Gilbert in 1600 was the first to announce the bold hypothesis that the earth is a great magnet, and that the needle assumes a N and S. direction because it is attracted by the dissimilar and repelled by the similar poles of the terrestrial sphere. He illustrated this hypothesis by magnetizing small globes of steel; but this illustration, though it served in a general way to represent the phenomena, is not strictly correct In the first place, the magnetism of the earth is not symmetrical like that of a steel magnet, but is to a considerable degree irregular; and secondly, it is not permanent, hut subject within certain limits to almost continual changes both in di-rection and intensity. Indeed, the magnetic needle is scarcely ever absolutely stationary from one moment to another, hut is constantly exhibiting minute variations. If the earth is a magnet, the free needle at any place should assume a definite direction; hut it does not follow from the hypothesis that this direction must be the true north and south, since the magnetic poles of the earth do not necessarily coincide with its geographical poles.

If the two poles be in the same meridian with a given place. the needle will at that place point to the true north; but if the magnetic pole lie either W. or E. of the meridian of a given place, the N. end of the needle will deviate either E. or W. of the true north, and the phenomenon of the declination or variation of the compass will be exhibited. That the needle does not point to the true north had long been known, and it was observed by Columbus in his first voyage of discovery that the direction of the needle is not the same for all portions of the earth. Thousands of observations have since been made to obtain the data for constructing charts to represent for the use of the mariner the declination in various parts of the earth.

Again. if we assume that the earth is a great magnet, it will follow that in passing from the magnetic equator, the needle which is accurately balanced, so as to settle horizontally at the former place, will incline or dip as we ad-vance to either pole. That this is really the fact was first discovered by Robert Norman in 1576. Furthermore, if the earth is a magnet, we should expect that the magnetic intensity or the strength of the action would not be the same at all points of its surface, and this inference has also been found to be true. By counting the vibrations of a delicate dipping needle, we find that the strength of the magnetism of the globe increases as we go from the equator toward the pole. The magnetic intensity, however, exhibited by observations of this kind, does not indicate as rapid an increase of force as we approach the magnetic pole as might be expected from such a distribution of magnetism as would result from a magnetized sphere of iron. In conformity with the three magnetic elements we have mentioned, namely the variation, the dip, and the intensity, it is customary to represent the magnetic condition of the earth at a given time by three systems of lines supposed to be drawn on the surface of the globe.

These are as follows: 1, the line drawn through all places where the needle points to the true north or south, to 5° W., to 5 E. l0° W. and 10° E., and so on, called the isogenic lines, or lines of equal variation or de-clination; 2, lines nearly at right angles to the former, drawn through all places exhibiting the same angle of dip of the needle, called isoclinal lines; and 3, a system of lines joining all places having the same magnetic intensity, and consequently known by the name of iso-dynamic lines. It is a problem of much practical importance in regard to the art of navigation, as well as to the study of the phenomena of terrestrial magnetism, that these three systems of lines should be accurately determined; and accordingly expeditions have been fitted out by different nations almost expressly for this purpose. All the observations, however, which have been made in regard to them, indicate the fact that they are not permanent, but are constantly undergoing a change, of which the law is exceedingly complex. Hal-ley's chart of declination for 1800 is very different from that of Barlow for 1833; and Han-steen's dip chart for 1780 does not represent the isoclinal lines of the present day.

The great practical object then of investigation in this branch of science is to discover the law of these changes, in order that, the position and form of these lines being determined for a given epoch, they may be calculated for any future time. The phenomena were first referred to a very small magnet at the centre of the earth, the direction of which is subject to irregular changes. Tobias Mayer, instead of supposing a magnet to be placed at the centre of the earth, conceived one to be situated at about the seventh part of the earth's radius from the centre, and from this hypothesis he was enabled to calculate the variation and clip in places not far distant from those in which these quantities had been determined by actual observation. Hansteen of Norway, who collected an immense number of observations, endeavored to represent the phenomena by the hypothesis of two small eccentric magnets of unequal strength placed at the centre of the earth, giving rise to four magnetic poles, two in each hemisphere. In order to represent the variations of the needle, the poles of each of these two magnets were supposed to perform a revolution around an intermediate line, with different velocities.

Gauss of Gottingen, however, made the first rigid investigation of the problem in accordance with a definite plan, He founded his research on the assumption that the terrestrial magnetic force, or that which is exerted on a needle freely suspended by its centre of gravity, is the resultant action of all the magnetized particles of the earth's mass. According to this assumption, the governing power which affects the needle is due to the magnetism of the earth itself, while the different perturbations to which the needle is subjected are the results of extraneous forces.

To give clearness of perception, he represents magnetization as consisting in the separation of two magnetic fluids, giving magnetic polarity to each particle, or in other words in a repulsive and attractive force acting inversely as the square of the distance. No change would be produced in the result by adopting the hypothesis of Ampere, in which magnetism is held to consist of constant magnetic currents; nor would there be any difference if terrestrial magnetism were ascribed to a mixed origin, as consisting partly of actual electrical currents and partly of permanently magnetized masses. Starting from these assumptions, Gauss obtained a general mathematical expression for the action of the whole globe on a magnetic needle, however irregular might be the distribution of the magnetism of the former. In other words, he obtained an expression by which, if the distribution of the magnetism of the earth were known, and the intensity of its action ascertained with reference to a unit of distance and intensity, the position of the needle and the magnetic force by which it was acted upon at any point could be determined; and conversely, if the action of the earth on the needle were known for a large number of places on the surface of the earth, the distribution of the magnetism might be considered the unknown quantity, and might be approximately found from the data thus afforded by observation.

In this way Gauss was enabled to give a method of constructing general charts to represent in every part of the earth the magnetic declination, inclination, and isodynamic lines, the intensity and direction of the magnetic force being known at a given number of places. The data necessary for improved charts of this kind have been furnished by the magnetic surveys made in various parts of the world in recent times, at the suggestion and principally under the direction of the British association. By repeating the construction of such charts for different epochs, the secular changes in different parts of the earth will become known; and it is hoped that, in due time, if the system of magnetic observations which has been established should be continued, the law of the changes will ultimately be fully ascertained. The investigations of Gauss have shown that the hypothesis of two movable magnets at the centre of the earth does not explain the phenomena of terrestrial magnetism. He defines a magnetic pole to be the place at which the needle points directly downward, or at which the dip is 90°. Indeed, he has pointed out the very obvious fact, that if there be two such points in the northern hemisphere, then there must be somewhere between the two a third point at which the needle would also assume the vertical position.

Gauss, however, arrives at the remarkable conclusion that the place of greatest magnetic intensity does not coincide with that which is usually denominated the pole; and it would appear that there may be a diffused space in the northern hemisphere around which the isodynamic lines may be drawn, representing apparently at least two centres of greater magnetic attraction. These phenomena are best represented by the hypothesis of magnetism due to currents of electricity in the earth, but as yet no definite hypothesis has been advanced as to the nature of such currents. It is true, they have been referred to thermo-electricity; but how the varying heat of the sun or the high temperature of the interior can give rise to currents constantly circulating round the earth, of such intensity and such flexures as would account for the observed direction and intensity of terrestrial magnetism, has not yet even approximately been made out. - What we have said in regard to the magnetism of the earth principally relates to its state at a particular time.

We shall now briefly give an account of the discoveries which have been made in regard to the changes to which terrestrial magnetism is subject; and for the data from which these have been deduced science is indebted to the several magnetic observatories established in different parts of the earth. These are furnished with improved instruments, which in their present perfect state constantly record, by means of photography, the minutest changes in intensity and direction of the magnetic force. The magnetic perturbations were at first sup-posed to consist of two classes, namely, periodical and fitful. Many perturbations, however, which had been regarded as fitful are now known to recur at regular periods, and are therefore not properly designated by this term. The changes of terrestrial magnetism are of three classes. The first consists in a movement of the magnetic poles, around the true poles of the earth, from E. to W. in both hemispheres. This motion is inferred from the secular changes which have been found to affect the position of the magnetic lines, as well as from the secular changes in the position of the magnetic needle at any given station.

The magnetic lines at any given epoch present great irregularity of shape, because very slight differences of magnetic declination, due to local peculiarities, may largely affect the position of the magnetic lines. But when the changes of declination at any given station are considered, they are found to correspond, at least during the period within which systematic observations have been made, to an oscillation such as would result from the motion of the magnetic poles around the true poles of the earth in a period of between six and seven centuries. Thus in 1576 the declination needle in London pointed 11° 15' E.; in 1657 or thereabouts the needle pointed due N; in 1760 it pointed W. by 19° 30'. The westerly declination attained its maximum in 1819, when it amounted to 242/3°. Since then the needle has been slowly travelling east ward, the present annual rate of decrease be ing more than 8'. The mean westerly decli nation for the year 1873 was 19° 30'. Again, in Paris, which lies 2° 20' E. of London, the needle pointed due N. in 1663. Its subsequent motion have closely resembled those of the London needle; but the Paris noodle ceased to move westward as early as 1817, and attained a maximum declination of only 22 1/2°. Now if we combine these facts with the changes of the inclination, we see at once that they point to a movement of the northern magnetic pole from a position between London and the N. polo in the middle of the 17th century to its present position in the extreme north of the American continent (or rather in the archipelago which lies beyond those parts northward). For in the middle of the 17th century the needle pointed northward, while afterward it pointed westward.

Then the magnetic pole lav at that time either directly beyond the N. pole of the earth, or somewhere on (or near) the arc joining London and the X. pole. But if the magnetic polo had lain beyond the true polo, the inclination would have been much less than that corresponding to a magnetic pole at the true polo of the earth, that is, loss than 51 1/20 Instead of this however, the inclination was much greater. Moreover, the incli-nation, which would then have been at a mini-muni had the magnetic pole been beyond the true pole, appears to have then been at a maximum. For though exact observations of the inclination have not hoen made during so many observations of the declination, we find that in 1720 the inclination was 74° 42' in London; in 1800, 70° 35'; in 1865, 68° 9'; in 1870, 67 55'; and in 1873, 67° 45'. The northern magnetic pole was therefore between London and the N. pole of the earth in the middle of the 17th century, and has since travelled westward, or in a direction from E. to W. around the true pole. If we assume the motion to he uniform (which is probably not the case), and that the needle at Greenwich responds uniformly to such motion (which is certainly not the case), we may calculate the period of polar revolution.

Thus, taking the magnetic pole as due N. in 1657, and in 1833, according to Ross's observations, as 95° W. of Greenwich, we have for the period of revolution 360/95 (18331657) years = 667 years about. Combining Ross's estimate with the Paris epoch, we get a period of '360/95'(1833 - 1663) years = 644 years about We may take 050 years as a not improbable period of revolution. It may be added, as confirming the above, that in Russia the magnetic inclination has now reached a minimum, while in Peking it is in-creasing. The cause of this change is at pres-ent entirely unknown; it has no analogy with any other class of physical phenomena with which we are acquainted. By a rough comparison of the isothermal lines and thelines of equal magnetic intensity, a general similarity has been observed, and hence the two have been considered as referable to the same causebut it will be perceived that this analogy does not hold, since the magnetic lines are in constant motion, while the isothermal lines retain very nearly a fixed position, or at least change in comparison with the other lines with extreme slowness. - The second system of changes has evident relation to the annual position of the earth in its orbit round the sun, and its revolution on its axis.

These were at first ascribed to the influence of the heat of the sun on different parts of the earth; but they have the remarkable characteristic of exhibiting notably the same amount in the southern hemisphere as in the northern, and in the tropical as in the temperate zones. The magnetic force is found to be greater in the months of December, January, and February, when the sun is nearest to the earth, than in those of May, June, and July, when it is most distant from it; whereas, were the effect due to temperature, the two hemispheres would be oppositely instead of similarly affected in each of these two periods. We must therefore ascribe the effect to the direct magnetism of the sun itself, and consider it established that this luminary like the earth possesses attracting and repelling poles, and that the effects on the needle result from the different positions of the earth in regard to these centres of action. The pole of the needle which is least distant from the sun makes a double diurnal movement in the following manner.

It arrives at its greatest western excursion four or five hours before the sun passes the meridian of the place, as if it were repelled; it then turns eastward with increasing celerity, and reaches the limit of its eastern excursion one or two hours after that passage. As the sun passes the inferior meridian, there is repeated in the night the same variation as that which took place in the day. To illustrate the action, let us suppose two globes, a larger and a smaller, placed upon the same plane, with their axes of revolution not precisely parallel to each other, as in the case of the earth and the sun; and let us further suppose that one globe is made to revolve round the other, the axis of the former being constantly parallel to itself. It is evident that in one half of the orbit of the moving globe the northern poles will be inclined toward each other, while in the other half of the orbit the southern poles will be similarly inclined; and if we further suppose that the magnetic axis of the sun, as in the case of the earth, does not differ very much from the axis of rotation, we shall have an explanation of the effects observed in the records of the diurnal motions of the needle.

The N. end of the needle, which is attracted by the N. pole of the earth, will be repelled by the N. pole of the sun, provided it has dissimilar magnetism to that of the earth, and consequently will decline from the sun; and as, on account of the revolution of the earth on its axis, this luminary appears on the E. of every place in the northern hemisphere in the morning and on the W. side in the afternoon, corresponding variations in the needle will be exhibited. In the other half of the year, for a similar reason, the S. end of the needle will be affected in an analogous but opposite manner; the strength of the magnetism of the earth will be increased by the nearer approach of the sun, in the same way that two magnets having their dissimilar poles opposite each other are increased or diminished in magnetic power by a diminution or decrease of distance. We are indebted for the interesting discovery of the polar action of the sun to Gen. Sabine of England, who has had charge of the reduction of all the magnetic observations of the English colonial observatories; and to Dr. Kriel of Austria for another of the same character, which leads us to extend the principle of magnetism to the moon.

It is found that there is a variation of each of the magnetic elements corresponding with the diurnal position of the moon in regard to the earth; but this resembles the tides in exhibiting two maxima and two minima in the course of 24 hours, regularly changing in time with the motion of the moon in her orbit around the earth. These phenomena indicate that the moon is not magnetic per se, that is, possessed of permanent magnetism, but its magnetic condition resembles that of soft iron developed by the continued but varying inductive influence on account of change of distance of the earth and the sun. That these changes in the magnetic elements cannot be due to heat in this case, must be evident, since the temperature of the moon as a mass is but little greater than that of celestial space. - The third class of variations, which Avas formerly denominated fitful, is now known in a certain sense to be periodical. They were called by Humboldt magnetic storms, and were found by Arago to accompany the appearance of the aurora borealis.

Although it is impossible to predict from our present knowledge the recurrence of individual cases of these great perturbations in the intensity and direction of the magnetism of the earth, yet they are known to increase in number and magnitude of action within the period of a little more than five years, and gradually to diminish through nearly an equal period, the whole cycle being completed in a little more than 11 years. The magnetic storms have been observed in the most distant parts of the earth, and no doubt can now exist as to their cosmical character. The lunar influence of which we have just spoken does not appear to participate in or be connected with this inequality. The periodicity of these apparently fitful variations of magnetism was first pointed out by Gen. Sabine, and has since been established by the investigations of Prof. Lloyd of Ireland, Dr. La-mont of Germany, and by those of Prof. Bache from the observations made under his direction at Girard college. But the most astonishing result in regard to this class of perturbations is that they coincide with the periodical recurrence of the maxima and minima of the spots on the sun. A German astronomer bchwabe, has established, by nearly 30 years of unremitting daily observation, the periodicity of this phenomenon.

He finds that the solar spots increase in magnitude for about 5 1/2 years, and diminish through an equal period, the cycle, as in the case of magnetic storms, being completed in about 11 years. The discovery of a connection of this remarkable kind gives to magnetism a high position in the scale of distinct natural forces, and assigns to it equally with gravitation a truly cosmical character. It is not impossible that the spots on the sun may be connected with the falling into its gaseous envelope of meteorites, and this suggestion is favored by an observation of Mr. Carrington of England, in which a remarkable appearance was observed on the surface of the sun, analogous to that which would have been produced by an occurrence of the kind we have mentioned. Recently Prof. Loomis of Yale college has published his analysis of the observations of many past years, apparently placing beyond all question the existence of a connection between the sun-spot period, terrestrial magnetic disturbances, and the frequency of auroras. One of the most interesting questions belonging to the future of this subject, is the possible existence of an association between the phenomena of the sun's colored prominences and the magnetic activity of the earth.

Observations by Prof. Young of Dartmouth college seem to show the extreme probability of such an association. Moreover, the observations which have been made on the prominences, by showing a connection between these objects and the solar spots, seem to force upon us the conclusion that some relation exists between the colored flames and the phenomena of terrestrial magnetism, since the partial dependence of these upon the sun's condition as to spots has been very nearly if not quite demonstrated. - It is not intended by what has been said to convey the idea that meteorological changes may not affect the position of the needle, and that even the magnetic condition of the atmosphere, according to the hypothesis of Faraday, may not produce appreciable results; but as yet the actions of these appear to neutralize each other, and to leave no definite record of their existence in the course of periods of considerable length. It is probable, however, that with the improved photometrical instruments and a more minute scrutiny of their records, the effects due to these causes will be shown.

Since tho agitation of the atoms of an iron bar is found to favor the development of magnetism by induction, it is not improbable that the magnetism of the earth may be disturbed during the continuance and shortly after the occurrence of an earthquake.

## Terrestrial Magnetism #1

See Magnetism, Terrestrial.