115 



minimum of the dips. Most commonly I had 

 to judge of the magnitude of the angle by tak- 

 ing, among a great number of very small oscil- 

 lations, the mean of the elongations toward the 

 north and the south. I constantly employed 

 two different needles ; their centring was veri- 

 fied by comparing the indications of the two ex- 

 tremities of the same needle, and inverting it, 

 or successively directing the divided face of the 

 limb east and west. I think I may be certain, 

 that observations made when under sail may at- 

 tain an average exactness of twelve minutes of 

 the centesimal division *. 



* The angles given by Borda's dipping compass are ex- 

 pressed in centesimal degrees and decimal parts. The veri- 

 fications of the instrument, which can be made on land, and 

 which I have constantly employed with Mr. Gay-Lussac, 

 during the course of the observations published in the Me- 

 moirs of the Society of Arcueil, are reducible to the following : 

 1st, giving a horizontal position to the azimuth circle, by 

 means of a bubble level, and a thread level j 2ndly, finding 

 the direction of the magnetic meridian, either a) by corre- 

 spondent dips, or b) by adding, on the azimuth circle, one 

 hundred degrees to the point which corresponds to the per- 

 pendicular position of the needle ; or finally c) by the minimum 

 of the dips ; 3dly, correcting the eccentricity of the needle, 

 by observing the superior and inferior points ; 4thly, examin- 

 ing whether the magnetic axis of the needle coincides with 

 it's physical axis, by observations to the east and the west ; 

 5thly, correcting the want of equilibrium in the needle by 

 changing the poles. The slight differences, which will be no- 

 ticed between the results published in this narrative, and 



