 or 17° of S. lat. it has considerable breadth, and does not approach the shores of South America nearer than 250 miles. Afterwards it increases in breadth and velocity, and approaches nearer the land. Opposite Cape Trio it is 200 miles from the coast, and runs about 30 miles per day. As the shore south of this cape falls off to the west, the current is at a greater distance; and though it soon changes its direction, it does not approach nearer than 250 miles to the coast, off the mouth of the Plata river, running all this way from 15 to 20 miles per day. It is crossed by the Plata river, just as the Guiana current by the Amazon river. Here too the current of the river is observable in the sea at a distance of upwards of 500 miles, but it seems not to have the least effect in changing the direction of the streamcurrent, which continues, though much weakened, farther to the south, and may be traced to the Straits of Magalhaens and Le Maire. As this current runs at a considerable distance from the shores of Brazil, the intermediate space is Occupied by other currents, which mostly, however, follow the direction of the monsoons which blow along this coast. The most remarkable, and at the same time the best Issuing from the Strait of Florida, the Gulf Stream runs northward along the shores of Florida to 31° N. lat., and afterwards nearly north-east along the shores of Georgia and of both Carolinas, as far as Cape Hatteras (about 35° N. lat.). In this course the current widens considerably in breadth, and decreases in velocity and temperature. Opposite the harbour of Charlestown, its breadth is from sixty to sixty-three miles; and at Cape Hatteras, from seventy-two to seventy-five miles. At the latter place it runs only three and a quarter miles per hour, or seventyeight miles per day, and its temperature has fallen from 86° to 83°. At Cape Hatteras, the north-western or inner edge of the current is twenty-four miles south-east of the cape. By the falling back of the coast north of Cape Hatteras, the current directs its western edge towards the north, while the main body continues its former course to the north-east to a considerable distance. At about 40° N. lat. it meets the extensive Nantucket and St. George's Banks, which turn it off seaward, and it never after approaches the land. From this point it runs in the direction E. N., brushing the southern extremity of the Great Bank of Newfoundland, and it continues in this line to 43° and 44° long., and 37° and 43° lat. Here, however, it bends by degrees to the east, south-east, and south, and having enclosed the islands of Flores and Corvo, which belong to the group of the Azores, its traces are lost in the water of the ocean. Sometimes, though rarely, the warm water of the current has extended to the shores of Europe. In this part of its course across the ocean, it is very difficult to ascertain the breadth of the current, because the warm water expands to the north and to the south to a considerable distance in the sea, where no current can be traced; in the former direction to a degree, or a degree and a half of latitude; and on the southern side, even to two degrees and a half; it has been met with at 3340 and 34° lat. The strongest current is commonly met with between 38° and 39° lat.; and it is the opinion of many intelligent navigators, that the breadth of what may be called a current does not exceed 100, or 120 nautical miles. The warm water sometimes only extends to 140 miles, and then it seems to occupy only the current, but at other times it is found to cover 186, 240, 270, and even 320 miles. It does not seem that this difference in the extent of the warm water is affected by the seasons, for both extremes have been found to exist in the same month (May), between 63° and 65° long. It is very probable that the main current does not always run in the same places, but |