In the ordinary operation of maritime surveying by the needle and sights, the errors are often likely to be considerable; more particularly where some of the angles are determined with the instrument placed on shore, mariners having been uninformed with respect to the existence of such disturbing forces. To such causes it is very probable are owing the discrepancies found in charts of new or distant coasts, where the instruments have even been good and the observer practically expert in its use. It is even probable that the charts of the west coast of Scotland, constructed by Mackenzie with abundant leisure and experience in his art, owe many of their very glaring errors to this cause, which on many of these coasts is exceedingly active. Want of care or want of accuracy, as far as this could be obtained by the means which he possessed, cannot be imputed to him, as the soundings, distances, and enumeration of even the most minute sunk rocks, is surprisingly accurate, rendering his work such, that even the most inexperienced pilot may enter these intricate harbours and come to his anchorages with no other guide, as I have abundantly verified on numerous occasions. If, as has been said, the well known map of North Wales constructed by Evans, was also surveyed by the compass, we may perhaps attribute a portion of the errors which it comprises to this neglected cause among others. It is true that the repeated comparison of bearings will afford some correction to inaccuracies of this nature, but a medium of bad observations can never produce a good one.
At any rate where such an instrument is to be used it will always be prudent to erect it on a staff at some distance from the ground, as we shall then at least avoid those disturbances which result from the contact, or very near approach of rocks when it is placed on or near the ground. This addition is easy and of little encumbrance to the operator, and as the compass and
