[ I 3 I ] 
to be 36° 2' 4 ; from whence the latit. of the fhip is 
59 0 n 7 \ and that of the fa&ory 58° 55' 
The prodigious difference between the latitude of 
Churchill fa&ory, as laid down from obfervations 
made by Hadley’s quadrant, and that deduced from 
the obfervations made with our agronomical quadrant 
on fhore, has often employed my mofl ferious at- 
tention ; but I cannot think on any probable caufe 
for fuch difference, unlefs it lie in the very great re- 
fradive power of the air in thefe parts. I have men- 
tioned how the ice and land appear to be lifted up,, 
when we (land on the fhip’s deck : and if the vifible 
horizon be lifted up in like manner, it muff make its 
apparent diftance from the fun, or, which is the fame 
thing, the fun’s apparent altitude lefs than it other- 
wife would be,; and confequently, the latitude greater 
than the truth; and alfo greater than it will be fhewn 
by a land quadrant, which depends not on the hori- 
zon, agreeable to what we find it in the cafe be- 
fore us 
■ ■ • Having mentioned this circumftance to the reverend Mr.. 
Mafkelyne, it immediately occurred to him, that the longitude- 
deduced from obfervations of the j) ’s diftance from the fun or a 
ftar, would be conftderably affected by this caufe, as not only 
the altitudes of the O, from whence the time at the fhip is 
found ; but alfo the latitude of the fhip, found by an obferva-- 
titin of the fun’s meridional altitude, or otherwife, will confpire 
to encreafe the fun’s diftance from the meridian, or angle at the. 
pole. 
I have therefore recomputed the longitude from my observa- 
tion of the moon’s diftance from the fun, taken Auguft the ^tb, 
176?, on a fuppofition that the mean error in any altitude taken 
by Hadley’s quadrant, arifing. from this caufe, is 10 minutes;, 
and find that, on fuch a fuppofition, which it muft be allowed 
appear to be extremely well' founded, the longitude will be ii' J 
S 2 ' ? Sen*- 
V. . 
