It is now exactly thirty years since the world rang with one of those discoveries which go down to the ages and at once insure the names of the makers of them being inscribed upon the muster-roll of the Immortals.

In the autumn of 1859, Kirchhoff and Bunsen announced that at last a way had been found of studying the chemical nature of bodies in space-nay more, that they had already begun the work, and found that the sun, at all events, was built up of matter identical with that of which the earth is composed.

In physical science, in most cases a new discovery means that by some new idea, new instrument, or some new and better use of an old one, nature has been wooed in some new way. In this case it was a question of a new idea and an old instrument. The instrument was the spectro

scope.

It forms no part of my present purpose to deal either with the principles involved in spectrum analysis or its history during the period which has elapsed since 1859. The task I have set myself in this article is a much more modest one.

First I wish to point out that during the thirty years the method of work which Kirchhoff and Bunsen applied to the sun has been applied to the whole host of heaven; by this I do not mean that every star has been examined, but that many examples of each great class-nebula, comet, star, planet-have been studied. The same kind of information has been obtained with respect to these bodies as Kirchhoff and Bunsen gleaned with regard to the sun; and the great generalization to which I have referred has been found to hold

good in the main for all. From nebulæ and stars existing in space in regions so remote that the observations have been of the utmost difficulty in consequence of the feebleness of their light; from comets careering through stretches of space almost at our doors, the same story has come of substances existing in them which are familiar to us here. In ascending thus from the particular to the general, from the sun to the most distant worlds, it is obvious that the field of observation has been enor

mously extended. Kirchhoff and Bunsen's view has been abundantly verified, as we have seen; but the question remains, Has this larger area of observation supplied us with facts which enable us to make a more general statement than theirs? It is possible that it has. Recent inquiry has suggested that if the study of meteorites be conjoined with that of the heav enly bodies, the story told by the spectroscope enables us to go a step farther, and to say that not only have we the same matter everywhere, but all celestial bodies, including the earth, are due to an exquisitely simple evolution of matter, in the form of meteoritic dust. We have no longer to rest content with the fact that all nature is one chemically; we have the

cause.

Secondly, I propose to make as short and simple a statement as I can, of the general idea of the new cosmogony suggested by the spectroscopic survey to which I have referred.

I must, in the first place, ask my readers to grant me the scientific use of their imagination; and in order that it may not be called upon to cope with questions as to whether space is infinite or not; or whether space and time ever had a beginning; we will not consider the possibility of the beginning of things, or attempt to define the totality of space, but we will in imagination clear a certain part of space, and then set certain possibilities at work.

How much space shall we clear? A very good idea of one of the units of space, which is very convenient for me to employ here-I mean the distance of the nearest star, or one of the nearest stars— can be obtained by stating the time taken by light in performing the journey between the earth and the stars, knowing as we do that light travels 186,000 miles in a second. In the case of the nearest stars the time thus required is about three and a half years. With regard to the twelfthmagnitude stars, we find that in all probability the distance in their case is so great that light, instead of taking three and a half years, takes three thousand five hundred years to reach us.

The space included in a sphere with this

radius will be sufficient for our purpose. The stars that we shall have to abolish for the purpose of this preliminary inquiry number something like six millions; the probability being that if we consider the stars visible, not in the largest telescopes, but in those which are now considered of moderate dimensions, their numbers may be reckoned at something between thirty and fifty millions.

Imagine then this part of space cleared of all matter. We shall have a dark void, and the probability is that all that dark void will, sooner or later, in consequence of conditions existing in other parts of space into which we have not inquired, be filled with some form of matter so fine that it is impossible to give it a chemical

name.

Next we may imagine that this something without a chemical name may curdle into something which is more allied with our terrestrial chemistry, and the chances are, so far as we know, that that first substance will be either hydrogen itself, or some substance seen in the spectrum of hydrogen or closely associated spectra.

It is just possible that at this point we enter the region of observation. In the nebula we are brought face to face with a substance (or substances) which, as far as our observations go, exists nowhere else except in the very hottest region of the sun that we can get at with our instruments. It is unknown here, and all attempts to match the spectrum by exposing terrestrial substances to the highest temperatures available in our laboratories have so far been unavailing. Both in sun and nebulæ this substance (or substances) is associated with hydrogen. This curdling process will go on until at length further condensation will take place, and instead of having simply the substance (or substances) to which I have referred, and hydrogen, we shall have an excess of hydrogen with an infinitely fine dust interspersed in it, which will go on condensing and condensing until at last we get dust of substances the existence of which is revealed to us in the spectra of bodies known to terrestrial chemistry; among these are magnesium, carbon, oxygen, iron, silicon, and sulphur.

This dust, fortunately for those interested in such inquiries as this, comes down to us in more condensed forms still, and it is in consequence of the messages which

they bring from the heavens that I am engaged in writing this article. Not only have we dust falling, but large masses; magnificent specimens of meteorites which have fallen from the heavens at different times, some of them weighing tons, are open to our inquiries. Although, therefore, it is very difficult for us to collect the dust, it is perfectly easy to produce it by pulverizing any specimens of these meteorites that we choose into the finest

powder. If we examine this dust spectroscopically, we find that, in addition to hydrogen, its chief constituents are magnesium, iron, carbon, silicon, oxygen, and sulphur..

I have, therefore, in this first sketch of a possible result of a process going on in our space-clearing at an early stage, not arrived at something that is unreal and merely the creation of the imagination, but something very definite indeed, which we can analyze and work with in our laboratories.

How it comes that this infinitely fine dust, finer probably than anything we can imagine, becomes at last, in the celestial spaces, agglomerated into meteoric irons and stones with which the earth is being continually bombarded, is one of the most interesting questions in the domain of science. Space is no niggard of this dust, for if we deal with agglomerations of it sufficient in quantity to give rise to the appearance of a falling star" to the unaided eye, we know that the number of such masses which fall upon the earth every day exceeds twenty millions.

66

We have then the idea before us that here and there in this space that we have cleared, we have initial curdling, as I have called it; we need not assume that these curdlings are uniform.

It is impossible with our present knowl edge to suppose that at any prior stage of the history of the heavens gravitation did not exist. It is impossible from what we know now to suppose that even the finest form of matter which entered our clearing in space was not endowed with motion. Given this matter, its motion and gravitation, let us next see what must very quickly follow.

Gravitation will give us a formation of centres; we shall get a rotation (moment of momentum) due to the prior existence of motion and to this formation of centres; we shall eventually in that way get con

densing masses of this curdled substance. The moment we have these centres formed, gravitation again will give us the motion of exterior particles toward these centres, and the condensation in one part of space will necessarily be counterbalanced by a clearing in another, so that, if we suppose that the curdling was not uniform to begin with, the uniformity will be less and less as time and this action go on.

Let us imagine that here and there we have isolated eddies, and here and there in the larger aggregations of the dust-in the most enormous swarms we can imagine we have also eddies; these eddies in volved in the larger curdlings will be associated with the phenomena of the general system of which they form an insignificant part. These cosmical molecules aggregating in this way will be, to compare great things with small, like the invisible molecules of a gas. It is not too much to say, as Professor George Darwin has recently shown, that we shall have in effect the whole mechanism of the kinetic theory of gases before us, but instead of dealing with invisible gaseous particles we shall have particles, large or small, of meteoritic dust. The kinetic theory tells us that if we have encounters we must have a production of heat; if we have production of heat we must have the production of radiation, although, if the heat be insufficient, the radiation may not produce light enough to be visible to the human eye.

It is a remarkable thought that all these changes to which I have so far drawn attention may have been going on in different parts of space for æons without any visible trace of the action being possible to any kind of visual organs. I refer to this because it is right that I should point out here that Halley, who was one of the first to discuss the possible luminosity of sparse masses of matter in space, and Maupertuis, who followed him, both laid great stress upon it.*

* "But not less wonderful are certain luminous Spots or Patches, which discover themselves only by the Telescope, and appear to the naked Eye like small fixt Stars; but in reality are nothing else but the light coming from an extraordinary great space in the Ether; through which a lucid Medium is dif. fused, that shines with its own proper Lustre. This seems fully to reconcile that Difficulty which some have moved against the Description Moses gives of the Creation, alleging that Light could not be created without the Sun.

When, then, these encounters, which we may call collisions, take place, and when the heat due to the arrested motion of the particles coming together, and the accompanying light are produced; must expect that that light will at first be very dim, and will require very considerable optical power to render it visible.

We may now consider some early results obtained in connection with this matter. Sir William Herschel, although not the first to examine into it, was the first to bring before us an idea of the magnificent spectacle which the heavens present to mankind, and he, without any difficulty, with his large instruments, began by dividing these dim bodies into nebulosities and nebulæ; the nebulosities extending over large spaces of the heavens, and being of very, very feeble luminosity.

When we pass from these we become acquainted with bodies which may be truly termed nebulæ, as opposed to nebulosities, and the most magnificent of these is that in Orion, which has recently been so grandly photographed by Mr. Common and Mr. Roberts, the latter using the intensifying action of four hours' exposure of the photographic plate, hereby revealing details that no human eye will ever see, thus demonstrating how true it is that these changes may go on for æons and æons, though the eye may never become acquainted with them.

There is a magnificent arrangement in the human eye which, though it invalidates it for some astronomical purposes, is convenient, because it enables us to go on using our eyes all our lives, whereas a prepared photographic plate can only be used

once.

By this arrangement, however long we look at an object, it does not appear brighter, but in the case of the photographic plate all the action upon it is totalled, so to speak, so that if the plate be exposed, say for two hours or sixty hours,

But in the following Instances the contrary is manifest; for some of these bright Spots discover no sign of a Star in the middle of them ; and the irregular form of those that have, shews them not to proceed from the Illumination of a Central Body, since they have no Annual Parallax, they cannot fail to occupy Spaces immensely great, and perhaps not less than our whole Solar System. In all these so vast Spaces it should seem that there is a perpetual uninterrupted Day, which may furnish Matter of Speculation, as well to the curious Naturalist as to the Astronomer."-Edmund Halley, Phil. Trans, vol. xxix, p. 392.

we shall go on getting impressed upon it more and more of the unseen.

Thus the nebula of Orion as seen is almost insignificant compared with the glorious object which the photographic plate portrays if the integrating power be allowed to go on for hours.

It seemed pretty obvious, since the light of such bodies is so dim that a large portion of it beats upon the earth and upon our eyes without having any effect upon either, that the temperature was low; and it seemed also that to test the idea that this luminosity might be produced, as I have suggested, by collisions of meteoric dust, the way was open for laboratory work.

Smash a meteorite, collect the dust, expose it to a low temperature: compare its spectrum with the spectrum of such a body as those we have been considering, and see by actual experiment if there is any similarity. This was done.

The result was almost identical. It seemed therefore that one had at last got to solid ground, and could go ahead. But how to go ahead in a scientific way? Naturally by developing the argument which had led us so far. Let us agree that the nebulæ are condensations of meteoritic dust, and see whether we are led to the true or the false by such a concession. Let us further grant that the condensations go on. What will happen next?

In certain regions of space the encounters-the collisions-will increase in number in consequence of the accumulation of meteoric dust in these regions; the temperature will therefore be higher and the light more intense.

Is there only one process by which the temperature can be increased? It did not take very long to recognize that there might possibly be three lines of action, each one of which would result in the production of a higher temperature.

In the first place, moment of momentum-rotation-being at our disposal to start with, it was obvious, in virtue of mechanical laws, that as the condensation went on the rotation would be accelerated; the motions of the particles of dust in the reaction, so to speak, would be more violent; the collisions therefore would produce more smashes, and more heat, and therefore more light.

We should get a central system and surroundings, such as Mr. Roberts has recently photographed in the great nebula

of Andromeda. The exposure he gave was four hours, and again this photograph brings us face to face with phenomena which will probably never be seen by the eye alone.

A central condensation, here and there fragments of spirals, and here and there dark gaps are seen. These gaps were observed by Bond and others years ago, but it remained for Mr. Roberts to demonstrate to us that they are produced by the wonderful in-draught action which we can now, by means of the photograph, see going on. We have a concentration toward the centre, the dark gaps representing to us either the absence of matter or the presence of meteoritic dust in a region where it is all going the same way, and in which therefore there are no collisions. Here and there we get regions of great luminosity, and associated with the spirals we get obvious loci of encounters. ternal swarms are also seen which have been thought, with great probability, to belong to the system-smaller condensations partaking in the general motion of the whole. Here then we are in presence of one possible cause of increased temperature.

Ex

There is another. One of the early results obtained by Sir William Herschel was that it was a very common thing for double nebulæ to make their appearance in his gigantic telescope. Now it is diffi cult for us to imagine that these double nebulæ, like their allied systems of stars, should not be in motion; and if we imag ine a condition of things in which one swarm is going around a larger one in an elliptic orbit, and occasionally approaching it, and mingling with it; we shall have at one part of the orbit the centres nearest together; so that a greater number of particles of meteoritic dust will be liable to encounters at this time than at others. Hence we shall get a cause of increased temperature of a periodic kind; there must be variable stars in the heavens and there are,

As a third possible condition we have the known movement of these swarms of dust through space. If we take note of the known movements of the star which forms the centre of our own system we can learn that these movements may be gigantic. We know that the sun is travelling nearly half a million of miles every twenty-four hours toward a certain re