370 Mr. J. R. Napiee on Ships' Compasses. 



produces oscillations, — the more violent according to the power of' the 

 needles. 



This I believe to be the case. The remedy for this appears to me to 

 be very ingenious. The card traverses on an axis fixed to the top of a 

 brass bell, which bell is placed on an ordinary pivot. 



The weight which prevents the needle from dipping is here transferred 

 to the bell, and the rotation or oscillation of the bell, he says, causes little 

 or no oscillation of the needle itself, except what is communicated to it 

 from friction against the axis and pivot. I do not know any simple 

 method of testing the correctness of Captain Walker "s views, but the re- 

 ports of the Admiralty, and other trials of the compass published in 

 Captain Walker's work, and also in the late Captain Johnstone's, prove it 

 to be at least equal to the best. 



While reading Dr. Scoresby's letter addressed to the Liverpool Under- 

 writers' Association, on the subject of the Compasses of Iron Ships, I was 

 startled by the remarks regarding the loss of the " Tayleur," when, from the 

 evidence, he says it appeared that the compasses were all correct on leaving 

 Liverpool, and all wrong after getting to sea. This recalled to my mind a 

 singular change in the compass deviations of the iron brig " Haiti;" the 

 change, however, was fortunately discovered before the vessel put to sea. 

 She was built on the Clyde, near Glasgow, with her head about south-west, 

 was swung as an experiment at the Broomielaw and the deviations noted, 

 then was towed to Goiirock Bay in order to be again swung and to have 

 the errors reduced by soft iron correctors, but it blew a gale, which pre- 

 vented the eiTors from being observed till the second day. 



The deviations shown by the curve were then observed and noted, the 

 results, as will be observed, differ considerably from those taken at Glas- 

 gow, (vide diagram, brig " Haiti.") 



Though the observations at Gourock are evidently not so correct as 

 could be wished, the curves passed through those that were thought to 

 be the most so, show a great and decided difference, whatever may have 

 been the cause of this change of the deviations, a compass placed up one 

 of the masts, or on a mast, or staff of its own, would in all probability 

 have avoided it ; for, in the first place, there would have been little 

 or no error in a compass placed at a proper height. Such changes, 

 however, if arising, as appears to be very probable, if not certain, from 

 the causes mentioned by Dr. Scoresby in his letter to the Liverpool 

 Underwriters, viz., from the soft ii'on when hammered in a given 

 direction, becoming highly penetrated with retentive magnetism, and so 

 retaining its magnetism when the iron is turned quietly in any other 

 direction, but being liable to change its character when vibrated in new 

 positions, such change seems more likely to happen at Liverpool than 

 on the Clyde, where vessels are fitted out in dock, and where the com- 



