having a depth of above 900 feet, have their beds above 300 feet below the surface of the ocean; Lake Ontario, with a depth of 500 feet, reaches a point 268 feet below the sea-level. The depth of the St Lawrence river-bed, as related to the sea, is not ascertained. Most of the Swiss lakes, too, having a depth often of more than 1000 feet, come under the same category with the lakes under consideration, waters from the mountains having gradually filled up chasms made at the time of the upheaval of the adjacent region. Some of these lake-basins may be deep enough to lie below the level of the ocean.

THE BITTER LAKES OF THE SUEZ ISTHMUS.

Some bitter salt lakes on the Isthmus of Suez, forming a chain from the Red Sea to the south-east corner of the Mediterranean, long claimed attention from their supposed singularity. During the occupation of Egypt by the French in 1799, a survey of the district was made with the level, in view of a prospective canal across the isthmus, connecting the Nile with the Red Sea. An account of that survey was published by Le Père, in his great 'Déscription de l'Egypte.' The result of the survey was very surprising; it assigned to the Gulf of Suez a height of 25 feet at ebb tide and 30 feet at flood tide above the level of the Mediterranean- -a result which seemed to agree with Pliny's account (vi. 23) of the elevation of the Red Sea above the level of lower Egypt. The salt swamps lying between the two seas, and known even to the ancients, lie, according to the same authority, 20 feet below the surface of the Mediterranean, and 50 feet below that of the Red Sea. These singular statements were not

K

received without considerable doubt as to their correctness; but during the military disturbances in that region, no revision of the investigations could be made. Certain circumstances connected with an unusual inundation of the Nile in 1800, when its waters flowed as far as the transverse valley called the Wady Tumilet, in which the salt lakes lie, and where traces of the ancient canal, constructed by the Egyptians between the seas, could be seen, seemed to confirm the result of the survey of 1799. The inference was a natural one-that the sandy Isthmus of Suez was an accumulation of dunes, and of the deposits of inundations of both the Mediterranean and Red Seas, and that the salt morasses in the middle are but a trace of the primitive bottom. There were not wanting defenders of the old measurement, Favier being the most prominent. Since 1845, five surveys have been made, in reference to the projected canal. These all contradict the results of 1799, and show that there is but the difference of four-sevenths of a foot between the level of the two seas, and that there is the same agreement there as in all other parts of the earth. Many hypotheses, built on the old measurement, have accordingly fallen to the ground.

THE REGIONS OF TRANSITION BETWEEN HIGHLANDS AND LOWLANDS; THE RIVER-SYSTEMS OF THE GLOBE.

Between the two great and most sharply-marked physical features the high plateaus and mountains, and the lands of very little elevation-there are regions of transition very numerous and exceedingly varied.

The conception of highlands and of lowlands having a

certain, constant, and absolute value, and it being immaterial whether the elevation be specially marked or not, provided it be uniform, the regions of transition find their most marked characteristics in their want of constancy, in their very change, and the rate at which the grade ascends from a low to a high elevation, or falls from a high to a low one. Their real value lies in the mutual compensation of highlands and lowlands, which is effected through the mediation of a third physical feature or system, which has received the name Lands of Gradation, or Terrace-Lands: these, by their gradual rise from the sea-level, serve as the means of transition from the lowest lowlands to the loftiest plateaus and mountains.

TERRACE-LANDS AND RIVERS IN THEIR GENERAL
CHARACTER.

Districts sloping to the sea, or lands of gradation, as we have called them, varying as they do in elevation and in relative situation to each other, are the true mediators between the districts but little above the level of the sea and others much more lofty. At the sources and the mouths of rivers they partake, more or less fully, in the characteristics of both highland and lowland. The manner of their mediation, as determined by the rate of the fall of water and by their direction, gives to every one of these regions of transition its peculiar character, and determines its conformation and its relation to the globe. And yet, no more than in lowlands and highlands, can we rid ourselves of some arbitrary data relating to the size of rivers, when we discriminate between those which we call large and those which we call small. As in all

other geographical distinctions, we must here be content with arbitrary approximation, and with the ordinary usages of speech. The comparison of streams, in regard to their breadth and fulness, determines their volume; the comparison, in respect of length and tributary waters, determines the compass of the river-system. The entire characteristics, breadth, depth of channel, length and extent of drainage, determine the status of the river, whether first, second, or third class, in relation first to those of the same continent, and then to those of the world. The Volga, for instance, is, in relation to Europe, a firstclass river, but, like the Danube, in relation to the entire globe, is merely in the second or third rank. Not the length alone determines the importance of rivers. The Thames, one of the smallest streams in Europe, is one of the most important. And aside from commercial considerations, a river of insignificant size may have great influence in consequence of its relation to the entire adjacent region. The little Bavarian Isar—a river which, so far as the great world is concerned, seems to have no importance-receives on its left side the water of 860 tributary brooks, among which are 44 rivulets; on its right the water of 433: these 1293 brooks and rivulets pour themselves into the Isar through 103 direct tributaries; and not these alone, but the waters of 136 lakes are embraced within the Isar system! Yet the Isar is only one of 24 branches of the Danube, and of the fourth rank even among these, and the Danube is by no means one of the great rivers of the globe. A short but navigable stream can have great influence over a territory limited in extent, and may make a long but shallow stream sink into insignificance in respect of comparative importance to the world. There are some great streams which

are of first magnitude in all their characteristics-rivers which drain hundreds of thousands of square miles in their course to the sea. The number of such is small, however; there are scarcely fifty on the whole globe. Besides these, there is a large number of rivers much shorter, and of much less volume, but not deficient in the attributes which give a stream value to man, and which serve to mediate between highlands and lowlands, to fulfil the requirements of navigation, and to drain regions of more or less magnitude. These can be classed in four ranks in the first place absolutely, and in the second place in relation to each continent. Yet, in classing them, it is necessary always to keep in mind that it is not size alone which gives a river its value, but a combination of all its characteristics, and its relative influence on the country through which it runs.

Looking at the direction of streams, we observe that there are some which flow northerly, as for instance those of Siberia, the Nile, the Rhine, the Elbe, and the Weser; there are those which flow southward--the Indus, Ganges, Euphrates, La Plata, Mississippi, and Volga, for example ; there are those flowing eastward-the Hoang-ho and Yang-tse-Kiang, the Amazon, the Orinoco, and the Danube, for instance; and some westward, instances of which may be found in the Gihon and Sihon, the Senegal, Gambia, Niger, the Colorado, the Seine, Loire, Garonne, and the Spanish rivers which enter the sea in Portugal.

And this characteristic, trite and unmeaning as it may at first seem, establishes, for the area which these rivers water, very diverse conditions. In like manner, too, their position, in relation to the oceans into which they flow, is very influential, in consequence of the action of the tide upon the lower course. The emergence of their