THE ORIGIN OP THE VALE OF MARSHWOOD. 179 



In order, therefore, to understand the drainage system of this part 

 of Dorset we must imagine a time when the surface of the land 

 sloped gently both northward and southward from the line above 

 mentioned. On this surface there was a certain accumulation of 

 clay, pebbles, cherts, and flints, the heavy and insoluble relics of 

 the Eocene, Greensand, and Chalk which had been destroyed ; 

 remnants of this deposit, which is generally called " the clay with 

 flints," still remain on the tops of the higher hills. 



The rain flowing down the southern slope of this surface 

 gathered into streams, which cut channels for themselves through 

 the Chalk and Greensand. They ran, of course, high above the 

 present surface, and their courses were prolonged far to the 

 southward before reaching the sea ; indeed, during the Miocene 

 and again in the later Pliocene time it is probable that most of 

 the English Channel was dry land, and that these Dorset streams 

 were merely tributaries of a large river which ran westward down 

 the valley of the Channel.* 



$o\v the slope along which these streams made their way was 

 planed across the summit of the low dome or pericline, which has 

 been described, and as we have calculated the base of the Green- 

 sand on this summit to have been about 100 feet higher than it is 

 at Lewesdon, where the thickness of Greensand at present is not 

 more than 130 feet, and as the surface sloped southwards from 

 Lewesdon, there cannot have been much Greensand left over the 

 central area of the pericline when the streams began to make their 

 valleys. Hence, as they deepened their channels they would 

 quickly cut through the Greensand on the central area and would 

 soon enter the Jurassic beds on which the sand rests ; these beds 

 are the Midford Sand, the Upper Lias clay (which is thin), and the 

 Marlstone Sands. 



As soon as any stream cut into the Upper Lias the water on the 

 overlying sands would issue in the form of springs. Thereby the 

 volume of the streams would be increased and at the same time 



* For a restoration of English geography at this time see " Building of 

 the British Isles," by the Author, Plate xiii. 



