the River Severn at Worcester. £51 



are the great preponderance of sandy and clayey rocks of various 

 degrees of induration, and of the saliferous marl of the New Red 

 Sandstone. Hence the river before it leaves Montgomeryshire, 

 where its channels have been confined to rocks of a more inso- 

 luble and harder character, and where its sources have been re- 

 cruited from a more abundant rainfall, is of necessity purer than 

 at any other part of its course. The difference between its purity 

 in this county and at Worcester is well shown by contrasting the 

 results obtained by the late Dr. R. D. Thomson* with my own 

 observations. Dr. Thomson examined the Vyrnwy and its most 

 important tributary the Banw at points not far from their origin, 

 the Clywedog and the Hafren or Severn at Llanidloes, and the 

 Severn lower down at Caer Sws, soon after the accession of 

 streams which must evidently have brought to it a purer water. 

 His results are as follows : — 



Grains in the gallon. 



Total Organic 



impurity. impurity. 



River Vyrnwy at Pont Hagel . . . 2*60" '72 



„ Banw at one mile below Can Office 2'84 1*16 



„ Clywedog at Llanidloes Bridge . 6*80 1-40 



„ Severn and Afon Dulas at Llanid-"~l q i q 9 -■ .gp 



loes Bridge J 



„ Severn at Caer Sws .... 7'24 1-44 



With these I will now compare the maximum and minimum 

 amounts of dissolved matter which I have found in the Severn 

 at Worcester : — 



Grains in the gallon. 



, A . 



River Severn at about a mile above" 



Total Loss by 



impurity. ignition. 



' Loj-OQ "I •T'zL 



Worcester after drought, July 1866J 

 Ditto after flood, February 1867 . . 11*65 -88 



The circumstances, however, in which the water at Worcester 

 contains the above minimum of dissolved matter are quite excep- 

 tional, and in all ordinary conditions of the river I have found 

 it to contain from 21 to 22 grains in the gallon. From the want 

 of complete analyses of the river at higher points, it is impossible 

 to trace the entrance of the constituents which cause this great 

 increase of impurity • yet when we consider that for almost its 

 entire course through Shropshire and Worcestershire the Severn 

 forms the principal drain of the New Red Sandstone and of the 



* Quoted by Mr. Bateman in his pamphlet on the supply of ^Yater to 

 London from the sources of the Severn, November 18G5. 



S2 



