[ l6 7 ] 
From hence we may fafely conclude that the dif- 
ference of longitude between any two places may be 
determined by one fingle obfervation of the contact 
of Mercury with the Sun’s limb, made at each place, 
i'o that the error in the determination will not exceed 
30" of time from the truth: whereas in the above 
33 obfervations of the eclipfes of the firft fatellite of 
Jupiter we find the limit between the errors to a- 
mount to 3' 44" of time. If we take a mean of the 
faid obfervations of the firft fatellite, the difference of 
longitude between Greenwich and Paris is — g' 2 g " , 
and if we rejebt thofe which differ the moft from the 
reft, the mean of the remaining 25 obfervations gives 
the faid difference — g' 40'', and the mean of thofe 
8 obfervations, which are reje&ed, gives the faid dif- 
ference = 8' 53", both which laft determinations can 
be proved to be very far from the truth by the obfer- 
vations of the late tranfit of Venus ; for by the faid 
obfervations of Venus it appears that the difference 
of longitude between Greenwich and Paris cannot 
exceed g' 33", as I faid before; and if the faid dif- 
ference is — 8 / 53 // , then the parallax of the Sun, 
by the Savile-houle obfervation compared with that 
of M. de la Lande at Paris, would amount to 20 " 
which we are fure it cannot be. 
Upon the whole therefore we may cnnclude, that 
the difference of longitude between the royal obfer- 
vatories of Greenwich and Paris (as determined by 
63 obfervations of the contaft of Mercury with the 
Sun’s limb made at each place) is— g’ i6 //- . This de- 
termination would have been perhaps more decifive, 
if I could have had recourfe to the books containing 
the obfervations of the late aftronomer royal, Dr. 
Bradley,. 
