Nor does the rarity of the atmosphere in which the comet was moving serve to help us in our difficulty. Doubtless we are little familiar with the effects which terrestrial elements would experience if they were distributed freely in the ether occupying the interplanetary spaces. But so far as our experience enables us to judge, we should rather look for intensity of cold than of heat under such circumstances. We see the heights of the Andes and of the Himalayas clothed in perpetual snow, though day after day the fierce heat of the tropical sun pours down upon them, and though there is no winter there (in our sense of the word) during which the snows are accumulated. We know that the explanation of this peculiarity lies in the extreme rarity of the air at a great height. It seems, therefore, reasonable to conclude that the cold of the interplanetary spaces must be far greater. Yet here we have an object whose light comes from the incandescent vapour of so fixed and unchangeable a substance as carbon, and thus, in place of an almost inconceivable intensity of cold we find the evidence of intense heat.

It seems impossible, at present, to suggest any explanation of the observed phenomena. That carbon exists out yonder in space in the state of luminous gas or vapour, is the one fact of which alone we can be certain. Dr. Huggins in his treatment of this fact suggests the possibility that the carbon may be divided into particles so minute, that as the comet approaches the sun, more of the sun's heat is gathered up, so to speak, and that thus the carbon is volatilised. He also points to phenomena of phosphorescence and fluorescence in illustration of the appearance presented by the comet's spectrum; but without suggesting any association between these phenomena and those presented by comets.

One cannot help associating the new views thus opened out to us respecting comets, with the discovery recently made that the meteoric bodies which flash singly or in showers across our skies belong in reality to the trains of comets. We have now every reason to believe that there is not a single member of the meteoric systems, not a single aerolite, bolide, or fire-ball, that has not belonged once upon a time to a comet. The evidence on which this view is founded, though it may seem insufficient at a first glance, is almost irresistible to those who can appreciate its significance. Let us briefly recapitulate the facts. ,

It has been proved that shooting-stars come from the interplanetary spaces, that they form systems, and that these systems travel in regular elliptical orbits about the sun. Two of the systems produce striking meteoric displays, viz. the system encountered by the earth on or about August 10, and the system encountered on or about November 13. Now it had been suggested that the members of the former system follow the track of the conspicuous comet which made its appearance in the year 1862; and it was proved that, assuming the orbit of the meteors to be very eccentric, and assigning to them a period of 147 years (that of the comet), their motions corresponded in the most remarkable manner with the orbital track of the comet. In fact the agreement was so close that very few who had examined the question could believe it to be accidental. But there were two objections on which some stress was laid. First, it had been necessary to make assumptions respecting the motion of the meteors; secondly, those assumptions were not rendered probable by anything which had been proved respecting any meteoric system. The examination of the November star-shower by a host of eminent mathematicians in 1866-7 led to results which at once removed these objections, and brought new evidence - and that of the most striking character - in favour of the theory that comets and meteors are associated. It had been supposed that the November meteors travelled in a nearly circular orbit within a period of somewhat less than a year. Adams proved that they travel in an orbit extending far out into space beyond the orbit of distant

Uranus. And the period of this orbit was calculated to be 33 1/4 years. Here then was strong confirmatory evidence in favour of the elliptic orbit and the long period assigned, by way of assumption, to the August meteors. But this was far from being all. Astronomers looked for a comet to be associated with the November meteors; and they found one - a small one, it is true, but with a well-defined character-an orbit calculated beyond suspicion of important error, and agreeing so closely in its motions with those of the November meteors that the chances were millions on millions to one against the coincidence being accidental. It hardly required, after this, that an association should be pointed out between other meteor-systems and other comets. Yet this has been done, and thus that which had already been demonstrated was illustrated by new proofs. We may say that nothing which men of science have dealt with has ever been more satisfactorily proved than the fact that meteors are the attendants on comets. , Now, how meteors are thrown off from cometic nuclei we are not yet able to say. They differ wholly in character from their source, and thus we learn that the gaseous nature of cometic nuclei is due to the action of causes connected with those to which the nuclear (structure of the comet's head is due. But whether the first formation of meteoric systems is associated in any way with the processes which result in the formation of a comet's tail, is not quite so clear. As yet no comet which has had a brilliant tail has been subjected to spectroscopic analysis, so that we cannot pronounce with any certainty respecting the structure of these singular appendages. Some astronomers are disposed to look on the formation of a track of meteors all round the orbit of a comet as due to the action of influences by which parts of the comet's mass are thrown into orbits of slightly longer period than that of the head, though closely resembling that orbit in figure. Be this as it may, it is certain that the great contrast in character between the meteoric bodies which form the train of a comet, and the gaseous nucleus and coma, remains yet among the mysteries which astronomers have been unable to clear up.

But so soon as it had been shown that a comet's head is formed of a certain well-known terrestrial substance, it was natural that the question should be asked whether this substance is to be found in meteors. Hitherto no great progress has been made in determining the elementary constitution of meteors which have not actually fallen upon the earth. It is so difficult to catch them during their brief transit across our skies that only a few substances, as sodium, phosphorus, magnesium, and so on, have been shown with any appearance of probability to exist in shooting-stars. Certainly carbon is not among the number of those elements which have been detected in this way. But at a recent meeting of the Astronomical Society, it was stated that several aerolites contain carbon in their structure, and Dr. De la Rue offered a fragment of one of these to Dr. Huggins for analysis. Certainly a strange circumstance that an astronomer who had analysed the structure of a body millions of miles away from the earth, should take into his hands and subject to chemical analysis a fragment which had once in all probability belonged to a similar comet!

In conclusion, I must notice that there has been a remarkable absence during the past few years of those brilliant and long-tailed comets which alone seem calculated to afford the spectroscopist the means of answering some of the difficult questions suggested above. The tail of Winnecke's comet was too faint to give a visible spectrum. In fact the comet itself was only just visible to the naked eye. When a blazing object like Donati's comet or the comet of 1861 comes to be subjected to spectroscopic analysis, we may hope for an amount of information compared with which that hitherto obtained is probably altogether insignificant.

From Frazer's Magazine for February and June 1869.