Dictionary of Greek and Roman Geography/Atlas

ATLAS (Άτλας: adj. Άτλας, fem. Άτλαντίς: , Άτλαντίκός, Atlanticus, Atlantēus), a name transferred from mythology to geography, and applied to the great chain of mountains in the NW. of Africa, which we still call by the same name. But the application of the name is very different now from what it was with the ancients. It is now used to denote the whole mountain system of Africa between the Atlantic Ocean on the W. and the Lesser Syrtis on the E., and between the Mediterranean on the N. and the Great Desert (Sāhăra) on the S.; while, in the widest extent assigned to the name by the ancients, it did not reach further E. than the frontier of Marocco; and within this limit it evidently has different significations. To understand the several meanings of the word, a brief general view of the whole mountain chain is necessary.

The western half of North Africa is formed by a series of terraces, sloping down from the great desert table land of North Central Africa to the basin of the Mediterranean; including in this last phrase that portion of the Atlantic which forms a sort of gulf between Spain and the NW. coast of Africa. These terraces are intersected and supported by mountain ranges, having a general direction from west to east, and dividing the region into portions strikingly different in their physical characters. It is only of late years that any approach has been made to an accurate knowledge of this mountain system; and great parts of it are still entirely unexplored. In the absence of exact knowledge, both ancient and modern writers have fallen into the temptation of making out a plausible and symmetrical system by aid of the imagination. Thus Herodotus (ii. 32, iv. 181) divides the whole of N. Africa (Libya) W. of the Nile-valley into three parallel regions: the inhabited and cultivated tract along the coast; the Country of Wild Beasts (ή ζηριώδης) S. of the former; and, S. of this, the Sandy Desert (ψάμμος καί άνυδρος δεινώς καί έρήμος πάντων, comp. iv. 184, sub fin.), or, as he calls it in iv. 181, a ridge of sand, extending like an eyebrow (όφρύη ψάμμης) from Thebes in Egypt to the Pillars of Hercules. A similar threefold division has been often made by modern writers, varying from that of Herodotus only in naming the central portion, from its characteristic vegetation, the Country of Palms (Beled-el-Jerid); and the parallel chains of the Great and Lesser Atlas have been assigned as the lines of demarcation on the S. and in the middle. Such views have just enough foundation in fact to make them exceedingly apt to mislead. The true physical geography of the region does not present this symmetry, either of arrangement or of products. It is true that the whole region may be roughly divided into two portions, the cultivated land and the sandy desert (or, as the Arabs say, the Tell, and the Sāhăra), between which the main chain of Atlas may be considered, in a very general sense, as the great barrier; and that there are districts between the two, where the cultivation of the soil ceases, and whore the palm chiefly, but also other trees, flourish, not over a continuous tract, but in distinct oases: but even this general statement would require, to make it clear and accurate, a more detailed exposition than lies within our province. In general terms, it may be observed that the Tell or com-growing country, cannot be defined by the limit of the Lesser or even the Great Atlas (terms themselves far from definite), but that it even extends, in some places (as in Tunis), beyond the latter chain; that the Sahara or sandy desert, spreads itself, in patches of greater or lesser extent, far to the N. of the great desert table-land, which the name is commonly understood to denote; that the palm-growing oases (wadys) are found in all parts of the Sahara, on both sides of the Atlas, but chiefly in aeries of detached oases, not only on the N., but also in the S. margin of the main chain of mountains; and that, where any continuous tract can be marked out as a belt of demarcation between the Tell and the Sahara, its physical character is that of pasture-land with numerous fruit-trees of various species. The Tell is formed by a series of valleys or river-basins, lying for the most part in the mountains near the coast, which form what is called the Lesser Atlas; and opening out, in the NW. of Morocco, into extensive plains, which, however, the larger they become, assume more and more of the desert character, for the obvious reason that they are less completely irrigated by the streams flowing through them. The lower mountain ridges, which divide these basins, seem generally well wooded; but, as they form the strongholds of the Berbers, they are little known to the Europeans, or even to the Arabs. The southern limit of the Tell cannot be defined by any one marked chain of mountain; but in proportion as the main chain retires from the sea, so does the Sahara gain upon the Tell; and, on the other hand, where, as in Tunis, the main chain approaches the sea, the Tell even reaches its southern side.

To the S. of the Tell, the Sahara, in the Arab sense of the word, extends over a space which can be toerably well defined on the S. by a chain of oases, running in the general direction of WSW. to ENE. from the extreme S. of the empire of Marocco, in about 28° or 29° N. lat., to the bottom of the Lesser Syrtis, between 33° and 34°. As far as can be judged from the very imperfect data we possess, this series of oases marks a depression between the S. slopes of the Athas system and the high tableland of the Great Desert. It thus forms a natural boundary between the "Barbary States," or that portion of North Africa which has always fallen more or less within the history of the civilized world, and the vast regions of Central Africa, peopled by the indigenous black tribes included under the general names of Ethiopians or Negroes. To the S. of this boundary lies the great sandy desert which we commonly call the Sahara; to the N., the Sahara of the Arabs of Barbary: the physical distinction being as clearly marked as that between an ocean, with here and there an island, and an archipelago. The Great Desert is such an ocean of sand, with here and there an oasis. The Sahara of Barbary is "a vast archipelago of oases, each of which presents to the eye a lively group of towns and villages. Each village is surrounded by a large circuit of fruit-trees. The palm is the king of these plantations, as much by the height of its stature as the value of its products; but it does not exclude other species; the pomegranate, the fig, the apricot, the peach, the vine, grow by its side." (Carette, l'Algérie Meridionale, in the Exploration Scientifique de l'Algérie, vol. ii. p. 7.) Such is the region confounded by some writers with the Desert, and vaguely described by others as the Country of Palms, a term, by the bye, which the Arabs confine to the Tunisian Sahara and its oases. As for Herodotus's "Country of Wild Beasts", whatever may have been the case in his time, the lion and other beasts of prey are now confined to the mountains, and do not venture down into the plains. The inhabitants of the Sahara are connected with the peoples N. of them by race and by the interchange of the first necessaries of life, receiving the corn of the Tell, and giving their fruits in return; while they are severed from the peoples of the S. by race, habits, and the great barrier of the true sandy desert. A particular description of the oases of the Sahara, and of the other points only indicated here, will be found in the work just quoted.

The only delimitation that can be made between the Tell and the Sahara is assigned by the difference of their products. But, even thus, there are some intervening regions which partake of the character of both. Carette traces three principal basins of this kind in Algeria: the eastern, or basin of lake Melrir, S. of Tunis and the E. part of Algeria, and W. of the Lesser Syrtis, characterized by the culture both of com and fruits; the central, or basin of El-Hodna, far NW. of the former, where both kinds of culture are mixed with pastures; and the W., or basin of the upper Shelif (the ancient Chinalaph), where cultivation is almost superseded by pasturage.

Such is a general view of the country formed by what we now call the Atlas system of mountains, the main chain of which defines the S. margin of the basin of the Mediterranean. The precise determination of this main chain is somewhat difficult. Its general direction is not parallel to that of the whole system; but it forms a sort of diagonal, running about WSW. and ENE., and nearly parallel to the line of oases mentioned above as the southern limit of the system. The true W. extremity seems to be a C. Ghir or Ras Aferni, about 80° 35' N. lat; and the E. extremity is formed by the NE. point of Tunis, Ras Addar or C. Bon. At this end it communicates, by branches thrown off to the S., with the mountain chain which skirts the eastern half of the Mediterranean coast from the Lesser Syrtis to the Nile valley; but this latter range is regarded by the best geographers as a distinct system, and not a part of the Atlas. The first part of the main chain, here called the High Atlas, proceeds in the direction above indicated as for as Jebel Miltsin, S. of the city of Marocco, where it attains its greatest height, and whence it sends off an important branch to the S., under the name of Jebel Hadrar, of the Southern Atlas, which terminates on the Atlantic between C. Nun and C. Jubi. The main chain proceeds till it reaches a sort of knot or focus, whence several ranges branch out, in 31° 30' N. lat and 4° 50' W. long. It here divides into two parts; one of which, retaining the name of the High Atlas, runs N. and NE. along the W. margin of the river Mulvia (the ancient Malva or Molochath), terminating on the W. of the mouth of that river and on the frontier of Marocco. From this range several lateral chains are thrown off to the N. and W., enclosing the plains of N. Marocco, and most of them reaching a common termination on the S. side of the Straits of Gibraltar: the one skirting the N. coast is considered as the W. portion of the Lesser Atlas chain, to be spoken of presently. From the usage of the ancient writers, as well as the modern inhabitants of the country, this so-called High Atlas has the best claim to be regarded as the prolongation of the main chain. But, on the ground of uniformity of direction, and to preserve a continuity through the whole system, geographers assign that character to another range, which they call the Great Atlas, running from the same mountain knot, with an inclination more to the E., forming the SE. margin of the valley of the Mulvia, and, after an apparent depression about the frontier of Marocco, where it is little known, reappearing in the lofty group of Jebel Amour in the meridian of Shershell, and thence continuing, in the direction already indicated, to C. Bon. Parallel to this range, and near the coast of the Mediterranean, from the mouth of the Mulvia to that of the Mejerdah (the ancient Bagradas) in Tunis, runs another chain, commonly called the Lesser Atlas, which may be regarded as an eastern prolongation of the High Atlas of N. Marocco; while its ridges may also be viewed as the walls of the terraces by which the whole system slopes down to the Mediterranean. These ridges are varied in number and direction, and the valleys formed by them constitute the greater portion of the Tell: the varied positions and directions of these valleys may be at once seen by the courses of the rivers on any good map of Algeria. In few places is there any tract of level land between the north side of the Lesser Atlas and the coast. Besides the less marked chains and terraces, which connect the Lesser Atlas with the principal chain, there is one well defined bridge, running WNW. and ESE. from about the meridian of Algier (the city) to that of Constantineh, which is sometimes described as the Middle Atlas; but this term is sometimes applied also to the whole system of terraces between the Great and Lesser Atlas. In the N. of Tunis (the ancient Zeugitana) the two chains coalesce.

The principal chain divides the waters which run into the Mediterranean (and partly into the Atlantic) from those which flow southwards towards the Great Desert. The latter, excepting the few which find their way into the Mediterranean about the Lesser Syrtis, are lost in the sands, after watering the oases of the Sahara of Barbary. Of the former, several perform the same office and are absorbed in the same manner; but a few break through the more northern chains and flow into the Mediterranean, thus forming the only considerable rivers of N. Africa: such are the Mulvia (Molochath) and Mejerdah (Bagradas). Of the waters of the Lesser Atlas, some flow S. and form oases in the Sahara; while others find their way into the Mediterranean, after a circuitous course through the longitudinal valleys described above; not to mention the smaller streams along the coast, which fall directly down the N. face of the mountains into the sea. Reference has already been made to the common error, which assumes to determine the physical character of the country by lines of demarcation drawn along the mountain ranges. On this point, Carette remarks (p. 26) that "in the east and in the centre, the region of arable culture passes the limits of the basin of the Mediterranean; while on the west, it does not reach them."

As to elevation, the whole system declines considerably from W. to E., the highest summits in Marooco reaching near 13,000 feet; in Tunis, not 5000. In its general formation, it differs from the mountains on the N. margin of the Mediterranean basin, by being less abrupt and having a tendency rather to form extensive table -lands than sharp crests and peaks.

The portion of this mountain system E. of the Molochath was known to the ancients by various names. [Maurentania: Numidia.] The name of Atlas seems never to have been extended by them beyond the original Mauretania (Tingitana), that is, not E. of the Molochath. The earliest notices we find are extremely vague, and partake of that fabulous character with which the W. extremity of the known earth was invested. On the connection of the name with the mythical personage, nothing requires to be added to what has been said under Atlas in the Dictionary of Mythology and Biography.

As a purely geographical term, the name occurs first in Herodotus, whose Atlas is not a chain of mountains, but an isolated mountain in the line of his imaginary crest of sand, which has been already mentioned, giving name to a people inhabiting one of the oases in that ridge. [Atlantes.] He describes it as narrow and circular, and so steep that its summit was said to be invisible: the snow was said never to leave its top either in summer or winter; and the people of the country called it the pillar of heaven (iv. 184). The description is so far accurate, that the highest summits of the Atlas, in Marocco, are covered with perpetual snow; but the account is avowedly drawn from mere report, and no data are assigned to fix the precise locality. With similar vagueness, and avowedly following ancient legends, Diodorus (iii. 53) speaks of the lake Tritonis as near Ethiopia and the greatest mountain of those parts, which runs forward into the ocean, and which the Greeks call Atlas.

It was not till the Jugurthine War brought the Romans into contact with the people W. of the Molochath, that any exact knowledge could be obtained of the mountains of Mauretania; but from that time to the end of the Civil Wars the means of such knowledge were rapidly increased. Accordingly the geographers of the early empire are found speaking of the Atlas as the great mountain range of Mauretania, and they are acquainted with its native name of Dyrin (Δύριν), which it still bears, under the form of Idrâr-n-Deren, in addition to the corrupted form of the ancient name, Jebel-Tedla. The name of Deren is applied especially to the part W. of the great knot.

Strabo (xvii. p. 825) says that on the left of a person sailing out of the staits, is a mountain, which the Greeks call Atlas, but the barbarians Dyrin; from which runs out an offset (πρόπους) forming the NW. extremity of Mauretania, and called Cotes. [Ampelusia]. Immediately afterwards, he mentions the mountain-chain extending from Cotes to the Syrtes in such a manner that he may perhaps seem to include it under the name of Atlas, but he does not expressly call it so. Mela is content to copy, almost exactly, the description of Herodotus, with the addition from the mythologers "caelum et sidera non tangere modo vertice, sed sustinere quoque dictus est" (iii. 10. § 1). Pliny (v. 1) places the Atlas in the W. of Mauretania, S. of the river Sala, (or, as he elsewhere says, S. of the river Fut) and the people called Autololes, through whom, he says, is the road "ad montem Africae vel fabulosissinum Atlantem." He describes it as rising up to heaven out of the midst of the sand, rough and rugged, where it looks towards the shores of the ocean to which it gives its name, but on the side looking to Africa delightful for its shady groves, abundant springs, and fruits of all kinds springing up spontaneously. In the day-time its inhabitants were said to conceal themselves, and travellers were filled with a religious horror by the silence of its solitudes and its vast height, reaching above the cloods and to the K^tere of the moon. Bat at night, fires were seen blazing on its crests, its valleyB were enlivened with the wanton sports of A^pans and Satyrs, and resonnded with the notes of pipes and fiates and with the cUng of dnims and cymbals. He then aHades to its being the scene of the ad- ventures of Hercnles and Perseus, and adds that the distance to it was immense. On the autborily of the voyage of Polybius, he places it in the extreme S. of Mauretania, near the promontoxy of Hercnles, opposite the island of Ceme. (Comp. vi. 31. s. 36.) Af^er Ptolemy, king of Mauretania, had been de«  posed by Claudius, a war arose with a native chief- tain Aedemcm, and the Roman arms advanced as &r as Mt. Atlas. In spite, however, of this opportunity, and of the resources of five Roman colonies in the province, Pliny insinuates that the Romans of eques- trian rank, who commanded the expedition, were more intent on collecting the rich products of the oonntry, to subserve their luxury, dian on making inquiries in the service of science : they collected, however, some information from the natives, which Pliny repeats. His own oontemporaxy, Suetonius Paufinus, was the first Roman general who crossed the Atlas : — a proof, by the bye, that the Marocco mountains only are referred to, for those of Algeria had been eroded by Roman armies in the Jugur- thine War. He confirmed the accounts of its great height and of the perpetual snow on its summit, and related that its lower slopes were covered with thick woods of an miknown species of tree, some- what like a cjrpress. He also gained some informa- tion respecting the country S. of the Atlas, as fiir as the river Geb. Pliny adds that Juba II. had given a similar account of the Atlas, mentioning especially among its products the medicinal herb eupkorina. Solinus (c 24) repeats the account of Phny almost exactly.

Ptolemy mentions, among the points on the W. eoast of Mauretania Tingitana, a mountain called Atlas Minor ^Aras ixdrrwr) in 6^ long, and 33P ]<y N. lat, between the rivers Duns and Cosa (iv. 1. § 2); and another mountain, called Atlas Major ^Aras fxei^tcv), the southernmost point of the province, S. of the river Sala, in 8^ long, and SeP Zff N. kt. (ib. § 4). These are evidently pro- montories, which Ptolemy regarded, whether rightly or not, as forming the extremities of portions of the chain ; bnt of the inland parts of the range he gives no information. (I^w, Tlraoeb, ^.; Pellissier, Memoiret hittoriqueM ei geographigue^ tur VA Igerie^ in the Exploration^ ^., vol. vi. pp. 316, foil.; Jackson, AeeowU of Marocco^ p. 10; Ritter, Erd- ttmde, vol. L pp. 883, foil.)[ P. S. ]