388 = Theory of the Situation of the Mugnetic Poles, 
I have, I flatter myfelf, cleared the magnetic theory of the 
abftrufe and embarraffing confiderations which have hitherto 
bewildered it, and thereby laid the foundation of a moft 
fimple, general, and ufeful method of determining a fhip’s 
correét place at fea, without any regard to the fun, moon, or 
ftars, to good or bad weather, to the time of the day, the day 
itfelf, months, years, feafons, or centuries, except fo far’as 
may be faid to regard the magnetic inclination, which muft 
always be known from obfervation; as alfo the latitude of 
the fhip, which mutft likewife be known, either from obferv~ 
ation, or dead reckoning. 
- JT would not have it aeiderithods however, that I mean to 
decry the ufe of lunar obfervations or chronometers. Far 
otherwife: I have devoted great application to the ftudy of the 
lunar (and other) aftronomy; and herein I am happy to add, 
IT have fucceeded to an accuracy and concifenefs of equa- 
tions and calculations far furpaffing Mr. Mayer’s, or any 
lunar tables yet offered to the public. But herein, notwith- 
ftanding the high efteem and veneration in which the great 
genius and labours of Euler, and the merits of Mr. Mayer 
muft ever be held by men of fcience, I found it neceffary to 
reject, for reafons which I fhall hereafter tranfmit for infer- 
tion in the Philofophical Magazine, the fuppofition that the 
motion of the fun or earth, according to Mr. Euler, and of 
‘the moon according to Mr. Mayer, are different now to what 
they formerly were. I am alfo the original inventor, and for 
which I have, for upwards of feven years, had a patent, of a 
double fextant for making lunar obfervations to the right and 
Jeft. I have alfo invented a metal quadrant, by which, with 
fore-adjuftments and obfervation, the diftance of the moon 
can be meafured from the fun or other object, from one 
Jimit or extreme point of the horizon to the other, 
From the multitude of lunar places which T have calcu- 
Jated from my tables, I am perfuaded they are fuch as fhall 
not, even at the equator, produce a mean error of five, nor 
an extreme error of fifteen miles in longitude. The firft of © 
the infiruments here fpoken of (a double fextant) meafures 
with convenience, without inverting the inftrument, the 
fimple limits of a fextant on either fide the fun or a fixed far: 
fe) to 
