made at Beverly. | 319 
good of its kind, and is furnifhed with a vernier, pointing out the 
azimuth to 5’; but the eye may pretty eafily determine. by it 
to 2’, and fometimes to 1°. To obferve by this compafs, I af- 
certained the going of my clock to great exactnefs, and on five 
different days, took feveral magnetical azimuths, both before 
and after the fun paffed the meridian, and noted the moments, 
which I have put down in apparent time. For thefe times, I 
have calculated the true azimuths by fpheric trigonometry, and 
have carried out the variation for each obfervation feparately. 
On two days, I alfo determined the variation, by taking mag- 
netic azimuths, at correfponding altitudes of the fun, making 
proper allowances for the change of declination, between the . 
obfervations of the forenoon and afternoon. On each of the - 
days, fome of the obfervations differ feveral minutes from others ; 
but this I cannot attribute to want of attention, as Lam con~ 
fcious that I made them with all the care in my power. The 
differences, I fufpect, principally arofe from the difficulty of de- 
termining, with entire exactnefs, when the fhadow from the 
hair was on the line beneath ; and when in two obfervations, 
the error. fhould be on different fides, the fam might make a 
number of minutes. But T have the fatisfaction of finding the 
mean refülts for the feveral days well agreeing with each other, 
which is-a good evidence that the refult of the whole muft be, 
at leaft, _very near the truth. 
Thefe obfervations and deductions are now humbly fubmit- 
ted’to the Academy, with withes that aid may fubferve the 
defigned nd 
Jury. 
