34 Remarks on tJie Formation of Alluvial Deposites. 



its lower part always extends with the stream to some distance 

 under the water, and then terminates in the usual steep decli- 

 vity. The form of alluvium thus produced is that of an acute 

 cone truncated by an obltise cone. 



I was enabled to observe this process very distinctly in Savoy 

 and Switzerland, in consequence of a practice, common in those 

 countries, of conducting the surface-water of the rivers along 

 artificial channels into small enclosures, which are thus, in the 

 course of a few hours, filled with fine sand. The sand is dug 

 out to improve the cultivated grounds, and the pool is left to 

 fill again. The appearances, of course, vary according to the 

 declivity of the channel, the swiftness and volume of the stream, 

 and the form of the reservoir. The plan and section. Fig. 24, 

 are designed to represent a case as free as possible from adventi- 

 tious circumstances : a 6 is a straight horizontal channel, through 

 which a stream charged with sand constantly flows. The side 

 of the reservoir is also straight, and the stream meets it at right 

 angles at the point b. The reservoir has an overflow at the 

 other end. The stream carries the grains of sand in nearly 

 straight lines, to all parts c of the circumference of the smaller 

 semicircle, from which, as soon as they have reached it, they 

 fall by their own weight down the steep declivity from c to d. 



In natural lakes, we frequently observe the process effected 

 on a greater scale. A small rill, carrying down sand, shews it 

 most distinctly ; and I have seen the form, produced without 

 any artificial arrangement, quite as regular as is represented in 

 the figures. But, as the conoidal deposite is not removed from 

 natural lakes as it is from the reservoirs constructed by alpine 

 husbandmen, the depositions of the stream, retarded as it is by 

 meeting the standing water, tend continually to raise both its 

 own bed and the central part of the semicircular area, over which 

 it is diff'used. Hence this area is gradually raised above the 

 level of the lake, although its border always continues to dip 

 under it. After this the stream either flows (from b) in a single 

 current, or divides into several channels, and, in process of time, 

 vegetation protects C'.'rtain parts of the raised area, so that the 

 stream, instead of acting equally in all directions from the cen- 

 tre {b) comes to be more or less confined, and hence the trun- 

 cated acute cone is more enlarged in some parts than in others. 



An obtuse cone, of great regularity, about half a-mile in dia- 



