which, consequently, can no longer be attributed to that force. We see them drawn out and lengthened in the direction of the abovementioned ideal line; but the matter which gushed forth (so to speak) on the sunshiny side is quickly arrested and driven back, while on the opposite side the matter rushes out, without meeting an obstacle, with an impetuosity of which no earthly phenomenon can give an idea. In a few days there is thus produced a tail diametrically opposite to the sun; and this tail may stretch before our eyes to a length of ten, twenty, thirty, or even sixty millions of leagues.
So great, in truth, is the complexity of cosmical phenomena, even when connected with unknown forces, that we might despair of success were we not offered a resource of which astronomical science has often availed itself—namely, the comparison of phenomena which occur at the same periodical intervals. Long before the discovery of universal gravitation, it was easy to see that the tides depended on the moon, since the periodicity is identical for the oscillations of the sea and the movements of our satellite. In like manner, the most complicated phenomena of meteorology, if they manifest a periodical character and their course agrees with the period of other strange phenomena, betray thereby their connection, in some way or other, with the latter.
Let us take, as an illustration, the variations of the dip of the magnetic needle. Every day, that needle, suspended freely, deviates in the morning from the position of equilibrium, and every day returns to it in the evening, after an excursion of variable extent. These regular movements evidently depend on the presence of the Sun above the horizon of any given spot; they also depend on its geographical situation, for they increase with the latitude on one hemisphere, and change their direction in passing from one hemisphere to the other. They are not due to a magnetic action proper to the sun; for, even supposing it to exist, the Sun, in consequence of his enormous distance, would be incapable of exercising a directing influence on a magnetic needle; but they result from some unknown action exerted on the electricity of the globe, and on the currents resulting from it—currents which themselves react on the direction of the needle, and often seriously disturb its movements.
How are we to give a precise account of this mode of action? In the midst of so many unknown details, how are we to lay hold of those which really require our attention? Observation, only, aided by this special form of empiricism—pointed out by M. Faye in his "Notice," and here attempted to be described—can help us in the matter. The diurnal variations of the needle have been noted for nearly a century past; it is remarked that they are not constant from one year to another; that they present maxima and minima, epochs of greatest and least activity; that these maxima occur every eleventh year. The phenomenon is periodical, and its period is eleven years.
