How to be obviated .] 
APPENDIX. 
531 
In some parts of this little discussion upon the attraction of land, I 
feel to have stepped out of my sphere ; but if the hints thrown out should 
aid the philosopher in developing a system of magnetism applicable to the 
whole earth, or even be the means of stimulating inquiry, the digression 
will not have been useless. I conclude this article with copying some pre- 
cautionary memoranda upon the use of the compass in marine surveying ; 
they were made for myself, in case of being hereafter called upon to sail 
in another Investigator, and may not be without their use to other officers. 
1. So soon as the guns are on board and the ship ready for sea, to 
nail small cleats on the binnacle for showing the place where the azimuth 
and surveying compass is to stand, when in use. To ascertain by repeated 
observations whether it be at North and South, or at what other opposite 
directions near them, that this compass gives exactly the same variation ; 
and to note these as the Points of no difference. 
2. Ascertain what the difference in variation is, when the head is 
placed at right angles to the points of no difference, on each side. Half 
this difference is the Error for eight points ; which being divided by the 
dip, will give the common multiplier for that hemisphere, and perhaps for 
both. 
3. To try the accuracy of the common multiplier as often as can con- 
veniently be done, by observations taken at various ports where the dip of 
the needle is different ; and more especially to ascertain whether observa- 
tions in the southern hemisphere give the same multiplier as in the north. 
4. No change to be made in the disposition of the iron work or guns 
during the voyage ; but if a change be indispensable, to ascertain as soon 
after as may be, what alteration it may have produced in the points of no 
difference, and in the multiplier. 
6. The direction of the ship’s head, by compass, to be noted to the 
nearest quarter point when the variation is observed, or bearings of land 
are taken ; this to be considered an indispensable part of such observations, 
since without it the true variation cannot be known, nor the proper allow- 
ance made to the bearings. 
6. On arriving upon the coast to be surveyed, to miss no opportunity 
of observing the variation, by azimuth if possible ; and on passing from one 
side of a projecting cape or island to the other side, to remark ii any differ- 
ence arise in the compass. This is best done by azimuth ; but it may be 
