No fact in the science of electro-magnetism is, perhaps, better established than the disposition of an ethereal current to place itself at right angles to the magnetic meridian, and conversely, when the current is not free to move, to place the needle at right angles to the current. Now, the terrestrial magnet or magnets, may be considered to be surrounded by a body of ether in rotation, which, in the earth, on its surface, and for some distance from the surface, is made to conform to the general rule, that is, to circulate at right angles to the magnetic meridian. Outside this again, the ether more and more conforms to the position of the axis of the vortex, and this position varying, it must exert _some_ influence on the surface currents, and, therefore, change in some degree the position of the magnetic meridian. The radial stream comes from the sun in parallel lines, and strikes the globe and its superficial ethereal envelope just as we have shown its action on the atmosphere; but in this last case the magnetic equator is not a great circle, neither can we suppose its effects to be an accumulation of a fluid which is imponderable at points 90° from the plane passing through the centre of the earth and sun, and coincident with the plane of the central meridian, and a depressing effect on that meridian. Its precise influence must be, from the nature of the cause, to deflect the circular current towards the poles, in places less than 90° from the meridian, and a contrary effect must be produced in places greater than 90° from the meridian. Let us assume, for argument's sake, that the magnetic poles of the earth correspond to the poles of rotation, the parallels of latitude will, therefore, represent the ethereal currents circulating around the globe. Now, at sunrise, the radial stream of the solar vortex is tangential to the surface, and, therefore, can produce no change in these currents. As the sun ascends say about 8 or 9 A.M., the radial stream striking only the surface of the earth perpendicularly in that place where the sun is vertical (which we will suppose at the equator), streams off on every side, as the meridians do from the pole, and the circles of latitude (that is the ethereal currents) being parallel to the equator, they are met by the radial stream obliquely, and deflected towards either pole. By this deflection they are no longer at right angles to the meridians. But, from the principle of reaction above noticed, the magnetic meridians will place themselves at right angles to the current, or, in other words, the magnetic pole will change its position on the surface of the earth with respect to that particular place. But, in other parts of the world, the meridians are in opposite phases at the same instant of absolute time; therefore, the magnetic poles are not points, but wide areas enclosing the magnetic poles of all the countries under the sun. As this conforms to observation, it is worthy our especial attention, and may be understood by the subjoined figure, in which the oblique curves represent the course of the tangential current in the different positions of the sun, the parallel lines representing the solar radial stream.