608 Mr. Dalby’s Remarks on Gen. Roy’s Account 
difference of longitude of \" .z of a degree, and confequently 
a variation of about 6 7/ in the longitude of Dunkirk or Paris. 
Was the diftance of the ftations about 36 miles, the error 
in longitude would be the fame as that in the horizontal angle, 
or 1 // . 
The length of the arc RG (PI. X. fig. 5.) is *7695 fathoms, 
and its value 1 7' 20" .06 as an arc of a great circle perpendicular 
to the meridian. Now, was the earth a fphere, the length 
of any arc, would be to the number of degrees it contained, as 
17695 to 17" zo /f . 06 ; but this is not accurately the cafe on a 
fpheroid ; though, on this account only, the error in longitude 
(which is in defeat) deduced from an arc of a great circle ob- 
tained in the above manner, muff be fmall to the extent of 3 
or 4 degrees (in the latitudes of the places of obfervation) on 
a fpheroid not more oblate than the earth. 
It may be obferved, that in determining the differences of 
longitude by the pole-ftar obfervations, the ftations fhould be 
as nearly eafc and weft from each other as the nature of the 
country will permit, becaufe in that direction, any errors which 
may be thought to arife from the uncertain inclination of the 
verticals on the fpheroid, will vanifh ; and, what is of more 
confequence, a longer arc of a great circle perpendicular to 
the meridian will thereby be determined than could be in any 
other direction with the fame diftance. On this account the 
ftations at Botley Hill and Hollingborn Hill (for one is feen 
from the other) are eligible. Their diftance is about 28^ 
miles, which would meafure near z\' of a degree of a great 
circle perpendicular to* the meridian. 
P. 220. Art. XI. feems to want correction : for, if Mg is 
a lefler circle parallel to the meridian GR, it will cut the great 
circles rM, Gg, at right angles. Hence, RMC — RMr 
( 1 4 ° 
