belonged to the same stream, although the true relationship of the phenomena was not immediately apprehended. Olmsted ascribed to the meteors a revolution about the sun once in every six months, bringing them to the intersection of their orbit with that of the earth every November 13th; but later investigators found that the real period was about thirty-three and one-quarter years, so that the great displays were due three times in a century, and their return was confidently predicted for the year 1866. The appearance of the meteors in 1832, a year before the great display, was ascribed to the great length of the stream which they formed in space—so great that they required more than two years to cross the earth’s orbit. In 1832 the earth had encountered a relatively rare part of the stream, but in 1833, on returning to the crossing-place, it found there the richest part of the stream pouring across its orbit. This explanation also proved to be correct, and the predicted return in 1866 was duly witnessed, although the display was much less brilliant than in 1833. It was followed by another in 1867. [Illustration: Curious forms of meteorite trains Nos. 1 to 6 show the changes undergone by a train left by a meteorite which passed near the “Great Dipper”; 7 shows the changes and drift of a train seen in the constellation Virgo; 8 is the singular train of the meteorite of February 22, 1909, near the Pole Star. (From _La Nature_.)] In the mean time Olmsted’s idea of a cometary relationship of the meteors was demonstrated to be correct by the researches of Schiaparelli and others, who showed that not only the November meteors, but those of August, which are seen more or less abundantly every year, traveled in the tracks of well-known comets, and had undoubtedly an identical origin with those comets. In other words the comets and the meteor-swarms were both remnants of original masses which had probably been split up by the action of the sun, or of some planet to which they had made close approaches. The annual periodicity of the August meteors was ascribed to the fact that the separation had taken place so long ago that the meteors had become distributed all around the orbit, in consequence of which the earth encountered some of them every year when it arrived at the crossing-point. Then Leverrier showed that the original comet associated with the November meteors was probably brought into the system by the influence of the planet Uranus in the year 126 of the Christian era. Afterward Alexander Herschel identified the tracks of no less than seventy-six meteor-swarms (most of them inconspicuous) with those of comets. The still more recent researches of Mr W. F. Denning make it probable that there are no meteors which do not belong to a flock or system probably formed by the disintegration of a cometary mass; even the apparently sporadic ones which shoot across the sky, “lost souls in the night,” being members of flocks