4 On Terrestrial Magnetism. 



in the Philosophical Transactions, the Miscellanea Ciiriosa, and 

 other works. On his return from this expedition, he combined 

 his own observations with those of others, and from them con^ 

 strutted a chart, with the curve lines of the various degrees of 

 variation of the compass drawn upon it, in directions corre- 

 sponding with those on the surface of tlie earth. The idea of 

 this chart was at that time perfectly new: and men of science 

 were led to imagine that the longitude might he discovered by 

 means of it with suiHcient accuracy for nautical purposes. Were 

 these lines permanent, or the changes which they undergo re- 

 ducible to any calculable laws, it might be determined by them 

 within certain limits; but this is not the case: and although 

 the accompanying paper elucidates several important points, and 

 brings before us the active labours of many ingenious indivi- 

 duals ; yet it has not been successhil in removing all the difficultits 

 that present themselves in its a])plication to the longitude. 



It is well known to philosophers, that there was a time when 

 the laws of the moon's motions were considered so inexplicable 

 as to defy all human attempts to reduce them to any simple 

 principles, capable of being apphcd for predicting her future si- 

 tuation for any given time with exactness: and the same could 

 then be said of the tides, the orbits of comets, and various other 

 particular parts of astronomy and natural philosophy: but by 

 the unwearied diligence and researches of a few profound ma- 

 thematicians, this uncertainty respecting them now no longer 

 exists. There are, imleed, few things of this kind that cannot 

 by degrees be brought to some system. Em]viric modes are 

 first applied of explaining and computing the several motions ; 

 tlien by investigating, comparing, and gradually approximating 

 to the observations, we come at length to causes which rest on 

 established princij)les, and ultimately every apparent anomaly is 

 accounted for, by a reasonable and satisfactory theory. 



Mr. Ralph Walker, who published a treatise on this subject 

 in \704, has computed a table* extending from the equator to 

 the latitude of GO" north, and from the meridian of London to 

 90' of west longitude, in which, by entering with the latitude on 

 the side, and the longitude at the top, the corresponding varia- 

 tion of the compass may be taken out of the area of the table ; 

 or, by means of the variation, and the latitude of the place, the 

 longitude may be obtained f. But such is the imperfect state 



of 



* A siiTiilnr one is tiiveii by Captain Cliristuplier I\]irl(!leton, to suit tlie 

 voyage- lioiii L()n'l(jii t.i Hudion's B:n', iti the Pliilobopliical rnuibactions, 

 No. 393, and iidsiptcd for tlii; year 17ao. 



+ It ih evident, tiiat sucli a table iiiust be usefni in more respects than 

 the one lutiitiuiicd iiere; lor it enableu scaiiicn, wliy know tlieir situation 



