TERRESTRIAL MAQNKTISM. 



127 



antarctic regions. The north and south magnetic poles, where 

 a freely suspended needle assumes the vertical position, have 

 thus been located, though not, it appears, actually reached. 

 These poles slowly change their position, in correspondence with 

 the general secular changes of the earth's magnetism ; the 

 motion, however, has not yet been ascertained with precision. 



During the present century a great advance has been made 

 towards the survey of the whole earth, within a comparatively 

 limited period, partly by official action leading to modern surveys 

 in India, America, Egypt, Japan, and elsewhere, and partly by 

 the establishment of the Department of Terrestrial Magnetism 

 of the Carnegie Institution of Washington. The latter body has 

 not only surveyed many land regions where magnetic observations 

 were scanty, but has also made a remarkably extensive oceanic 

 survey embracing the whole globe, during the years 1910-1917. 

 For this purpose a specially non-magnetic ship, the " Carnegie/' 

 has been constructed, whereby magnetic determinations at sea 

 have been rendered far more accurate and speedy than is possible 

 in ordinary ships. The work of this institution has materially 

 improved the accuracy of recent magnetic charts of the world. 



The magnetic maps so far referred to, while they are in the 

 form most useful for navigation, do not give a good picture of 

 the actual direction of the earth's lines of magnetic force. As 

 regards the direction of horizontal force, which is what the com- 

 pass shows, maps such as Duperrey's (as on the lantern slide), 

 which give the lines of force themselves, afiord the best idea. 

 They converge to the magnetic poles in the two hemispheres, 

 and their direction differs most from that of the meridians in 

 polar, regions. 



Why the earth is a magnet, and why its magnetism changes 

 from year to year, are mysteries which have as yet hardly at all 

 yielded to attempts at explanation. One fact, however, con- 

 cerning the seat of its magnetism, has been surely established. 

 All except possibly a small fraction of the whole is known to have 

 its origin tvithin the earth, and not in the atmosphere, nor in 

 extra-atmospheric regions. It has been conjectured, with some 

 probability, that the rotation of the earth is the prime cause of 

 its magnetization, and that any large rotating body would 

 probably be likewise magnetic. The fact that the sun also 

 appears to be a magnet, and one having its magnetization, on 

 the whole, similarly related to the axis and direction of rotation, 

 tends to support this theory of the origin of the earth's 



