Fisher oti the Variation of the Compass. 83 



In the other ship, the Trent, the line of no deviation was about 

 NNE., or SSW. ; therefore, the magnetic axis made an angle 

 with the ship's axis of about 45°, traversing the ship from larboard 

 bow to starboard quarter. 



In the Dorothea, in which vessel the magnetic axis made an 

 angle of about 67° with the ship's axis, it traversed in an opposite 

 direction from the starboard side before the beam, to the larboard 

 side abaft the beam. 



If now the magnetic axis of the ships, besides their respective 

 inclinations to the axis of the ships, have them inclined upwards, 

 having their south poles nearer the stern of the vessel than their 

 north poles, then can all the observations in both ships be re- 

 conciled ; since the constant inclination observed during the 

 voyage of the north end of the needle in the binnacle compasses 

 towards the head of the ship can be accounted for, from the 

 proximity of the southern magnetic pole of the ship ; and we have 

 the less difficulty in conceiving this to be the case from the 

 curious property iron obtains of becoming magnetical from being 

 insulated in a vertical position, the south pole being constantly 

 uppermost in northern latitudes *. 



The difficulty of applying a correction for the error proceeding 

 from local attraction, must be evident when we consider the con- 

 ditions on which a formula for the purpose must necessarily 

 depend, viz., the magnetic dip and intensity of the magnetic 

 force in different latitudes, conditions which are themselves 

 matters of still greater difficulty. 



I shall now give the observations as they were taken, with the 

 circumstances connected with them, that they may afford data 

 in conjunction with others recently made, for any future inves- 

 tigation on a subject so important to navigation. 



To the observations of Lieutenants Franklin and Beechey, of 

 the Trent, I am particularly indebted for their skill in taking 

 them, and accuracy in registering every connecting circum- 

 stance. I cannot, therefore, do better than insert them, with the 

 remarks made at the time. 



* Since writing the above I find I liave been anticipated in this idea, 



by Captain Scorcsby, jun., iu an excellent paper of his on the same subject, 



G 2 



