TERRESTRIAL MAGNETISM. (APP. /?.) 333 



seen accurately shown on the divided circle of the 

 compass-card. This is a very valuable quality for 

 use in rough weather at sea, or when there are 

 flying clouds which just allow a glimpse of the 

 object, whether sun or star, to be caught, without 

 .allowing time to perform any adjustment, such as 

 that needed in the old Admiralty azimuth compass 

 to bring the hair, or rather the estimated middle 

 of the space traversed by the hair in the rolling of 

 the ship, to coincide with the object. The same 

 degree of error as on the horizon, but in the 

 opposite direction, is produced by imperfect ad- 

 justment in taking the bearing of an object at an 

 elevation of 38. 



Thus for objects from the horizon up to 38 of 

 altitude the error in the bearing is less than 12 per 

 cent, of the error of the setting. For objects at a 

 higher elevation than 38 the error rapidly in- 

 creases ; but even at 60 altitude the error on the 

 bearing is a little less than half the error of the 

 setting ; and it is always easy, if desired, to make 

 the error of the setting less than a degree by 

 turning the instrument so that the marker which 

 you see below the lens, shall point within a degree 

 of the position marked on the circle of the compass- 

 card by the image of the object. 



For taking star azimuths the azimuth mirror 

 has the great advantage over the prism compass, 

 wit /i its then invisible hair, that the image of the 

 object is thrown directly on the illuminated scale 



