108 



at levels which must have been subject to alternations of encroachment 

 and silt deposit, long after the districts mentioned were dry land, 

 subject only to atmospheric disintegration ; and may we not also, 

 attribute the persistency with whicli mariue plants cUng to such 

 spots as Longdon Marsh, not so much to its having been a shallow 

 backwater of the Severn, as to its having been a deep hollow of the ol d 

 strait, forming, when isolated from it, a reservoir, first of truly salt and 

 then of brackish water, whose saline particles are still to a gi-eat extent 

 retained by the soil. Upon this supposition we can conceive why such 

 plants as Scirpus mcuritimus, and others mentioned in Buckman's Essay, 

 are found growing here and in similar localities, thoiigh absent in others 

 throughout tracts more frequently inundated by the Severn, and constantly 

 washed by its tides. 



We may here make a few brief remarks upon the correlation of the 

 lake period, upon which Mr. Symonds has more than once ably addressed 

 us, and of the existence of which no more doubt can be entertained 

 than of the Strait of Malvern, with the freshwater deposit to wliich 

 I have called your attention. 



The conversion of the bed of the strait into dry land must, of course, 

 be attributed to its elevation, and not to the i-ecession of its waters. 



The physical characters of the country, through which the Severn 

 flows, suffice to inform us that its course must have been baiTed at 

 various points, by obstacles which have since been cleared away by its 

 own long operating powers. 



We know that the oolitic capping of the hills gi'adually thins out 

 westward, and from the entire absence of any trace of it, except in its 

 drifts, we believe that it did not extend from its present escarpments 

 farther than the upturned edge of the forest coal basin in this district, 

 against (not upon) which it rested conformably with the lias beds, which 

 we still find in that position at Awre, Westbuiy, Purton, and Beachley ; 

 we can therefore readily imagine the damming iip of the waters of the 

 Severn and its tributaries, to involve the existence of series of large 

 freshwater lakes, for any indefinite period. 



Now if we glance at the course of the river upon the geological map, 

 we must remark -that it has finally forced itself into communication 

 with the sea, at px-ecisely the point, which, given the sti-atigi-aphical 

 conditions, might have been j^redicated ; where, in fact, the iinconformable 

 conjunction of two rock-masses of very difi'erent degrees of coherence, 

 one being particularly susceptible of frost and general atmospheric 

 influences, afforded the most direct and readiest line of egress. 



We know that the rocks upon the opposite side of the channel, which 



