fact, has become of these débris which failed to reappear at the time when their regular period should have brought them within sight of our globe?

Thus everything inclines us to believe that scattered here and there upon the planetary shores, or floating upon the waves of the ethereal ocean, shattered and broken-up comets exist; remains of shipwrecks suffered by millions of comets; waifs of aërial barks unable to accomplish the voyage without paying tribute to the element in which they float. Nevertheless such fragments, more or less disintegrated, do not wander at will in space; they move in orbits whose forms are dependent upon the modifications which the disturbing influence has brought to bear upon the original motions. We may suppose that these orbits continue to be closed curves of which the sun is the focus, or that they have become changed to hyperbolas. We may even suppose the disturbing influence to have been such that these disintegrated fragments have become satellites of the disturbing planet. Perhaps all these results and many more have been brought about in the course of time.

The number of comets which penetrate into our world is, in all probability, so immensely great, that during the hundreds of millions of years which we may assume to have elapsed since the beginning of the world the interplanetary spaces must have been furrowed by prodigious multitudes of currents of matter from disintegrated comets, fragments of comets, which the planets, in their regular course about the sun, cannot fail frequently to encounter. It can hardly be otherwise. And indeed at the present day there is nearly certain evidence that this actually is the case; and, although originally the proof of the existence of these material currents may have been arrived at in another manner, the fact of their existence is certain. Their origin alone can be matter of doubt.

SECTION III.

COMETS AND SWARMS OF SHOOTING STARS.

Periodicity of the meteor-swarms; radiant points; number of swarms recognised at the present day-Periodical maxima and minima in certain meteoric currents; thirty years period of the November swarm-Parabolic velocity of shooting stars; the swarms of shooting stars come from the sidereal depths.

THESE Considerations bring us to the theory recently elaborated by the learned Italian astronomer M. G. V. Schiaparelli, Director of the Observatory of Brera, at Milan.

According to this theory there exists between comets and shooting stars a connexion and community of origin, which henceforth we may regard as certain, as it is supported both by logical deduction and observation. We shall now explain by what train of ideas this assimilation between phenomena which at first sight appear so foreign to each other has passed from the phase of simple hypothesis into that of a theory which observations of great value permit us at the present time to consider demonstrated.

Let us first of all pass in review the facts upon which the theory is based.

The shooting stars which may be observed on any clear night throughout the year are notably more numerous at certain times, the dates of which are nearly fixed, as, for example, August 10, November 13 or 14, and April 20. They then appear in sufficient numbers to be considered, not as isolated

meteors, but as groups or swarms of meteors. This connexion soon became still more manifest when it was found that the stars of each swarm were moving in trajectories, not distributed at random over the celestial vault, but so disposed that if prolonged backwards all or nearly all passed, if not through the same mathematical point, at least through one very circum

scribed region of the heavens. To obtain an idea of this remarkable conjunction of the apparent trajectories of a stream of shooting stars the reader has only to glance at fig. 69, representing the tracks among the stars of a certain number of these meteors which appeared on the night of November 13-14, 1866.

The point of emanation or convergence of the trajectories of a swarm is called the radiant point. Now, not only have the meteors which appear at the fixed dates we have mentioned, both on the same night and on several successive nights, the same radiant point, but this radiant point does not vary in position, or varies very little, in the course of years for successive apparitions of the same meteor-current.

At the time when researches were first commenced respecting these singular phenomena the known streams of periodical return were few in number; those of August 10 and November 14 were at first alone recognised. It was thought that the shooting stars of ordinary nights were scattered without any apparent connexion. But more careful observations showed that in reality these sporadic shooting stars obeyed laws similar to those of the other swarms; a great number of streams were thus distinguished, and the positions of their radiant points determined. At the present time 102 are known.'

But another fact of great importance was established, partly by historical researches in regard to the previous apparitions of similar phenomena, and partly by the continuous observations of contemporary astronomers. The fact in question is this:The periodicity of meteoric swarms is not only annual, so that the same nights of each succeeding year are remarkable for displays of meteors sufficiently abundant to distinguish these nights from those immediately adjacent, but it also happens that, as regards certain streams, the display varies in a manner which clearly enables us to distinguish the recurrence of periodical maxima and minima. We will cite a remarkable instance of this periodicity, viz. that of the swarm of the middle of November. In 1799 the stream of shooting stars was of

wonderful intensity. The same phenomenon was repeated thirty-four years later; that is to say, in the year 1833. On tracing back the records of similar phenomena the display of 1766 was found to have been not less abundant at the same date. Similar instances were likewise found at more remote epochs, and it appeared that the intervals between these extraordinary apparitions were always either thirty-three or thirtyfour years, or a multiple of these numbers. It seems certain, then, that for this particular swarm there is a periodical maximum which recurs about every third of a century. Thus Olbers did not hesitate to predict, long beforehand, the recurrence of a maximum for the year 1867. Professor Newton, of America, furnished more precise data, and defined the period with greater accuracy, assigning to it a duration of thirty-three years and a quarter, and fixing the date of its next return for the night of November 13-14, 1866.

The meteor-swarm was true to the prediction.

It soon became clear that these singular phenomena could only be explained by considering the different swarms of meteors as of extra-terrestrial or cosmical origin. Other circumstances, which we shall enlarge upon in a separate work, contributed, in conjunction with those which we have just related, to prove that shooting stars form currents of celestial particles which circulate independently in space and describe regular orbits like comets. These numerous currents traverse the interplanetary spaces in all directions, and by their occasional rencontres with the earth give rise to the production of shooting stars. Our globe, or simply its atmosphere, penetrating more or less deeply to the centre of one of these groups, a meeting takes place, the velocity being sometimes equal to the sum and sometimes to the difference, or to any intermediate amount, of the respective velocities of the two bodies. In any case there is a loss of vis viva, or rather a trans