tlErORT ON A LEVEL LINE. 9 



iience of a more rigorous nature. The mean level of mean water 

 at one point of the coast of the island, taken for a semilmiation 

 (and probably still more if taken for several lunations), may be 

 asserted to agree with the mean level at another point taken 

 in the same manner, within a very few inches. Perhaps the 

 agreement, if places situate on the open sea were taken, is still 

 nearer; for Portishead, and even Wick rocks, may be affected 

 by the narrowness of the Bi'istol Channel, which may elevate 

 the low water there, as it certainly does in a river. It appears 

 very probable that the level of mean tide at different places on 

 the open coast agrees as nearly as the operation of leveling can 

 determine. 



17. This result is not only very curious in itself, but pregnant 

 with important practical consequences. It is very clear, from the 

 slightest consideration of our results, that nothing but error and 

 confusion can result from processes, such as have often been em- 

 ployed up to the present time, in which heights are determined 

 from " the level of the sea," this level being understood to be 

 that of low water spring tides. Such heights are not measured 

 from a level at all, but from a surface of which some parts are 

 sixteen feet lower than others within the limits of our operations, 

 and probably above twenty feet, if we take the extreme cases on 

 the shores of our island. The only method of stating heights 

 which can have any pretensions to accuracy, is that of reckoning 

 them from a conventional fixed datum upon the solid land ; to 

 which datum the sea as well as the land must be referred by 

 proper leveling operations. 



18. As a specimen of the doubt and confusion which have 

 hitherto prevailed on this subject, I may quote a passage from 

 Mr. Telford's report on the project of a ship canal, intended 

 to connect the Bristol Channel with the English Channel, and 

 following nearly the same course as our level line. He says 

 " the total distance from Beer Harbour [near Axmouth] to 

 Bridgewater Bay [in which are Wick rocks] is forty-four miles 

 five furlongs. The fall from the summit to high water at an or- 

 dinary tide in Bridgewater Bay is 23 1 feet ; but by taking an- 

 other tide at Beer, the fall was found to be 233 feet." 



The vague mode in which this result is expressed, an '^ordinary 

 tide " being taken in Bridgewater Bay, and " another tide " at 

 Beer, without any indication whether any correction was re- 

 quired for the difference of tides, and whether the result could 

 pretend to any accuracy, is, I conceive, an instance of the impos- 

 sibility of referring any elevations to the sea in a satisfactory man- 

 ner, till it is determined how we are to allow, not only for the 

 difference of high and low water, but for the different heights of 



