295

THE CORONA AS A PHENOMENON OF ERUPTION.

BEFORE adducing fresh evidence in favour of the eruption theory, I think it desirable to make a few preliminary remarks on the subject of the accumulation of evidence in favour of theoretical views. It has been urged once or twice, and that by persons to whose opinion I feel bound to pay respectful attention, that in advocating a theory I have shown myself somewhat too apt to adopt a tone resembling special pleadingthat I have endeavoured to show not merely that certain arguments favour the theory, but that all the known circumstances of the case point in the same direction. This has been said, in particular, of those reasonings by which I endeavoured to convince others as I was convinced myself, long since, that the corona is a solar appendage, and not, as had been urged, a mere phenomenon of our atmosphere or due to the illumination of matter lying between the earth and the moon. I dwelt very earnestly on my views about the corona, because I felt that when the evidence was duly weighed, the atmospheric theory could not but be regarded as disposed of, and that the labours of eclipse observers could be directed to determine the nature of

the solar surroundings, instead of being wasted on the inquiry whether in the corona solar surroundings were in question at all. It was urged (and not untruly) that a variety of circumstances seemed to favour the atmospheric theory, and that the consideration of these circumstances had led others to entertain an opinion different from that which I was advocating.

The circumstance that the observations made during the late eclipse have at length definitely convinced even those who had been most doubtful, that the corona is a solar appendage, and that the able observer, Janssen, has, in the most uncompromising terms, enunciated the opinion which I advocated so earnestly two years ago,' places me in a position to defend with some confidence that method of advocacy which has been depreciated.

It seems to me, then, that those who oppose the accumulation of evidence in favour of a theory-that is, the attempt to show that all the known circum

1 He writes, 'Le résultat de mes observations à Sholoor indique, sans aucun doute, l'origine solaire de la couronne et l'existence de matières au-delà de la chromospéhre;' and again, Rien de plus beau, de plus lumineux, avec des formes spéciales qui excluent toute possibilité d'une origine atmosphérique terrestre. . . . Je crois tranchée la question de savoir si la couronne est due à l'atmosphère terrestre, et nous avons devant nous la perspective de l'étude des régions extra-solaires, qui sera bien intéressante et féconde.' The expressions here used are singularly emphatic, and they are the more effective that Janssen himself had adopted no definite theory of the corona, while M. Faye, to whom they were addressed, had once been the chief advocate of a non-solar interpretation—though (as appears from some expressions in Janssen's letter) M. Faye had of late adopted the theory now shown to be the just one.

stances favour the theory-have not sufficiently kept the fact in mind that the only circumstance which can justify the advocacy of a theory at all, is the conviction on the part of the theorist that the theory is the true one. If he merely conceives that the theory is more probable than some other theory or theories which have been discussed, he should only present those facts which appear to favour the theory, making careful mention of the circumstance that other facts seem less favourable or eveu contradictory to the theory. Until he can show that not one known fact opposes the theory, or seems difficult of explanation in connection with it, he has no right to advocate the theory at all (in the proper sense of these words). But the case is altogether different when he is convinced that a theory is the correct one. For this conviction can only arise from the fact that every known fact has been carefully compared with the requirements of the theory, and found to be altogether in accordance with it. It is only to false but plausible theories that Voltaire's description applies: A theory is like a mouse which passes through nineteen holes, and is stopped at the twentieth.' The very first duty of a theorist who conceives that he has lighted upon the true explanation of any physical phenomenon, is to test his theory by bringing it into the presence of every single known fact respecting that phenomenon, and to forsake it at once if a single fact is opposed to it. Supposing it stands such a test, how can the theorist otherwise present it to the world than as in accordance with all

known facts, showing carefully that this is so in the case of each particular fact? His doing this may have an air of special pleading to those who conceive that he is bringing forward facts thought of after the theory had been adopted; but when it is understood that those facts have been the tests to which the theory was exposed before being adopted, it will appear not only natural, but necessary that they should be adduced.

It seems to me, indeed, that the course reprehended as resembling special pleading is so far from being objectionable that it is to be regarded as the only proper course for the theorist; and I am sure that if this method were applied rigidly to every theory suggested to the minds of scientific inquirers, we should have fewer ill-digested theories under discussion than at the present time. Speaking for myself, I can safely assert that if the method has on a few occasions caused me to be somewhat confident respecting theories which I have advocated and still hold, it has again and again caused me to abandon theories which I had begun to regard with favour and should undoubtedly have still entertained but that on some one or other point they would not bear the test thus applied to them.

Now it will be noticed that the eruption theory of the solar corona was not advocated in my former paper. It was described, and the evidence, which seemed favourable to it, was discussed at considerable length; but although no evidence suggested itself as unfavourable to the theory, there was still not that overwhelm

ing array of facts necessary in the case of a theory which is to be deliberately advocated.

move.

In particular, I recognised certain difficulties which no evidence available when I wrote enabled me to reFor instance, although it seemed reasonable to infer that matter propelled from the sun must be enormously retarded by his atmosphere, and that the rarer matter so erupted would be more retarded than the denser, yet there was no direct evidence of such retardation. All the prominences which had been regarded by Zöllner, Respighi, and others, as phenomena of eruption, had been observed when already formed, and the estimate of the velocity of ejection had been simply based on the height attained by the erupted matter. In no single instance had the actual upward motion of prominence matter been watched, measured, and timed. Again, the theory seemed to require that the eruption-prominences should be somewhat more markedly limited to the solar spot-zones than at the time appeared to be shown. Respighi had found the prominences on the spot-zone larger and better developed than those in the sun's equatorial or polar regions; yet prominences were not wanting in the last-named parts of the sun, and some of those seen there have been, as Respighi's pictures show, of no inconsiderable magnitude. Yet again, it seemed that if other matter, as the metallic elements, were propelled along with the glowing hydrogen of the prominences, the spectroscopic analysis of the prominences ought to reveal traces of the presence of such matter