VOL. XC.] PHILOSOPHICAL TRANSACTIONS. 795 



barrow, and Bodmin Down, have been determined from that of Trevose Head. 

 Of the remaining stations, some are derived from Maker Heights, others from 

 Dunnose : most of them are mean results, that is, each station has generally been 

 found 2 ways ; and as it will serve to show what errors proceed from irregularity of 

 refraction, and imperfection of observation, I shall exhibit a few particulars. 



Height of Deduced from Feet. Mean. 



( Maker Heights 1 169 



Black Down < 1 16*0 



t Carraton Hill 1 152 



C Black Down 609 



St. Stephen's Down X 6*05 



LCarraton Hill 600 



C Bradley Knoll 779 



Westbury Down < 775 



( Beacon Hill 771 



{Mendip Hills -. . 703 

 700 

 Westbury Down 6*96" 



C Mendip Hills 335 



Moor Lynch 2 330 



I Ash Beacon 325 



rDundon Beacon 46* 



Lugshorn Corner 1 49 



LGreylock's Foss-way 52 



{Highclere 1014 

 1011 

 Beacon Hill 1009 



r Bull Barrow 653 



Ash Beacon 1 655 



(.Bradley Knoll 657 



The above will sufficiently show what dependence is to be placed on the heights 

 deduced from observed angles of elevation or depression ; the results are indeed often 

 less consistent, and frequently unsatisfactory ; but generally they run on a parallel 

 with these. The data from which all the heights have been computed, accompany 

 the article. The measurement of the base on Sedgemoor, showed a fall of about 7 

 feet from Lugshorn Corner to Greylock's Foss-way : therefore, supposing that fall 

 to be gradual and constant, all the way from the latter station to the surface of the 

 sea at Bridgewater Bay, we shall get 24 feet, for the height of Lugshorn Corner 

 from the surface of the sea. The altitude of this station, deduced from that of 

 Trevose Head, is 49 feet ; and subtracting 3 feet from it, the height of the bank on 

 which the instrument stood above the moor, we get 4() feet for the height of the moor 

 at Lugshorn Corner, above the level of the sea at Bridgewater Bay. But this 

 height, supposing the fall regular, is proved to be 24 feet. There is therefore a 

 difference of 22 feet, granting the whole of this to be an error on the side of the 

 survey ; but as the general surface of the moor at Bridgewater Bay is several feet 

 above the surface of the sea, we may take a moiety of 24 feet, for the error of the 

 computed height of the station at Lugshorn Corner. 



The refractions contained in this account, like those in the former papers, tend 

 to prove, that when rays of light pass horizontally, and considerably distant from 



5 1 2 



