260 PROCEEDINGS COTTESWOLD CLUB 1909 
joins the stream coming in at Berkeley Pill which is, as 
nearly as may be, opposite. At both of these points 
there is a considerable inflow of water requiring a channel 
to convey it to the sea, but the predominant influence is 
on the left side. The original opening of the Gloucester 
and Berkeley Canal at Sharpness Point and the present 
opening a little lower down, co-operate with the stream 
coming in at Berkeley Pill to keep the main channel on 
this side. 
Thus, as appears, there is a disposition towards one 
good channel on the left side and the formation of new 
land on the right or Lydney side, but, in order to pro- 
mote this, a long front, unbroken by the entrance of any 
stream, is highly desirable. This would be attained by 
intercepting all above the Guscar rocks and bringing the 
united streams into one channel passing through between 
the two natural piers which they form. In my paper of 
1882, I remarked that the natural history of rivers seemed 
“to have a bearing on the formation of warths (new lands) 
and the maintenance of navigation channels.” This, I 
think, is illustrated here as it is at The Nooze. 
The two channels, one on each side, shown in fig. 10, 
are seen to be united by one directed obliquely across the 
stream past the Guscar rocks. This forms a looping 
which closely resembles the looping of inland streams. 
The apparent breaks where one line of stream joins 
another show the accuracy of the survey. They are not 
really breaks in the lines of stream, but shallows formed 
by strong currents passing the outlets of relatively weak 
streams and projecting some of their silt into the channels 
of these weaker streams, less perfectly scoured than their 
own. 
The main channel, having remained on the left side 
for some distance, crosses over to the right in order to 
pass through the opening between the Narlwood rocks 
