246 PROCEEDINGS COTTESWOLD CLUB 1909 
hardness or softness of rock-formation may decide this, or 
it may be that one channel will serve the purpose of two 
united streams better than any other, and may so attract 
neighbouring streams as to divert them from their courses. 
Effacement of the channel at a particular point may compel 
the stream above it to go back on its course, describing a 
curve just as shown in fig. 2. So, too, effacement of a 
A B 
Fic. 2.—(A) Ten miles of the Severn below Gloucester; (8) One thousand miles of the 
Nile below Khartoum: to illustrate the similarity of curves of different sizes, 
and in different geologic conditions. 
channel may result in, or may be the result of, the separ- 
ation of one part of a drainage area from that in which it 
had been included. Former looping at a higher level is, 
I think, well illustrated by the case shown in fig. 3. If, 
too, we look at the relation between the upper part of the 
Leadon and that of the Coddington brook behind the 
Malvern Hills, separated only by a slight elevation, a former 
union suggests itself. The two streams flow in opposite 
directions, but both ultimately reach the Severn. A little 
north of Leominster there is a divergence of streams, one 
going to the Teme and on to the Severn, the other to the 
Lugg, a tributary of the Wye, and so, ultimately, to the 
