274 £ & BUCKMAN [april 



with Jurassic rocks and buried beneath Cretaceous strata; but it had 

 not been understood that the Thames streams flowed across this area for 

 a long time before the Severn had existence. It seemed to be supposed 

 that the Severn and the Thames were contemporaneous, that the Severn 

 cut a channel between Palaeozoic and Mesozoic rocks, and so initiated 

 the Mesozoic escarpments ; that the Cotteswold escarpment extended 

 farther west to nearer the present position of the Severn ; and has only 

 been worn back therefrom by the further cutting done by the Severn. 



Yet I afterwards learnt that some dozen years before Professor 

 Davis put forth his views, Dr. T. S. Ellis (5), of Gloucester, had 

 imagined the possibility of rivers flowing across a non-existent Severn 

 valley. He says, " If we could have it admitted that the trend of the 

 country was originally to the east before the present [Severn] valley 

 was formed at all, following a slope as the Cotteswolds now dip ; then 

 one could imagine a stream flowing through the gap between Malvern 

 and May Hills along the line of the Leadon and escaping over the 

 Cotteswolds at Witcombe, so first marking out one of the west-to-east 

 streams of which apparent signs remain." 



To his paper he appends a map showing the application in three 

 cases : — A. The Severn to Tewkesbury flowing up the valley of the 

 Swillgate to Cheltenham, and up the Chelt valley to the Coin at 

 Andoversford ; B. The Leadon flowing past Gloucester and over the 

 Cotteswolds near Birdlip, and so into the Thames system ; C. A river 

 from west of May Hill flowing up the Stroud valley along the course 

 of the Frome. 



In his arguments for his theory from the physical features of the 

 country he makes another good point. " The peculiar course of the 

 tributaries of the Severn in our district seems to have a bearing on my 

 suggestion that the first streams flowed across the line of our valley. 

 It is remarkable that nearly all those on the left bank flow towards the 

 river in a direction against that of the Severn itself. . . . Can these 

 tributaries on the left side be occupying channels originally marked out 

 by streams flowing in the opposite direction ? " 



Lately I have had the great advantage of accompanying Professor 

 Davis to some of our Cotteswold streams, and he has urged me to give 

 an account of the Severn. 



This introduction is rather long ; but I wish to show my indebted- 

 ness to others. I feel that the ideas to be set forward are not mine 

 in any great degree. And, in fact, I know not how much is from 

 myself, or how much has been suggested by the written or verbal 

 communications of others. 



II. General Considerations. — At the first initiation of the drainage 

 of an area the direction of streams is determined by the dip. Such 

 streams are termed consequents by Professor Davis (4). Now it is 

 necessary to consider what would happen in a tilted area composed of 

 rocks of varied power in resisting denudation, and different porosity. 



