48 ON THE GENERAL DISPOSITION OF THE 



in the sea. Lakes are, in some instances, the imme- 

 diate sources of rivers, but rather in a geographical 

 than a physical sense ; as they are themselves fed by 

 streams, and ought rather to be considered as basins 

 interposed in their courses. It is not in Loch Tay 

 that we must seek the source of that river, but in the 

 innumerable streams which contribute to swell the 

 Lyon and the Tumel, the Isla, the Garry, the Lochy, 

 the Dochart, and the Almond. 



As rivers which are fed either by snow or rain vary 

 irregularly in summer and winter, so those that are 

 much indebted to snows, as the Wolga, are fullest in 

 the early heats of May and June. From the same 

 causes, many mountain streams in Peru and Chili, 

 flow only in the day. A clue to discover the origin of 

 rivers whose sources are unknown, is thus sometimes 

 offered, by considering the places and tynes of perio- 

 dical rains or summer thaws, and by comparing these 

 with the increase of their waters. This species of 

 reasoning has often been applied to the questions, yet 

 obscure, which concern the origin and courses of the 

 Niger and the Nile. From these periodical rains or 

 thaws, arise the inundations which are the sources of 

 fertility to Egypt, and which are conspicuous in the 

 Indus, the Ganges, the Plata, and many other rivers 

 throughout the world. 



The direction of rivers is necessarily regulated by 

 the form and elevation of the lands whence they arise 

 and through which they flow ; and, as far as particular 

 tracts are concerned, the division of streams to oppo- 

 site directions, points to the highest level. The par- 

 ticular history of these courses is often interesting, but 

 would lead beyond the bounds to which the present 

 sketch must be limited. The length of the courses of 

 some of the larger rivers of the world is very great ; 



