8. 8. Buchman — River Development. 373 



post-Carboniferous, post-Jurassic, post-Cretaceous — we obtain a rock- 

 structure witli lines of outcrop arranged as in Fig. 5. Then, if we 

 fill in the names of the various rocks, we shall see that it gives us 

 what may be called the fundamental plan of the geological structure 

 of Mid- Wales, Mid-England. In a generalized way it accords with 

 the geological map ; its discrepancies therefrom are no more than 

 the omission of minor local anticlines would account for, together, of 

 course, with local curvature of anticlines. 



One point before proceeding further. Apparently Caledonian 

 and Charnian directions of strikes and folds are not strictly due 

 to impulse at right angles to the respective directions, that is, from 

 north-west or north-east, but they are the immediate result of 

 a combination of the north and south with the east and west lines 

 of foldings which are dominant in the Eastern Hemisphere. With 

 a Malvernian axis dipping south, Caledonian direction is given to 

 the outcrop of beds on the east, Charnian to those on the west. 

 With an Armorican axis dipping east, Charnian direction is given 

 to the strata on the north, Caledonian direction to the strata on 

 the south. 



We have already seen that, from a dipping anticlinal axis, streams 

 run in an oblique direction. Therefore, from the east of the south- 

 dipping Welsh backbone the streams would first run obliquely — that 

 is, in a south-east direction — at right angles to the Caledonian strike. 

 From the north of the east-dipping Mendip-Pewsey axis the streams 

 would run north-eastwards at right angles to the Charnian strike. 

 In all cases the less the axial dip the less the oblique direction. 

 Then these two series of dip streams would meet together in the 

 synclinal trough north of the Mendip-Pewsey axis, in the line of 

 the Kennet-Thames : that is a trough stream. Such is the initial 

 stage of river development. 



The second stage commences when the covering rock, whatever it 

 may be, has been stripped off enough to expose the outcrops of older 

 rocks ; then differential denudation begins ; the soft rocks are 

 denuded faster than the hard ones, so that the latter are left to form 

 escarpments. Strike streams are started along the outcrops of soft 

 rocks ; the growth of strike streams results in river capture. There 

 is no better illustration of that than the manner in which the dip 

 streams of the Cotteswolds are caught by the strike stream which 

 has developed along the line of soft rocks in the Vale of White 

 Horse. 



The third stage follows on the capture process — rivers working 

 back against the dip are started in the valleys of the beheaded dip 

 streams. They necessarily follow up those valleys because those 

 valleys are the lowest ground, made so by the original dip streams. 

 So the phenomenon of breached escarpments with the water-parting 

 in the low ground of the valley and streams flowing in opposite 

 directions from that parting is seen to be exactly the result of the 

 oourse of events here detailed. 



What is to a certain extent abnormal, so far as the Severn is 

 concerned, is the fact of its being an invader in what is strictly 



