1849.] On the Embankments of Rivers. 699 



side was but little above the bottoms of the respective hill streams. 

 The mass of waters was thus sustained in the air by natural and artifi- 

 cial means, and perhaps on the occasion of a great flood, finding a lower 

 level within a given distance, as in the direction of a parallel stream, 

 than down their own channels, they at once struck out new courses for 

 themselves towards D, as shown in the sketches. The change in the 

 More took place during the present generation, but thirteen years ago 

 when last seen, it had made but small progress in forming a well 

 defined bed in the lower portions of its new course, and it had covered 

 about twenty square miles of good soil with interlacing ribbons of bar- 

 ren sand. Frisi mentions similar occurrences in the smaller rivers of 

 upper Italy, and even in the Po itself.* 



Courses of rivers in the central plains. — At great distances from the 

 hills, the velocities of streams decrease, large quantities of debritus 

 cease to be borne, and the rivers continually endeavor to sink themselves 

 to the final level of the ocean. Hence when the soil through which 

 they flow is of a firm character, their channels become deep and regu- 

 lar in section, and the uncertain reflections of the upper portions are 

 converted into well formed windings. But the tendency of these wind- 

 ings is to be cut through at the neck, and the variations of channel are 

 thus kept within limits greater or smaller according to the size and 

 force of the river, but usually somewhat narrower than in the upper 

 plains. Generally speaking, in these central tracts rivers do not over- 

 flow the country, as in the case of the Ganges about Mirzapore, but 

 the tortuous courses of smaller streams are sometimes invaded by the 

 inundations of greater rivers running parallel to them. Thus in sketch 

 No. IV. the winding course of the Qweea has now become submerged 

 by the waters of the More, as the old channel of that river (below C) 

 has long been by the inundations of the Bhagiruttee. 



The courses of rivers in the lower plains. — As rivers approach the sea 

 they for the most part flow through plains of small descent, and which are 



* Frisi on Rivers, p. 20. It may be doubted whether the Ganges has yet become 

 finally settled in its belt of variation with reference to the Jumna, which latter river 

 I consider to run on a lower level than the Ganges. Thus with a little aid from 

 art, the one might very likely be led from near Cawnpore into the other below 

 Calpee — a change which would shorten the new canal, if it is to be used mainly for 

 navigation, as it should be. 



4x2 



