The Declination or Variation is determined by the following very simple process :—Let C T denote the magnetic meridian ; c r, the astro which can be turned horizontally, so that the axis of the magnet can be constrained to assume a position approximately perpendicular to the magnetic meridian. Beneath the magnet, a graduated transparent scale, c, is suspended, and the readings of this scale are viewed by means of a telescope, T, whose axis is in the magnetic meridian. Now the horizontal force, x, evidently acts perpendicularly to the magnet, hence every change that it undergoes will cause the magnet to be deflected from its normal position. These deflections are read off by means of a vertical wire placed in the focus of the telescope ; hence, having determined the change in the horizontal force which corres ponds to one division of the scale, the reading of the scale at any time will show us the increase or decrease of the horizontal intensity—the zero reading of the scale corresponding to the normal value of the horizontal force.
The Vertical Force Magnetometer is another differential instrument for observing A T, a small variation in the vertical component Y. It nomical meridian. Then the axis of a magnetic bar, e C D, suspended by a single thread attached to a point aliore its centre of gravity, will assume a horizontal position in the line C T ; and a graduated glass scale being attached to the magnet, its zero-reading, which is taken to indicate the normal position of the magnet, is made to coincide with a vertical wire in the focus of the fixed telescope T. The angle r C T is the declination or inclination of the magnetic to the astronomical meridian, and is measured by a theodolite which is placed so that its axis is made to coincide with the line C T ; and being then turned in azimuth until a distant object is seen which is known to be in the meridian of the place, the azimuthal angle thus measured is the normal declination.
This instrument, which is sometimes called the Unifilar Magneto meter, likewise serves to make differential observations of the declina tion ; for there is no difficulty in determining the angular value of each division of the scale attached to the magnet: hence, if the vertical wire in the telescope coincides with the zero-reading of the scale at any specified time, when the needle is supposed to be in its normal position, any other reading will denote the angular deviation of the magnet from its original position, and therefore determines the corresponding variation of the magnetic meridian.
Such is a brief outline of the methods employed in determining the absolute value of the magnetic elements, and in watching the periodic changes they are constantly found to undergo. It is obviously beyond the limits of this article to enter into the numerous and elaborate details of the cautions which must be adopted by the observer ; the methods employed to diminish the effects of instrumental errors ; and the corrections which are applied to many of the observations to counteract the influence of temperature, which has the effect of changing the magnetic condition of the bars. For a more complete
description of the magnetometers, and all the practical details necessary for observing and computing the observations, the reader is referred to the Instructions of the Royal Society to the Directors of Maguetical and Meteorological Observatories.' We now proceed to notice some of the most prominent phenomena in connection with terrestrial, magnetism. It has been already stated that the magnetic elements are subject to periodic changes. These changes are generally slow, and exhibit a certain amount of regularity, though the laws which govern them have not yet been arrived at. One of the most apparent, perhaps, of these fluctuations is seen in the declination magnet, that end which is towards the north moving slowly westward during the forenoon, and returning to its normal position about ten in the evening. It then moves towards the east, and returns to its former position about ten in the morning. These changes evi dently establish the fact that the motion of the magnetic meridian is in some way connected with that of the sun. The other magnetic elements also undergo similar though less striking changes; and it is to be observed that each succeeding day will not show exactly the same set of hourly variations as its predecessor, nor are the observations of one year identical with those of another.
But, besides these known periodical fluctuations, which past expe rience teaches us to look for from time to time, there are sudden and unaccountable disturbances in the magnetic elements arising, it is supposed, from some sudden derangement in the magnetic condition of the earth.
This phenomenon exhibits itself in a sudden and sometimes violent agitation of the suspended magnets. During these magnetic storms, as they are called, the magnets oscillate to and fro, sometimes slowly and regularly, at other times with such rapidity that the observer is unable to note the time or arc of the vibration ; and often the scale by which the position of the magnet is viewed is completely carried out of the field of the telescope. On some occasions the magnet seems to be acted upon by a succession of rapid jerks; at other times a quivering motion is detected, so much so as to render the scale indistinct. These unusual disturbances, the particulars of which have, when observed, been carefully recorded at the different observatories, do not appear to have electricity as their origin, since they do not necessarily take place during an electric storm. The most remarkable feature attending them is the fact that simultaneous disturbances of the magnets at different places are found to take place during a display of the aurora borealis [TERRESTRIAL LIGHT], thus connecting that phcnomenon with terrestrial magnetism.