remarkable a difference. Accordingly Colonel Beaufoy announced his belief that some mistake had been made by Arago. But presently the London needle gave signs of having exhausted the energy of its westerly tendencies. In April, 1819, the end of the westward march came about, and from that day to this the London and Paris needles have both been steadily returning towards the north.

Some points are very worthy of notice in this matter. In the first place, the westerly sway of the Paris needle occupied nine years less than that of the London needle. The range of the former was 221⁄2 degrees, while that of the latter was 2423 degrees. We know not what amount of faith can be placed in the old observations; but it appears highly probable that they were very rough in comparison with those made in recent times. Still there can be little doubt that the needle really pointed due north earlier in London than in Paris. In this case then there is a somewhat perplexing anomaly; for, on account of the regular manner in which the phenomena of terrestrial magnitude appear to travel westward, it was to have been expected that Paris would have had a northerly needle earlier than London. In Copenhagen, for example, the needle pointed due north earlier than either in London or in Paris. Undoubtedly, however, there are circumstances of a local character which tend considerably to modify the march of the magnetic needle. In different parts of the globe the declination of the needle is different. If we divide the world into regions of westerly needles we find a somewhat irregular outline defining the several districts. This outline which, of course, is a line of no variation, has attracted a large share of the attention of our men of science. But in reality its peculiarities are little worthy of the attention they have received. It is very important that seamen should know where the line of "no variation" runs, and also along what lines a given declination prevails. Thus the "Halleyan lines," as (after Halley, the first to draw them) these curves have been called, are very valuable in themselves. But the indications which scientific men have attempted to draw from them are founded on a very unsound basis. A little consideration will show that very insignificant changes or irregularities of change in the earth's magnetism would suffice to make the line of "no declination" shift about in the wildest manner. We are not, therefore, to infer from the strange looped curves with separate ovals, which are depicted in works on terrestrial magnetism, and are shown to wave to and fro across whole continents,

loops changing into ovals and ovals into loops, or both disappearing for a while to reappear in new positions, that the earth's magnetic forces are in a strange state of disturbance and turmoil. The main fact to be noticed is that, on the whole, the line of "no declination" travels westward round the earth.

At present the earth may be divided into three regions, two of westerly and one of easterly magnetic declination. The principal westerly region includes nearly the whole of Europe and of the Atlantic ocean, the whole of Africa, Arabia, Turkey in Asia, and a part of Persia, a small slice from the north-eastern part of Brazil, Greenland, Labrador, and Canada, and the western part of Australia. The smaller western division is merely an oval in the eastern part of the Asiatic continent, extending from Yakutsk to the Phillippine Islands in a north and south direction. All the remainder of the globe belongs to the division of easterly needles.

Hitherto we have been considering only the horizontal needle; but one of the most important peculiarities of the magnetised needle consists in what is termed the dip or inclination. If we suspend a non-magnetised needle by a line passing through its centre of gravity, the needle will rest in any position. But if the needle be now magetised it will be found to be no longer thus free. If placed in a horizontal position and then suffered to seek its position of rest, the northern extremity will immediately dip downwards very sensibly. In fact, in our latitudes the needle tends more nearly to the vertical than to the horizontal position. In the southern hemisphere the opposite end of the needle dips. Hence between the two positions there must be one of horizontality. For example, suppose an observer were to set out from London, where the northern end dips, to the Cape of Good Hope, where the southern ends. As he journeyed he would find the needle gradually shifting from one position to the other. This must happen in two ways; but we need only concern ourselves with the observed process of change, according to which our traveller would find the northern end of the needle gradually rising as he travelled farther south. Therefore, since at the end of his journey he finds the northern end uppermost, whereas at the beginning the southern end was uppermost, it follows that at some part of his journey the two ends were exactly on the same level. Supposing he noted the place where this happened, on a map, he would have determined one point on the "line of no inclination." Now imagine a number of travellers setting out from all places in the latitude of London to

all places in the latitude of Cape Town, and that each recorded the place where the magnetised needle was horizontal. Then a line drawn through all these spots would be the "line of no inclination." It would surround the earth's equatorial regions, but would not be coincident with the equator. It has been found in fact that at present the line of no inclination lies to the south of the equator from about east longitude 6° round (toward the west) to about west longitude 150°, or through less than a complete semicircle, the remainder of the curve lying north of the equator. Then, again, if observers set out northwards from the latitude of London, or southwards from the latitude of Cape Town, they would find the inclination of the needle increasing. And if it were possible to search over the whole of the arctic and antarctic regions they would find two points, one in each region, where the magnetised needle would hang in an absolutely vertical position.

As a matter of fact, only one of these points has been actually determined, the difficulties which have attended antarctic voyaging having hitherto prevented our navigators from attaining the southern magnetic pole. To Captain Sir James C. Ross science owes the discovery of the northern pole, and the closest approach which has yet been made towards the neighbourhood of the southern pole. The northern magnetic pole lies (or perhaps we should rather say, lay at the time of Ross's arctic voyage) in 70° N. lat. and 117° W. long.; that is, to the north of the American continent, near the Gulf of Boothia. The southern seems to lie nearer the true pole, and the position assigned to it by Ross, from a careful examination of all the indications observed when he was nearest to it, lay in S. lat. 75°, and E. long. 154°.

It is customary in works on popular science to refer to the northern pole as though it were a fixed point; but there cannot be the least doubt that it changes its position. It would not be easy, of course, to watch the process of change, the pole being always so difficult of approach. But the mere fact that the dip of the needle is continually changing in every part of the earth, proves incontestably that the magnetic poles must be continually shifting. In London, for instance, the dip has been subject to a continual process of diminution from the period of the earliest recorded observation; and since the magnitude of the dip is a measure of the proximity of the magnetic pole, it follows that, during all that period, this pole has been receding from London. The same is true for Paris, where the series

of observations covers a yet larger period. In 1671, near which epoch it will be remembered the Paris needle pointed due north, the dip of the needle was fully 75°. A hundred years later it had diminished by two degrees and a half. At present it is about 66°. The inclination has diminished in London from 72° in 1786 to 70° in 1804, and thence to 68° at the present time.

Now it is not very difficult to form a general notion of the processes of change by which both the inclination and direction of the needle have been caused to vary in the manner indicated above. When the needle pointed due north in London or in Paris, doubtless the northern magnetic pole lay pretty nearly due north. I say, "pretty nearly," because we know that the directive powers of the needle are largely influenced by local peculiarities, and therefore the exact direction of the magnetic pole is not indicated by the needle. Now, lying due north of London and not coinciding with the true pole of the earth, the magnetic pole might either have been between London and the pole of the earth, or beyond the latter point. How are we to tell which of these positions was actually held by the magnetic pole in those days? Very easily. If the magnetic pole were between London and the pole of the earth, that pole was at its nearest to London; the inclination was therefore at its maximumn, and would thenceforth diminish. On the other hand, if the magnetic pole were beyond the pole of the earth the reverse would take place. We have seen that the inclination did actually diminish from the year 1671, and has been diminishing ever since. It follows, therefore, that the magnetic pole was between London and the pole of the earth at some epoch in the latter half of the seventeenth century.

We can also determine in what direction the pole moved away from this position. For the declination was easterly before that epoch, and westerly afterwards. Hence the pole must have moved towards the west. This result is confirmed by a fact which modern researches have established; the fact, namely, that the line of no inclination is shifting round from east to west. We may conceive, then, of the earth's magnetism as having a certain axial direction. The axis thus indicated passes through the earth's centre (we may assume) and is inclined some 23° to the earth's polar axis. Further the magnetic axis rotates around the polar axis, so as to trace out the surface of a gigantic double cone, having for its axis the polar axis of the earth. Can we form any estimate of the period of a complete revolution of the magnetic axis? It would scarcely be safe to attempt

this in the present state of our knowledge. The indications we have for our guidance are mixed up with many peculiarities. which are due to the local action, and it is very difficult at present to say what proportionate influence must be assigned to the general terrestrial change and the peculiarities of local influence. If we judged from the Paris needle, we should ascribe a period of about six hundred years; if from the London needle, the period would seem to be some thirty years longer. One other peculiarity affecting the magnetism of the whole earth must be considered before we direct our attention to those peculiarities which mark the daily and annual behaviour of the needle in any given place—which peculiarities, however, though seemingly more minute and limited, have led to the detection of the most widely significant of all the laws which affect the magnetic needle.

The two features hitherto dealt with include in reality only one relation, the declination and inclination being merely circumstances of the direction of the needle. But having found that in a given place the needle seeks to rest in a particular direction, we are led to consider whether the energy of this tendency is measurable, whether it is constant while the place or the time of observation is changed, and, if variable, what are the laws according to which the variation progresses. The property here referred to is commonly spoken of as magnetic intensity. But how are we to measure a force so minute? For, though the directive power of the magnetic needle is really the representative of a very important force, yet it is itself so feeble that (as we know) the minutest resistance serves to counteract it. Although it might be possible to apply delicate contrivances for estimating the intensity in a direct manner, the plan which has been found most readily available is indirect. In place of actually determining the resistance which the magnetised needle is capable of overcoming when seeking its position of rest, the observer sets the needle vibrating; and by counting the number of vibrations taking place in a given interval of time, he is able to estimate the relative intensity of the directive force at different times or in different places. For, the greater the power, the more rapid will the vibrations be. The principle, in fact, corresponds precisely to that which has been applied to determine the varying force of gravity at different parts of the earth's surface. Thus determined, the intensity has been found to vary very appreciably in different parts of the earth. Speaking generally, it is least where the dip is least; but this law is not exact,