352 T. 8. Ellis— Ewer Curves. 



Now we have the floor of the river channel, including the 

 expanded portion, and that of the affluent at its mouth covered 

 with mud, as shown in Fig. 2. When the river sinks to low water 

 the lowest line of channel has to be determined. Clearly, a channel 

 leading from the affluent down-stream is necessary, so that, if one 

 for the river formed on the opposite side, there would be two. But, 

 in fact, the river "falls into and adopts for itself" (to quote words 

 from my paper) the channel kept open by the affluent. Hence the 

 curve as seen in Fig. 3. The deflected river, having been directed 

 against the bank on the affluent side, washes it away at and below 

 the mouth of the affluent. The process being continued the river 

 recedes, generally until the edge of the alluvium has been reached, 

 unless some other diverting influence turns it back, while the 

 undisturbed mud on the opposite side consolidates as it grows until, 

 becoming covered with verdure, it forms part of the meadow on the 

 river margin. 



I am reminded that streams often wind freely where no affluents 

 are seen. But it must be remembered that, before the area adjoining 

 the stream became covered with grass, the surface drainage had to 

 run direct into the stream on either side. The series of curves 

 taking in affluents which can be seen in a brook running through 

 the mud of a lake when the dam below has been removed illustrates 

 this, and it is well shown in Professor Huxley's diagram of a catch- 

 ment basin given in his Physiography. 



An affluent is sometimes seen entering a river on the concave 

 side of a broad curve, which must be due to some stronger influence 

 diverting the river from it. This may be the drainage from a hill- 

 slope, not, it may be, enough to make any one stream of considerable 

 size, but, by flowing into the river-bed more rapidly, more effectually 

 keeps open a river channel than the sluggish affluent flowing across 

 the plain can do. 



Sir A. Geikie considers " some slight weakness in the river-bank " 

 to be the origin of a river curve. A breach in the continuity of 

 any structure, be it garden wall or river-bank, involves something 

 more than a slight weakness. But in the case of a river-bank 

 a channel must be kept open from the breach seawards, which prevents 

 the formation of any support for the bank near the affluent. Thus 

 there is a double cause of weakness. Large sums of money have 

 been spent in putting elements of strength into river -banks. 

 I suggest that it might be better spent in removing elements of 

 weakness, by uniting affluent streams and bringing them, when 

 united, into the river at fixed points, leaving between them a long 

 unbroken front. 



Some writers allege that the influence of a side-stream is to 

 drive the river to the opposite side. This, as it seems to me, is 

 contradicted on every map. In my paper I likened an ordinary 

 conjunction of two streams to a capital letter Y, where the faint 

 upper line is in direction continuous with the lower part of the 

 strong line. I did, however, recognize that if a stream were pre- 

 cipitated down the steep side of a valley with sufficient force at 



