ee oe 
VOL. XVI. (3) THE LOWER SEVERN 257 
joined at Tewkesbury in the form of a capital letter Y. 
Surely, any map will show this to be the general rule. 
Exceptions there are—Mr Marr goes to the Rhone Valley 
to find them—but I do not remember a case which I 
cannot explain. For instance, if a river have a hill on 
one side of it, giving a steep watershed, the channel will 
skirt the hill even if there be a stream coming in on the 
opposite side. So, too, a number of small streams com- 
ing in on one side seem to do more to divert a river 
channel towards that side than a single one of larger size 
will do. In my paper of 1882 I admitted that if “a 
tributary be precipitated down the steep side of a valley” 
it may, “if the force be sufficient,” produce a “ resultant,” 
“according to the law of the parallelogram of forces ” and 
so divert the river away from the tributary stream. I have 
not, however, seen this. In the great majority of in- 
stances tributaries join rivers on the convex side of the 
curves, whatever the size of the river or the character of 
the soil in which it flows. 
A little further down is the great curve, at Stonebench, 
round Minsterworth Ham. Here, as stated in my paper 
of 1882, I first saw reason to believe that river-curves 
could not be explained by obstructions or by any real 
explanation yet given, and suggested the influence of 
tributaries as the real cause. The one just mentioned is 
shown in fig. 2A. I hold, and I give this figure as evi- 
dence, that no one, if given a drawing, of a river's course 
and told nothing of the scale on which it was drawn, 
could tell in what kind of soil its course had been shaped. 
Yet, one may feel sure, some essential principle common 
to both must underlie the formation of courses so differ- 
ent in size and flowing in such very different material as 
in the two cases shown in fig.2. They are both relics of 
old figure-of-eight loopings. In each case arms of loops 
have been effaced because unnecessary. 
