89 



direct and rapid fall to the sea, which enabled the Severn not only to 

 cut its own system adrift from the Thames but thereafter to proceed 

 to rob the latter river of half its tributaries along the Cotteswold 

 escarpment. 



Again a similar struggle has taken place between the Frome 

 and, probably, the Parrot. Here again the Dorset-river has been 

 worsted by its rival working quickly backwards and downwards in 

 the flat Somersetshire plain and up to the escarpment slopes of the 

 retreating West Dorset highlands. It is as well to give the caution 

 that none ot these processes may have been the work of the actual 

 existing rivers as we know them to-day. It is convenient to name 

 them so as to gain a ready topographical guide to the events that are 

 outlined, and indeed these modern rivers are now no doubt the im- 

 mediate representatives of the protagonists of former days. But 

 rivers are always shifting their courses laterally over considerable 

 areas, and in some cases the modern valleys are greatly diminished, 

 and the old bottom gravels may now be found on terraces at a height 

 of 200 or 300 feet above and miles away from the present rivers. 

 Therefore when we speak, for instance, of the Frome of past ages, 

 we mean in effect that river which occupied broadly the same country 

 and had the same general direction as the modern river, which latter 

 now meanders in a valley within that of its ancestor. 



Now the Frome Solent having lost territory not only at its head 

 but also by the incursion of the sea upon its lower course, has this 

 latter catastrophe had any effect upon the lateral tributary streams ? 

 Seeing that the way to the sea has thus been shortened and simpli- 

 fied we might expect to discover some display of enhanced action — 

 to find for instance an F (figs. 1 and 2) in action in some direction. 

 When I had the pleasure of addressing the Society on this subject, 

 I pointed out the fine valley, now dry save for a very insignificant 

 stream, which runs due east into the Test valley at Mottisfont. 

 The railway between Romsey and Salisbury takes advantage of this 

 valley. There is only a low ridge between its head and the neigh- 

 bouring Avon to the west at Alderbury. On the opposite bank the 

 Avon receives a stream, The Ebble, which comes also due eastward 

 from the northern slopes of Cranborne Chase. Not having examined 

 the ground in detail I did not then suggest what seemed nevertheless 

 a very likely conclusion . Mr. Clement Reid, however, in his late 

 paper read before the British Association at Portsmouth, has in- 

 stanced the Hampshire Avon as a once minor tributary of the Frome- 

 Solent, which gaining power by the shortening of its passage to the 

 sea, has worked back and captured the drainage of Salisbury Plain. 

 I do not know if he mentioned this Alderbury Mottisfont valley, but 

 his authority seems to bear out my previous assumption that all that 

 system the upper Avon, Wylye, Nadder, and others that unite near 

 Salisbury once passed by this route into the Test. The lie of the 

 Avon valley sides still seem to show some vestiges of the " col " 

 which once formed the watershed of the head waters of the Avon at 

 a point about a mile south of Downton. 



It may be as well, partly by way ol retrospect, to emphasize one 

 or two points which must be borne in mind in searching for these 



