Page:Transactions of the Geological Society, 1st series, vol. 3.djvu/340

 by repeated observations on its steadiness at different elevations above the surface, and at different points upon it. It is not a sufficient trial that detached specimens of the rocks in the vicinity exert no action on it, since the variations will often be sensible from the joint action of a considerable mass of polar matter, when small pieces of the same substance produce no effect. The dipping needle would give the most correct information respecting the existence of these local causes of error, but it is an instrument unfortunately too nice and expensive for ordinary purposes. In all cases where such a disturbing force is suspected to be in existence, and where accuracy is necessary, we should not be content until the ground has been examined and the actual variation ascertained by the observation of the magnetic azimuth, wherever it is practicable to procure such an observation. It would however tend still more to the removal of all possible errors arising from this cause, if surveyors were to reject the use of the needle altogether, and depend solely on the back angle; since although every one who lays claim to accuracy will correct the one observation by the other, yet the temptation arising from the facility of using the needle alone, is perhaps too great to be always resisted.