We have seen how the passage of the axis of the vortex may derange the electric tension of the parts passed over; but there is another mode of action not yet adverted to.
When the moon is at her perigee, the axis of the vortex passes through the centre of gravity of the earth and moon at C, and cuts off the segment RR. At the apogee, on account of her greater distance, and of her consequent power to _push_ the earth out from the axis of the vortex XX, the segment R'R' is only cut off by the axis; and the angle which the axis makes with the surface will vary with the arcs AR and A'R'; for these arcs will measure the inclination from the nature of the circle. In passing from the perigee to the apogee the axis will pass over the latitudes intermediate between R and R' in both hemispheres, neither reaching to the equator E, nor to the pole P. Let us now suppose a meridian of the earth, represented by the line NRS, N being north, and S south, and the surface of the atmosphere by N'S'; XX still representing the axis of the vortex, ordinarily inclined 34° or 35° to the surface. Let us also conceive the rotation of the earth to cease, (the action of the vortex remaining the same,) thus leaving the axis over a particular longitude. If the ether possesses inertia, there will be an actual scooping out of the upper portions, driving them southward to a certain distance, where the atmosphere will be piled up above the ordinary level. There will, therefore, be a strong contrary current at the surface of the earth to restore the equilibrium, and if the action be violent, the surface wind will be increased; so that if it be considered tangential to the surface at S, its own momentum will tend to make it leave the surface and mount up to T; and in this way increase the action due to the ether. Now, although the axis is never stationary, but travels round the earth in less than twenty-five hours, yet there is a tendency to this mode of action; and it is even sometimes palpable to the observer when the axis has passed immediately to the northward; for the pinnate shafts and branching plumes of the cirri often reach far to the south of the southern boundary of the storm. These shafts are always longer when radiating from the northward than when proceeding from the southward. The cause is understood by the above figure. At such a time, after dark, the auroral shafts will also be seen over the storm to the northward, but will be invisible to those beneath. There is this to be observed, however, that the visibility of the ethereal current (or the aurora) is more frequent when the passage of the vortex is not attended with any great commotion, its free passage being perhaps obstructed by too dry an atmosphere; hence it becomes more visible. But it may be asserted that a great aurora is never seen except when a vortex is near, and to the northward, and within a few hours of its passage over the meridian. We have, however, seen partial auroras to the south when none existed north, and also cases when the radiation was from west, but they are never as bright as in the north. They are all due, however, to the same cause; and we have frequently followed a vortex for three days to the northward, (that is, seen the effects of its meridian passage,) at 700 miles distance, by the aurora, and even by the lightning, which proves plainly that the _exterior layers_ of our atmosphere can reflect a flash of lightning, assisted by the horizontal refraction, otherwise the curvature of the earth would sink it ten miles below the horizon.