836 



terminations, by observers of different nations, all comprised between 

 the years 1828 and 1848 ; each determination being in the majority 

 of cases a mean of several distinct and independent observations, and 

 all reduced to the epoch of 1840, by the rates of secular change de- 

 rived from a comparison of Hansteen's map of 1787 with the present 

 state of the phenomena. 



A considerable portion of the determinations thus co-ordinated 

 having been obtained on board ships in which measures were taken 

 to supply the means of correcting the errors occasioned by the in- 

 fluence of the iron which the ships contained, the author in an ac- 

 companying memoir has discussed at some length the variable part 

 of the corrections required for that purpose, being that portion of 

 the correction which varies as a ship changes her geographical po- 

 sition. He infers from the observations generally, and especially from 

 those of H.M.S. Erebus during the Antarctic Expedition in 1839- 

 1843, that the disturbance occasioned by the iron is chiefly, if not 

 entirely, due in wooden ships to the induced magnetism of the iron ; 

 but that its changes are not so rapid as those of the terrestrial mag- 

 netic Inclination and Force during changes of geographical position, 

 and that the magnetism of the ship is consequently at such times 

 liable to become more or less in arrear, if the expression may be 

 permitted, of the change which the magnetic inclination and force 

 have undergone ; that in fact there are considerable portions of a 

 ship's iron which are not permanently magnetic on the one hand, 

 nor perfectly soft so as to undergo instantaneous change with changes 

 of inclination and force on the other hand ; and which derive mag- 

 netism by induction from the earth, but conform gradually rather 

 than instantaneously to the changes of terrestrial magnetism corre- 

 sponding to changes in the ship's place. 



The bearing which this new view of a ship's magnetism would have 

 on the endeavour to counteract the disturbance occasioned by the 

 iron, either by permanent magnets or by soft iron, is noticed, and the 

 conclusion is arrived at, that in order to correct magnetic observations 

 made on board ship, it is desirable to determine experimentally the 

 variable term in the correction formulae, at intervals of not many days 

 apart when a ship is changing her geographical position. An apparatus 

 and method of observation to accomplish this purpose are described, 

 which are stated to be extremely simple, to require no unusual cir- 

 cumstances of weather and no reference to celestial objects, and to 

 occupy but a very fev\^ minutes. 



In conclusion, a comparison is instituted between the declinations 

 computed by M. Gauss's general theory and those now derived from 

 observation over a field of considerable extent ; and on this compa- 

 rison the author founds the following remark, "that the general 

 theory will require to have its numerical coefficients reconstructed 

 before it can become available for practical purposes ; and that those 

 who desire to take a correct view of the magnetic phenomena, must 

 for the present at least, have recourse to the maps constructed di- 

 rectly from the observations themselves." 



