The orbit of the Perseids cuts that of the earth almost perpendicularly. The bodies are generally supposed to be the result of the disintegration of an ancient comet which travelled in the same orbit. Tuttle's Comet, which passed close to the earth in 1862, also belongs to this orbit; and its period of revolution is calculated to be 131 years. The Perseids appear to be disseminated all along this great orbit, for we meet them in considerable quantities each year. The bodies in question are in general particularly small. The swarm has, however, like most others, a somewhat denser portion, and through this the earth passed in 1848. The _aphelion_, or point where the far end of the orbit turns back again towards the sun, is situated right away beyond the path of Neptune, at a distance of forty-eight times that of the earth from the sun. The comet of 1532 also belongs to the Perseid orbit. It revisited the neighbourhood of the earth in 1661, and should have returned in 1789. But we have no record of it in that year; for which omission the then politically disturbed state of Europe may account. If not already disintegrated, this comet is due to return in 1919. This supposed connection between comets and meteor-swarms must be also extended to the case of the Leonids. These meteors appear to travel along the same track as Tempel's Comet of 1866. It is considered that the attractions of the various bodies of the solar system upon a meteor swarm must eventually result in breaking up the "bunched" portion, so that in time the individual meteors should become distributed along the whole length of the orbit. Upon this assumption the Perseid swarm, in which the meteors are fairly well scattered along its path, should be of greater age than the Leonid. As to the Leonid swarm itself, Le Verrier held that it was first brought into the solar system in A.D. 126, having been captured from outer space by the gravitative action of the planet Uranus. The acknowledged theory of meteor swarms has naturally given rise to an idea, that the sunlight shining upon such a large collection of particles ought to render a swarm visible before its collision with the earth's atmosphere. Several attempts have therefore been made to search for approaching swarms by photography, but, so far, it appears without success. It has also been proposed, by Mr. W.H.S. Monck, that the stars in those regions from which swarms are due, should be carefully watched, to see if their light exhibits such temporary diminutions as would be likely to arise from the momentary interposition of a cloud of moving particles.