But besides these variations which we have mentioned, there are changes steadily going on, by which the isodynamic, isogonic and isoclinic lines are permanently displaced on the surface of our planet. These must be attributed to changes of temperature in the interior of the globe, and to the direction in the progress of subterranean fires, which it may also be expected will change the isogeothermal lines. But there are changes, which although of long period, are yet periodic, one of which is obviously due to the revolution of the lunar nodes in eighteen and a half years, and the revolution of the apogee in nine years. The first is continually changing the obliquity of the axis of the vortex, and they both tend to limit the vortices in their extreme latitudes; but the planet Jupiter has an indirect influence, which is probably equal, if not greater, than the action of the moon, in changing the magnetic declination. From the investigations of Lamont, it would appear, that the period of the variations of magnetic declination is about 10⅓ years, while, more recently, R. Wolfe has suggested the connection between this variation and the solar spots, and assigns a period of 11.11 years, and remarks, that it "corresponds more exactly with the variations in magnetic declination than the period of 10⅓ years established by Lamont. The magnetic variations accompany the solar spots, not only in their regular changes, but even in their minor irregularities: this latter fact is itself sufficient to prove definitely the important relations between them."[28] As the planet Jupiter exerts the greatest influence on the sun, in forcing the centre from the mechanical centre of the system, the longitude of the sun will in a great measure depend on the position of this planet; and, in consequence, the sun will generally revolve around this centre in a period nearly equal to the period of Jupiter. The sidereal period of Jupiter is about twelve years, but the action of the other planets tend to shorten this period (at least, that has been the effect for the last twenty or thirty years), and bring it nearly to the period assigned by M. Wolfe to the variations in the magnetic declinations. As this has its influence on the radial stream, and the radial stream on the declination, we see at once the connection between them. When we come to a consideration of the solar spots, we shall exhibit this influence more fully. AURORA BOREALIS.