1899] THE DEVELOPMENT OF RIVERS 279 



The escarpment breaches connected with these valleys must be 

 separated from those connected with the type of V., which indicate not 

 a through stream but merely local traffic. Their history has been 

 this : when the escarpment stood farther forward there was sufficient 

 collecting ground for the rainfall : a spring could issue at the back of 

 the escarpment and its stream formed a valley ; the escarpment has 

 been worn back ; the collecting ground has gone ; the spring has 

 disappeared ; the escarpment has been worn back farther ; the head of 

 the valley with the place where the spring once rose has been taken 

 off. The head of the valley forms a notch or breach in the present 

 escarpment. 



In drainage restoration a through stream with a distant source 

 must not be accorded to valleys of the type of V., breaching escarp- 

 ments, only a short local stream must be given. 



Among other numerous phenomena of rivers and valleys one 

 further feature must be noticed now, as it is important in drainage 

 restoration. It presents itself as a sharp 

 turn in the course of a stream ; not a 

 local twist, but an entire change of the 

 general direction. 



When a subsequent or obsequent ) 



stream captures a consequent stream an 

 abrupt angle is generally formed at the 

 junction. This angle is called the " elbow 

 of capture " by Professor Davis. 



Fig. 5 Shows elbows of successive FlG - 5.— Elbows of capture near 



/ k . N n rr,, , StrOud. 



capture (A, A) of I names consequents by 



the obsequent Frome ; Fig. 6 illustrates the restoration of the drainage 

 as it was before the captures had been effected. And at the base of 

 the diagram, below Ashton Keynes, is seen the elbow of capture (A) 

 of the there subsequent Thames having taken the Churn. It was the 

 continuation of the Churn which cut the Ogbourne Valley to Marl- 

 borough. It was the capture of the Churn by the Thames which has 

 caused that valley to be different from the Goring gorge. 



The elbow of capture must be distinguished from a right-angle 

 bend made by the joining of a tributary subsequent to a consequent 

 stream ; and especially from the bend which may be at the head of a 

 consequent stream which has been shortened by the successful growth 

 of an obsequent, so that the subsequent stream becomes the longer. 

 The Upper Kennet, where it turns west to the south of Avebury, is a 

 case in point. 



It has been necessary to put forward these general considerations, 

 before attention could be devoted to their particular application. Yet 

 again, before they can be applied to the history of the Severn, the 

 development of the Thames must be noticed. 



III. Development of tlir Thames.— As seen at the present day the 



