102 ADDITIONAL OBSERVATIONS OF THE 



October 7th, 1839, afford presumptive evidence of this. In order to obviate all 

 danger of mistake in reversing the poles, a small dot of ink was placed upon 

 each needle, and the position of this mark, whether up or down, was always en- 

 tered at the head of the observations. The following observations are arranged 

 in the order of their dates: — 



Magnetic Dip at Hudson, Ohio, Latitude 41° 15' N.; Longitude 5h. 26m. W. 

 Place of observation the same as formerly described. 





Date. 



Hour. 



Needle. 



No 



Readings. 





Dip. 



339 



, Aug. 16th, 



7—9, A. M. 



No. 1, 





40 



72^ 



54'.4 



a 







No. 1, poles reversed, 





40 





42.9 









Mean of No. 1, 





80 





48 .6 









No. 2. 





40 





22 .2 









No. 2, poles reversed, 





40 





69.6 









Mean of No. 2, 





80 





45.9 









Mean of two needles. 





160 



72 



47.3 



The mean of the observations with the two needles is nearly the same, yet 

 the observations with No. 2, in its two magnetic states, differs by 47', indi- 

 cating an inequality in the arms of the needle. If the needle could always be 

 magnetised with the same intensity, the effect of this inequality would be elimi- 

 nated by reversing the poles; yet, as this condition is not easily fulfilled, it is 

 better to make the arms of the needle of equal weight. The heaviest end of 

 No. 2 was therefore rubbed on a hone, and the observations repeated. 



Date. Hoar. Needle. No. Readings. Dip. 



1839, Aug. 18th, 8—9, A. M. , No. 2, 40 72° 36'. 8 



" " " " " No. 2, poles reversed, 40 51 .9 



" " " " " Mean of No. 2, 80 72 44.4 



Here the inequality is reduced to 15'. The heavy end was again applied to 

 the hone, and the following observations made : — 



Date. 



Hour. 



Needle. 



No. Readings. 



Dip. 



1839, Aug. 19th, 



9—10, A.M. 



No. 2, 



40 



72° 51'.9 



a a a 



a (( 



No. 2, poles reversed, 



40 



57.7 



(i (( (( 



[( a 



Mean of No. 2, 



80 



72 54.8 



The inequality is here reduced to 5'. 8, and no farther change was made in 

 the needle during the whole of the subsequent observations. The mean of the 



