There is a river of first-rate magnitude which has no delta, though it seems to possess all those qualities which are supposed to, be requisite to the formation of such an alluvial tract : the St. Lawrence in North America reaches the sea by a kind of bay, which extends upward of 300 miles, and gradually increases in width from three to above one hundred miles. One would suppose that the form of this bay would render it subject to be easily filled up by the earthy matter brought down by a river whose course exceeds 1800 miles ; and yet we, do not find that an alluvium of any extent has been formed along the banks of this wide textuary, except on the right bank below Quebec. This single instance might throw some doubt on the opinion that deltas are formed by rivers in the way above mentioned, if the peculiar nature of the St. Lawrence did not suggest an explanation of this deviation from the common course of things, which rather con firms than refutes the established principle. The St. Lawrence is the only largo river which traverses a great number of lakes. Even after having left the five great Canadian lakes, it passes through those of St. Francis, St. Louis, and St. Peter. In each of these lakes the current ceases, and it is only perceptible where the river again issues from the lake, All the earthy matter therefore which the river collects and keeps suspended in its course from one lake to the other is deposited in these lakes. Thus this large river brings no debris and earthy matter, or very little, to its wide estuary, which can nut therefore be changed into a delta by the accumulation of such matter.
Mast large rivers, as already observed, admit of this division of their course into three parts, an upper, middle, and lower course; but the exceptions are far from being rare. it sometimes happens that the characteristic features by which the middle courts, is distin guished, occur in the upper course. This takes place when a river originates on an elevated table-land, and traverses a considerable part of it. Thus the lulls, the Sutlej (an affluent of the Indus), and the Sampoo, rise on the elevated table-land of Tibet, and drain a portion of it : in this part of their course they resemble the middle course of the Rhine or Danube. But where they leave the plain and enter the mountain-region of the Himalaya, they resemble the moun tain-streams of the Alps, except as to the volume of water. When the Indus and the Sutlej have descended into -the plains of the Punjab, they ammo the character of the lower Rhine and lower Danube. 'the Sampoo, after leaving the monntain-region, traverses a hilly tract of great extent, the valley of Mem, before it enters the alluvial plain of Bengal. There are otber rivers, in which only the characteristic features of the middleand lower course can be recognised: the nnmber dismal is considerable, and sonic of them are of the first magnitude.
Thus the Volga and Mississippi, neither of which rises in a mountain region, hut in a hilly tract, in the greater part of their comae present the characteristics of the middle course of the Rhine and Danube, but tewartle their mouths they traverse a Large plain. The number of rivers whose whole course lies through a hilly or undulating country is still greater, as is the case with nearly all the rivers of England and Southern Scotland. There are also rivers which in their whole course traverse a mountain region, but they are all small ; such are some of the rivers In North Scotland and in Sweden, and nearly all the rivers of Norway, and those on the west coast of South America.
The number of rivers which do not reach the ocean is not great, if we except those which fall into the Caspian Sea and Into the lake of Aral. The other rivers without an outlet always terminate their course in a lake. It was formerly supposed that the water of some of them was absorbed by a dry soil, and that they were lost in the sand ; and this opinion still prevails as to some rivers which descend from Mount Atlas southward to the Sahara. But the point remains doubtful. Among other river, which have no communication with the sea, some few traverse elevated table-lands, consisting of plains surrounded by continuous mountain-ranges, through which the waters cannot find an outlet, and consequently collect in the lower part of the plains, and form lakes large enough to part with all their surplus water by evapo. ration. Such rivers occur in the valley of Tenochtitlan in Mexico. The most remarkable is the Disaguadero, in the valley of Titicaca In Bolivia, which runs about 300 miles, and is hurt in a lake or In swamps. The llyarkan or Yerkan, in Chinese Turkistan, is still larger, hut its character is imperfectly known. Another kind of such Lakes occurs in the plains of Mexico an 1 of South America, And almost exclusively in those parte which have no rain or very little. On the table-land of Mexico the greater number of rivers between 24' and 30' N. lat. terminate In lakes; and in the states which compose the Argentine Republic, rivers of this kind arc numerous between 28' and S. lat., west of 114' W. long. As very little rain falls in some of these coun tries, and in others none at all, the rivers are supplied with water by the rains which fall at certain seasons on the mountains in which they originate, and by the sprinpi which exist there. But as the supply of water is very moderate, it doss not give force sufficient to the currents to carry therm through those extensive tracts which separate them from the sea. It is remarkable that some of these rivers and all the lakes in which they tenninate, are silt in South America ; and it is probable that this is also the case with most of those on the Mexican Latheme.