- REPORT ON A LEVEL LINE. •> 



10. It is very difficult to explain the cause from which this 

 seeming error arises, or even to conceive any cause from which 

 it can arise. The errors arising from the curvature of the earth, 

 and from any permanent refraction, are eliminated by the con- 

 dition of equal distances in the fore and back observations. The 

 difference does not seem to arise from the effect of the sun's rays 

 on the instrument, for it is not removed by shading theinstrument 

 Avith white paper ; nor fi'om any rise of the peg between the fore 

 and back observation, for it is not confined to soft ground. It 

 appears to go on increasing with the time during which the ob- 

 servations are continued, and is such an error as would result, 

 if we suppose that in every interval of time between the back and 

 the fore observation, something takes place by which the staff 

 is apparently (by refraction or otherwise) less elevated (or more 

 depressed) at the fore observation than it had been at the pre- 

 ceding back observation. For these elevations are supposed to 

 be equal in the process ; and if the elevation of the fore point by 

 refraction or any other cause be the smaller, the point will ap- 

 pear to be lower when it is really on the same level. This state- 

 ment, however, is made rather with a view of explaining the 

 nature of this error than of assigning its cause. 



1 1 . But since it is thus probable that this apparent error arises 

 from some constant and general cause, it is clear that we shall 

 get rid of its effects in each case by taking the mean of the first 

 and last results. We may therefore suppose the mean difi:erence 

 of levels obtained by leveling between two points, first in one 

 direction and then in the other, to be accurate within limits very 

 much smaller than the errors above mentioned. We may venture 

 to confide in this result to a fraction of an inch. 



12. The relative heights of the parts of the lines surveyed 

 being determined by the operations of which we have been speak- 

 ing, marks were fixed at various points, by means of which the 

 position of the line now measured may hereafter be again dis- 

 covered. These mai'ks are the following. A place was selected 

 in the solid rock on the shore just below tlie fort at Portishead ; 

 and in this was inserted horizontally a cylinder of iron, two 

 inches diameter and fifteen inches long, containing in its centre 

 a brass wire one eighth of an inch in diameter, which marks 

 the position of the standard point, about eight feet above the 

 highest high Avater. This mark is on the property of James 

 Adam Gordon, Esq., of Naish House, who kindly gave per- 

 mission for its being placed there. Tiie mark at East Quan- 

 tockshead is on a farm called Perry Farm, the property of J. 

 F. Luttrell, Esq., of Dunster Castle. It consists in a block 

 of granite, a ton and half weight (the gift of the corporation of 



