783. The action of terrestrial magnetism does not reside in the iron formation alone, but in its antagonism to the reduced ores. The terrestrial magnetism is a tension between iron and silver. The lead imparts the poles. 784. As the earth is a globe, so are the two masses of iron situated nearer to its axis than the masses of silver and lead. If therefore the latter exert also any influence upon the direction of the magnetic needle, yet nevertheless the poles of magnetism must coincide nearly with the poles of the earth's axis, because the exciters of magnetism operate in this direction. 785. The magnetic meridian runs indeed in general from pole to pole, but as it is not the earth's axis, which determines the magnetism, but the metallic masses, so the direction of the meridian deviates from the earth's axis in accordance with these masses. 786. The magnetic needle can therefore assume a different direction on every part of the earth, according as its relative position is varied between two principal masses of metals, and even as is indicated by the movement of the compass with the change of geographical longitudes, it is removed from one metallic mass, approaches another, and by this becomes more powerfully attracted. Upon the whole, however, the direction must tend toward the poles. Such are the phenomena of _declination_ of the magnetic needle; and we can now comprehend why this is present. 787. But there must be also spots upon the earth, where the needle points straight to the north, probably, when it is between two metallic masses, or stands at a certain angle to them. These are the _lines without declination_, of which there are many as is well known, but which have not as yet been reduced to any law. Nor can they ever be so, because we shall never become acquainted with the metallic beds.