4 Eir.HITI REPORT — 1S38. 



leys of considerable magnitude. Such a couiitiy would require 

 to be leveled by a series of very short distance^ ; and this cir- 

 cumstance would not only add greatly to the labour and expense 

 o: the operation, but would render doubtful, in a very material 

 degree, the accuracy of the result. It was therefore judged ad- 

 visable to be content with extending the level rorth-eastward 

 to Portishead and Bristol. By this means the east and west- 

 extent of the surface surveyed became nearly equal to the ori- 

 ginal north and south line ; and the level line which rests upon 

 the Qiuuitocks at one extremity, crosses the Mendips and the 

 Leigh Down Hills, connecting a great number of different geo- 

 logical formations. 



7. This line from Bridgewater to Portishead was leveled 

 between May 15, and July 6, 18.3S; and tide observations were 

 made at the latter place in May 18.3/, and July 1838. 



8. Both in the leveling and in the tide observations, every 

 precaution Avas taken to avoid mistakes and to ensure accuracy. 

 As leveling operations of a very delicate kind have rarely been 

 performed, and are nowhere sufficiently described, it is con- 

 sidered worth Awhile to record the raetliod employed in this in- 

 stance, and an Appendix is added containing this description. 

 It may here be observed, that the most important precaution, 

 that of making the distances of the staff from the telescope 

 equal in the fore observation and the back observation, was 

 throughout attended to ; and that all the lines were leveled in 

 both directions, proceeding from the beginning to the end of the 

 line, and then returning back fron\ the end to the beginning. 



9. By employing this method of verification, an apparent 

 error in the process is brought into view, for which it is difficult 

 to account, but which is so constant in its occurrence that we 

 caimot help supposing it to depend on some general cause. The 

 error consists in this ; — that in proceeding \Aith the leveling 

 opei'ation along a line which is really level, the further end con- 

 stantly appears, from the observation, to be the lower end ; and 

 the amount of this depression appears to increase with the di- 

 stance. Hence, when v.e go to the end of a line and then return 

 I0 the starting point, we find the resulting elevation of the point 

 lower than its real elevation. The difference arising from this 

 cause is never considerable, but is always in the same direction, 

 and generally (in the same series of operations) greater in pro- 

 portion as the distance is greater. Thus in the line from Bristol 

 to Portishead (11 miles) it was 1*07 inches; from Bridgewater 

 to Axmouth (40 miles) it was 4*11 inches; from Bridgewater to 

 East Quantockshead (16 miles) it was 1*94 inches; from Bridge- 

 water to Portishead (^9 miles) it was 7'6 inches. 



