616 ADDITIONS. 



the agreement, as lie says, far beyond his hopes. They show, he says, 

 that the Theory comes near to the Truth. 



Correction of Ships' Compasses. 



The magnetic needle had become of importance when it was found 

 that it always pointed to the North. Since that time the history of 

 magnetism has had its events reflected in the history of navigation. 

 The change of the declination arising from a change of place terrified 

 the companions of Columbus. The determination of the laws of this 

 chano-e was the object of the voyage of Halley; and has been pursued 

 with the utmost energy in the Arctic and Antarctic regions by naviga- 

 tors up to the present time. Probably the dependence of the magnetic 

 declination upon place is now known well enough for the purposes of 

 navigation. But a new source of difficulty has in the meantime come 

 into view ; the effect of the iron in the ship upon the Compass. And 

 this has gone on increasing as guns, cables, stays, knees, have been 

 made of iron ; then steam-engines with funnels, wheels, and screws, 

 have been added ; and finally the whole ship has been made of iron. 

 How can the compass be trusted in such cases ? 



I have already said in the history that Mr. Barlow proposed to cor 

 rect the error of the compass by placing near to the compass an iron 

 plate, which from its proximity to the compass might counterbalance 

 magnetically the whole effect of the ship's iron upon the compass. 

 This correction was not effectual, because the magnetic forces of the 

 plate and of the ship do not change their direction and value accord- 

 ing to the same law, with the change of position. I have further 

 stated that Mr. Airy devised other means of correcting the error. I 

 may add a few words on the subject ; for the subject has been further 

 examined by Mr. Airy* and by others. 



It appears, by mathematical reasoning, that the magnetic effect of 

 the iron in a ship may be regarded as producing two kinds of deviation 

 which are added together; a "polar-magnet deviation," which 

 changes from positive to negative as the direction of the ship's keel, in 

 a horizontal revolution, passes from semicircle to semicircle ; and a 

 " quadrantal deviation," which changes from positive to negative as the 

 keel turns from quadrant to quadrant. The latter deviation may be 

 remedied completely by a mass of unmagnetized iron placed on a level 



8 Phil. Trans. 1856. 



