Now to such a problem as this - a problem whose grandeur cannot but be recognised even by those who reject the conclusions adopted by Madler - the new method of research is applicable with peculiar force. For instance, if the stars of Taurus are circling round a particular orb also in Taurus, it will be manifest, on a moment's consideration, that they can have only a slight motion either of recession or approach with respect to the sun. When from our station on the earth we see Venus or Mercury nearly in the same direction as the sun, we know that at the moment either planet has only a thwart motion, being then either at its greatest or least distance from us. So that if the new method were applied to stars in Taurus, and showed that swift motions of recession or approach are there in progress, it would at once dispose of the attractive but too speculative theory of the German astronomer.

This has not yet been accomplished; in fact, since Dr. Huggins' instrument was mounted and in order, the constellation Taurus has not been well placed for observation by the new method. But in the meantime, evidence of the most convincing nature has been obtained to show that Madler's theory is unsound.

We have seen that the theory was based, in the main, on a certain general community of apparent motion among the stars in Taurus. Madler took it for granted that this community of motion is exceptional. It did not occur to him to examine the motions of stars in other parts of the heavens, to see whether perchance a like feature might not present itself elsewhere.

Having been myself led by other inquiries than Madler's to the conclusion that the stellar motions might afford useful information as to the structure of the heavens, I thought it desirable to make a chart showing all the known stellar motions in such a way that wherever a community of direction exists it would be at once apparent in the chart. Little arrows affixed to the star-discs on the map, showed by their direction and length the nature and amount of the stellar thwart motions. When the map was completed, it was easy to see that the community of motion in Taurus was only one instance, and by no means the most striking which could be recognised, of a phenomenon which I have since called star-drift. Certain sets of stars are seen to be moving athwart the heavens, nearly in the same direction, and nearly at the same rate, in such sort as to show that they form distinct families of suns, travelling onwards-each family as a single group through the celestial spaces.

If this view is just, Madler's theory is at once shown to be unsound; since the stars in Taurus thus appear as simply a drifting family of stars, one among several such families.

All that was required to make the proof convincing was, that one of these sets of drifting stars should be shown to be either approaching the earth or receding from it as a single group.

Now, among the instances of star-drift, there was one in the Great Bear which presented some very striking features. Five stars in this constellation, known as Beta, Gamma, Delta, Eta, and Zeta, were seen to be travelling, not merely at the same rate and in the same direction, but on a course precisely opposite to that which they would have had if their apparent motion had been due to the sun's motion in space. Moreover, all these stars are large and conspicuous; while one of them, Zeta, is distinguished by having two companions, one very close to it, and the other so far away that its motion around Zeta is only completed (according to Madler's computation) in a period of about 2,000 years; so that, if all the five large stars form a single system, the cyclic revolutions of the system must require millions of millions of years for their completion.

I selected this family of stars as affording a convenient means of testing (crucially) the accuracy of my theory of star-drift. If that theory is just, all these stars must be either approaching or receding at a common rate. If the theory is unsound, the chances are enormous against their possessing a common motion of approach or recession. I expressed a strong feeling of confidence that whenever Dr. Huggins applied the new method of research to these stars, he would find that they are either all approaching or all receding, and at one and the same rate. When I expressed this opinion, I knew that before many months had passed, the matter would be decided one way or the other.

Nothing could be more complete than the confirmation of my views by Dr. Huggins' observations. In his table of stellar motions, Dr. Huggins brackets together the five stars in question as possessing a common motion of recession at the rate of about twenty miles per second. Moreover he finds, from the nature of their spectra, that they are all alike in physical constitution.

It is hardly necessary to insist upon the importance of this result. It proves, first, that in this instance-and therefore presumably in the other instances - of apparent star-drift, there is a distinct family or group of stars, travelling bodily onwards amidst the star-depths. It is shown that the motions taking place within this star-family are small compared with the common motion of the group. It can be inferred that the group is relatively isolated, since otherwise we should find other stars in the Great Bear sharing in the motion of these five; and also, if there had been a 'disturbing orb at a moderate distance from the group, the members of the family would ere this have lost their uniformity of motion. Whatever may be the centre around which these five stars are moving as a single group, the distance of that centre must exceed enormously the dimensions of the group, precisely as the distance of the sun from Jupiter's satellite family enormously exceeds the dimensions of that system. Yet the distances separating the stars of the Great •Bear are themselves amazingly vast. The distance between Beta and Zeta of the Great Bear cannot be less than 100,000 times the distance separating our earth from the sun, and is probably far vaster. What then must be the distance of the centre of motion, as seen from which this enormous space is reduced to an almost evanescent are!

It seems not unlikely that we ought to regard the family of stars here recognised as bearing the same general relation to the stellar universe (or to that portion of it to which our sun belongs) that a group of meteors bears to the solar system. All the drifting star-families may not indeed travel around one and the same centre; or there may be no true centre, but only a central region, round which these movements take place: but it is impossible to consider thoughtfully any instance of community of stellar motions without feeling that it implies a common influence affecting in the same or nearly the same way each member of the drifting star-family. If there is but one such centre, whether it be a single orb, or a central region of thickly clustering stars, there now seems to be at least a possibility that we may find where this centre lies. When only a few more star-families have been recognised, and their motions of approach or recession determined, it will be a problem of no inordinate difficulty to deduce the position in space of the regions round which these motions are taking place, or else to prove (which would equally be a solution of the problem now before us) that no such region exists, and that the stars drift around more centres than one.

Whatever success may attend the efforts made to explain the stellar motions, there can be no doubt that the problem is well worthy of the most thorough investigation. There is, indeed, something startling in the thought that man, placed as he is on a tiny orb-an orb rotating swiftly on its axis, carried swiftly round the sun, and borne along with him in his swift motion through space-man, shortlived and weak, and unable by his unaided vision to perceive a thousandth part of the star-system, should yet attempt (and not unhope-fully) to master the secret of its structure and motions. It may be that what has hitherto been done is but the beginning of the series of labours by which, if ever, that end will be accomplished; or it may be that we are nearer to the mastery of the problem than we at present imagine: but, in any case, there is but one course by which success can be achieved. Piece by piece the facts on which our reasoning is to depend must be gathered together; while at every stage of the inquiry, the full meaning of observed facts must be as far as possible evolved. Success will not be obtained by observation alone, nor by theorising alone; but by that combination only of observation and theory to which we owe all the most important discoveries hitherto effected by astronomers.

Fraser's Magazine for November 1872.