C 84 ] 
•lame divifion as before ; yet a mean of the nvo will be the 
ti'uthj as nearly, I believe^ as it is poffible for any inftru* 
ment to give it. 
Quere i ft. May not a part of this fmall difference be 
attributed to the direction of the magnetic influence 
(whatever that be) in the fteel bar? and if fuch an expe- 
riment could be tried upon the prefent azimuth com- 
paffes, is it not probable, that the variation in them would 
be at leaft as fenfible? Quere ad. May not this be the 
caufe that two of the beft of them will differ a fmall mat- 
ter from one another ? Quere 3d. Would the ends of the 
needle being made thus 1 — > , inftead of the fquare form 
be, in Ibme meafure, a remedy for this fmall variation? 
I am, moft^efpecftfully, sir, 
^ Your moft obedient humble fervant, 
J. LORIMER> 
Vm. 5/7/ 
