TERRESTRIAL MAGNETISM. (APP. A.} 323 



shore. The adjustable magnetic deflector before 

 you is designed for carrying out in practice 

 Sabine's method more rapidly and more ac- 

 curately, and for extending it, by aid of 

 Archibald Smith's theory, to the complete de- 

 termination of the compass error, y 

 with the exception of the con- 

 stant term " A" of the Admiralty 

 notation, which in almost ever}* 

 practical case is zero, and can 

 only have a sensible value in 

 virtue of some very marked 

 want of symmetry of the iron- 

 work in the neighbourhood of 

 the compass. 1 When it exists 



1 I had a curious case lately of effect of 

 unsymmetrical iron on a midship steering 

 compass, due to a steam-launch about 30 

 feet long placed fore-and-aft on the port 

 side of the deck with its bow forward and 

 its stern 5 or 6 feet before the thwart-ship 

 line through the position of the compass 

 (Fig. 41). The compass having been ad- 

 justed by the globes and magnetic correctors 

 to correct the quadrantal error (D), and the 

 semi-circular error, it was found (as was 

 expected) that the compass was correct on 

 the east and west points, but showed equal 

 westerly errors of about 3^ on the north 

 and south points. There were, therefore, 

 approximately equal negative values of 

 "A" and "E" each i^. The captain 

 was, of course, warned of the change he will find when he is 

 relieved of the steam-launch at Rangoon, the port of his destina- 



V 2 



