ORIGINAL COMMUNICATIONS. 



The Development of Rivers ; and particularly the 



Genesis of the Severn. 



By S. S. Buckman, F.G.S. 



I recognise the difficulty of the task. I do not pretend to be able to 

 perform it adequately ; but I have the hope of placing the subject on 

 what may be a somewhat surer basis for future work. 



I. Introduction. — A few words as to the evolution of my ideas, 

 showing how greatly I am indebted to other people, may not be amiss. 

 In some topographical articles (1), and in a communication to the Geo- 

 logical Society (2) in reference to some valleys now holding tributaries of 

 the Severn, I pointed out that their upper parts could only have been 

 excavated by streams which were tributaries of the Thames ; that 

 therefore there must have been high ground — lost Cotteswold Hills, 

 in fact — to the north and west of the present escarpment; and 

 that the Thames must have been a larger and longer river than now. 

 I said that the diversion of a west-to-east stream, tributary to the 

 Thames, must have been brought about by the working back of a 

 Severn tributary — the one which now forms the Chelt ; and that in 

 course of time the present Chelt must serve the same trick on another 

 Thames tributary, the Coin, because it could give so much quicker fall. 



But in extending the Cotteswold Hills farther to the west, I 

 certainly had no idea of filling- the Severn valley completely with 

 Jurassic rocks, and supposing the non-existence of the Severn itself. 

 However, when I read the paper by Osborne White (3) citing the 

 opinion of Professor W. M. Davis (4), that, in effect, the Severn valley 

 was filled with Secondary strata, and that the Thames streams 

 originally headed back in Wales, what was till then obscure became 

 suddenly clear : the phenomena of the Severn and its tributaries seemed 

 easily understandable. And the more the idea is worked out the more 

 does the evidence in its favour seem to accumulate. 



I ought perhaps to make this somewhat clearer. Of course it has 

 always been recognised that the Severn valley was originally filled up 



19 NAT. SC. VOL. XIV. NO. 86. 273 



