Front matter

ROUGH WAYS MADE SMOOTH

A SERIES OF

Familiar Essays on Scientific Subjects

BY

RICHARD A. PROCTOR

NEW IMPRESSION

LONGMANS, GREEN, AND CO.

39 PATERNOSTER ROW, LONDON NEW YORK AND BOMBAY

1903

All rights reserved

'Let knowledge grow from more to more' Tennyson

PREFACE.

It is scarcely necessary for me to explain the plan of the present work, because I have already—in introducing my 'Light Science for Leisure Hours,' my 'Science Byways,' and my 'Pleasant Ways in Science'—described the method on which, as I think, such treatises as the present should be written. This work deals with similar subjects in a similar way; but I think the experience I have acquired in writing other works on the same plan has enabled me to avoid some defects in the present work which I have recognised in the others.

The list of subjects indicates sufficiently the range over which the present volume extends. Some of them might be judged by their names to be in no way connected with science, but it will be found that none have been treated except in their scientific significance, though in familiar and untechnical terms.

RICHARD A. PROCTOR.

S.S. 'Arizona,' Irish Sea October 18, 1879.

CONTENTS.

PAGE

THE SUN'S CORONA AND HIS SPOTS. 1

SUN-SPOTS AND COMMERCIAL PANICS. 26

NEW PLANETS NEAR THE SUN. 32

RESULTS OF THE BRITISH TRANSIT EXPEDITIONS. 58

THE PAST HISTORY OF OUR MOON. 81

A NEW CRATER IN THE MOON. 98

THE NOVEMBER METEORS. 111

EXPECTED METEOR SHOWER. 117

COLD WINTERS. 125

OXFORD AND CAMBRIDGE ROWING. 148

ROWING STYLES. 169

ARTIFICIAL SOMNAMBULISM. 178

HEREDITARY TRAITS. 205

BODILY ILLNESS AS A MENTAL STIMULANT. 236

DUAL CONSCIOUSNESS. 259

ELECTRIC LIGHTING. 289

ROUGH WAYS MADE SMOOTH.

THE SUN'S CORONA AND HIS SPOTS.

One of the most important results of observations made upon the eclipse of July 29, 1878, indicates the existence of a law of sympathy, so to speak, between the solar corona and the sun-spots. The inquiry into this relation seems to me likely to lead to a very interesting series of researches, from which may possibly result an interpretation not only of the relation itself, should it be found really to exist, but of the mystery of the sun-spot period. I speak of the sun-spot period as mysterious, because even if we admit (which I think we cannot do) that the sun-spots are produced in some way by the action of the planets upon the sun, it would still remain altogether a mystery how this action operated. When all the known facts respecting the sun-spots are carefully considered, no theory yet advanced respecting them seems at all satisfactory, while no approach even has been made to an explanation of their periodic increase and diminution in number. This seems to me one of the most interesting problems which astronomers have at present to deal with; nor do I despair of seeing it satisfactorily solved within no very long interval of time. Should the recognition of a sympathy between the corona and the sun-spots be satisfactorily established, an important step in advance will have been made,—possibly even the key to the enigma will be found to have been discovered.

I propose now to consider, first, whether the evidence we have on this subject is sufficient, and afterwards to discuss some of the ideas suggested by the relations which have been recognised as existing between the sun-spots, the sierra, the coloured prominences, and the zodiacal light.

The evidence from the recent eclipses indicates beyond all possibility of doubt or question, that during the years when sun-spots were numerous, in 1870 and 1871, the corona, at least on the days of the total solar eclipses in those years, presented an appearance entirely different from that of the corona seen on July 29, 1878, when the sun was almost free from spots. This will be more fully indicated further on. At present it is necessary to notice only (1) that whereas in 1870 and 1871 the inner corona extended at least 250,000 miles from the sun, it reached only to a height of some 70,000 miles in 1878; (2) in 1870 and 1871 it possessed a very complicated structure, whereas in 1878 the definite structure could be recognised only in two parts of the inner corona; (3) in 1871 the corona was pink, whereas in 1878 it was pearly white; (4) the corona was ten times brighter in 1871 than in 1878; lastly, in 1871 the light of the corona came in part from glowing gas, whereas in July, 1878, the light came chiefly, if not wholly, from glowing solid or liquid matter. I must here point out, that the evidence of change, however satisfactory in itself, would be quite insufficient to establish the general theory that the corona sympathises with the solar photosphere in the special manner suggested by the recent eclipse observations. There are few practices more unscientific, or more likely to lead to erroneous theorising, than that of basing a general theory on a small number of observations. In this case we have, in fact, but a single observed correspondence, though the observations establishing it form a series. It has been shown that so far as the special sun-spot period from the minimum of 1867 to the minimum of 1878 is concerned, there has been a certain correspondence between the aspect of the corona and the state of the sun's surface, with regard to spots. To assume from that single correspondence that the corona and the sun-spots are related in the same way, would be hazardous in the extreme. We may indeed find, when we consider other matters, that the probability of a general relation of this sort existing is so great antecedently, that but slight direct evidence would be required to establish the existence of the relation. But it must be remembered that before the eclipse of 1878 was observed, with the special result I have noticed, few were bold enough to assert the probable existence of any such relationship; and certainly no one asserted that the probability was very strong. I believe, indeed, that no one spoke more definitely in favour of the theory that the corona probably sympathises with the sun-spots than I did myself before the recent eclipse; but certainly I should not then have been willing to say that I considered the evidence very strong.

We must then look for evidence of a more satisfactory kind.

Now, although during the two centuries preceding the invention of the spectroscope and the initiation of the solar physical researches now in progress, observations of eclipses were not very carefully conducted, yet we have some records of the appearance of the corona on different occasions, which, combined with the known law of sun-spot periodicity, may enable us to generalise more safely than we could from observations during the present spot-period, though these observations have been far more exact than the older ones. I propose to examine some of these. Necessarily I must make some selection. I need hardly say that even if there were no such relation as that which seems to be indicated by recent observations, and if my purpose were simply to prove, either that such a relation exists or that it does not, I could very readily bring before the reader of these pages what would seem like the most satisfactory evidence that the relation is real. I must ask him to believe, however, that my purpose is to ascertain where the truth lies. I shall neither introduce any observation of the corona because it seems specially favourable to the theory that the corona sympathises with the photosphere, nor omit any, because it seems definitely opposed to that theory. To prevent any possibility of being unconsciously prejudiced, I shall take a series of coronal observations collected together by myself, on account of their intrinsic interest, several years ago, when I had not in my thoughts any theory respecting periodic changes in the corona—the series, namely, which is included in the sixth chapter of my treatise on the sun. Each of these observations I shall consider in connection with the known condition of the sun as to spots, and those results which seem to bear clearly, whether favourably or unfavourably, on the theory we are enquiring into, I shall bring before the reader.

Kepler, whose attention had been specially drawn to the subject of the light seen round the sun during total eclipse, by certain statements which Clavius had made respecting the eclipse of 1567, describes the eclipse of 1605 in the following terms:—'The whole body of the sun was completely covered for a short time, but around it there shone a brilliant light of a reddish hue and uniform breadth, which occupied a considerable portion of the heavens.' The corona thus seen may fairly be assumed to have resembled in extent that of 1871. A bright corona, reaching like that seen during the eclipse of July 1878 to a height of only about 70,000 miles from the sun's surface, would certainly not have been described by Kepler as occupying a considerable portion of the heavens, for a height of 70,000 miles would correspond only to about a twelfth of the sun's diameter; and a ring so narrow would be described very differently. It seems, then, that in 1605 a corona was seen which corresponded with that observed when the sun has had many spots on his surface. Now we have no record of the condition of the sun with regard to spots in 1605; but we know that the year 1615 was one of many spots, and the year 1610 one of few spots; whence we may conclude safely that the year 1605 was one of many spots. This case then is in favour of the theory we are examining.

In passing we may ask whether the observation by Clavius which had perplexed Kepler, may not throw some light on our subject. Clavius says that the eclipse of 1567 which should have been total was annular. The usual explanation of this has been that the corona was intensely bright close to the sun. And though Kepler considered that his own observation of a broad reddish corona satisfactorily removed Clavius's difficulty, it seems tolerably clear that the corona seen by Clavius must have been very unlike the corona seen by Kepler. In fact the former must have been like the corona seen in July, 1878, much smaller than the average, but correspondingly increased in lustre. Now with regard to the sun-spot period we can go back to the year 1567, though not quite so securely as we could wish. Taking the average sun-spot period at eleven years, and calculating back from the minimum of spots in the year 1610, we get four years of minimum solar disturbance, 1599, 1588, 1577, and 1566. We should have obtained the same result if we had used the more exact period, eleven one-ninth years, and had taken 1610·8 for the epoch of least solar disturbance (1610·8 meaning about the middle of October, 1610). Thus the year 1567 was a year of few sun-spots, probably occupying almost exactly the same position in the sun spot period as the year 1878. Clavius's observation, then, is in favour of our theory.

But another observation between Clavius's and Kepler's may here be noticed. Jensenius, who observed the eclipse of 1598 at Torgau in Germany, noticed that, at the time of mid-totality, a bright light shone round the moon. On this occasion, remarks Grant, the phenomenon was generally supposed to arise from a defect in the totality of the eclipse, though Kepler strenuously contended that such an explanation was at variance with the relation between the values of the apparent diameters of the sun and moon as computed for the time of the eclipse by aid of the solar and lunar tables. The corona, then, must have resembled that seen by Clavius, and since the year 1598 must have been very near the time of fewest spots, this observation accords with the theory we are examining.

The next observation is that made by Wyberd during the eclipse of 1652. Here there is a difficulty arising from the strange way in which the sun-spots behaved during the interval from 1645 to 1679. According to M. Wolf, whose investigation of the subject has been very close and searching, there was a maximum of sun-spots in 1639 followed by a minimum in 1645, the usual interval of about six years having elapsed; but there came a maximum in 1655, ten years later, followed by a minimum in 1666, eleven years later, so that actually twenty-one years would seem to have elapsed between successive minima (1645 and 1666). Then came a maximum in 1675, nine years later, and a minimum in 1679, four years later. Between the maxima of 1639 and 1675, including two spot periods, an interval of thirty-six years elapsed. There is no other instance on record, so far as I know, of so long an interval as this for two spot-periods. In passing, I would notice how little this circumstance accords with the theory that the sun-spots follow an exact law, or that from observations of the sun, means can ever be found for forming a trustworthy system of weather prediction, even if we assumed (which has always seemed to me a very daring assumption), that terrestrial weather is directly dependent on the progress of the sun-spot period. But here the irregularity of the spot changes affects us only as preventing us from determining or even from guessing what may have been the condition of the sun's surface in the year 1652. This year followed by seven years a period of minimum disturbance, and preceded by three years a period of maximum disturbance; but it would be unsafe to assume that the sun's condition in 1652 was nearer that of maximum than that of minimum disturbance. We must pass over Wyberd's observations of the corona in 1652, at least until some direct evidence as to the sun's condition shall have been obtained from the papers or writings of the observers of that year. I note only that Wyberd saw a corona of very limited extent, having indeed a height not half so great as that of many prominences which have been observed during recent eclipses. If the theory we are examining should be established beyond dispute, we should be led to infer that the year 1652 was in reality a year of minimum solar disturbance. Perhaps by throwing in such a minimum between 1645 and 1666, with of course a corresponding maximum, the wild irregularity of the sun-spot changes between 1645 and 1679 would be to some degree diminished.

We are now approaching times when more satisfactory observations were made upon the corona, and when also we have more complete records of the aspect of the sun's surface.

In 1706 Plantade and Capies saw a bright ring of white light extending round the eclipsed sun to a distance of about 85,000 miles, but merging into a fainter light, which extended no less than four degrees from the eclipsed sun, fading off insensibly until its light was lost in the obscure background of the sky. This corresponds unmistakably with such a corona as we should expect only to see at a time of many sun-spots, if the theory we are examining is sound. Turning to Wolf's list, we find that the year 1705 is marked as a year of maximum solar disturbance, and the year 1712 as that of the next minimum. Therefore 1706 was a year of many sun-spots—in fact, 1706 may have been the year of actual maximum disturbance, for it is within the limits of doubt indicated by Wolf. Certainly a corona extending so far as that which Plantade and Capies saw would imply an altogether exceptional degree of solar disturbance, if the theory we are considering is correct.

In 1715 Halley gave the following description of the corona:—'A few seconds before the sun was all hid, there discovered itself round the moon a luminous ring about a digit' (a twelfth) 'or perhaps a tenth part of the moon's diameter in breadth. It was of a pale whiteness or rather pearl colour, seeming to me a little tinged with the colours of the Iris, and to be concentric with the moon.' He added that the ring appeared much whiter and brighter near the body of the moon than at a distance from it, and that its exterior boundary was very ill-defined, seeming to be determined only by the extreme rarity of the luminous matter. The French astronomer Louville gave a similar account of the appearance of the ring. He added, however, that 'there were interruptions in its brightness, causing it to resemble the radial glory with which painters encircle the heads of the saints.' The smallness of the corona on this occasion corresponds with the description of the corona seen in July 1878; and though Louville's description of gaps is suggestive of a somewhat different aspect, yet, on the whole, the corona seen in 1715 more closely resembles one which would be seen at a time of minimum solar disturbance, if our theory can be trusted, than one which would be seen at a time of maximum disturbance. Wolf's list puts the year 1712 as one of minimum disturbance, with one year of doubt either way, and the middle of the year 1817 as the epoch of maximum disturbance, with a similar range of uncertainty. The case, then, is doubtful, but on the whole inclines to being unfavourable. I may remark that because of its unfavourable nature, I departed from the rule I had set myself, of taking only the cases included in my treatise on the sun. For the corona of 1715 is not described in that treatise, as indeed affording no evidence respecting this solar appendage. The evidence given in this case is probably affected in some degree by the unfavourable atmospheric conditions under which Halley certainly, and Louville probably, observed the eclipse. In any case the evidence is not strong; only I would call attention here to the circumstance that if, as we proceed, we should come to a case in which the evidence is plainly against the theory we are examining, we must give up the theory at once. For one case of discordance does more to destroy a theory respecting association between such and such phenomena, than a hundred cases of agreement would do in the way of confirming it.

In 1724, Maraldi noticed that the corona was broadest first on the side towards which the moon was advancing, and afterwards on the side which the moon was leaving. From this we may infer that the corona was only a narrow ring on that occasion, since otherwise the slight difference of breadth due to the moon's eccentric position at the beginning and end of totality would not have been noticeable. Now, the year 1723 was one of minimum disturbance, with one year of doubt either way. Thus 1724 was certainly a year of few sun-spots, and may have been the actual year of minimum disturbance. The corona then presented an appearance according with the theory we are considering.

Few eclipses have been better observed than that of the year 1733. The Royal Society of Sweden invited all who could spare the time to assist, as far as their ability permitted, in recording the phenomena presented during totality. The pastor of Stona Malm states that at Catherinesholm, there was a ring around the sun about 70,000 miles in height. (Of course these are not his exact words; what he actually stated was that the ring was about a digit in breadth.) This is the exact height assigned to the coronal ring by the observers of the eclipse of last year. The ring seemed to be of a reddish colour. Another clergyman, Vallerius, states also that the ring was of this colour, but adds that at a considerable distance from the sun it had a greenish hue. This suggests the idea that the outer corona was seen also by Vallerius, and that it had considerable breadth. The reddish colour of the inner light portion would correspond to the colour it would have if it consisted in the main of glowing hydrogen. If that really was its constitution, then the theory advanced by one observer of the last eclipse, that at the time of minimum solar disturbance the glowing hydrogen is withdrawn from the corona, would be shown to be incorrect. For 1733 was the actual year of minimum solar disturbance. The pastor of Smoland states that 'during the total obscuration the edge of the moon's disc resembled gilded brass, and the faint ring round it emitted rays in an upward as well as in a downward direction, similar to those seen beneath the sun when a shower of rain is impending.' The mathematical lecturer of the Academy of Charles-stadt, M. Edstrom, observed these rays with special attention: he says that 'they plainly maintained the same position, until they vanished along with the ring upon the re-appearance of the sun.' On the other hand, at Lincopia no rays were seen. On the whole it seems clear from the accounts of this eclipse that the inner corona was bright and narrow; rays issued from the outer faint ring; but they were very delicate phenomena, easily concealed by atmospheric haze, and thus were not everywhere observed. As rays were seen in July 1878, there is nothing in the evidence afforded by the eclipse of 1733, occurring at a time of few spots, which opposes itself definitely to the theory we are considering. But the reddish colour of the corona as already noticed is a doubtful feature: in July, 1878, the bright inner corona was of a pearl colour and lustre.

During the eclipse of February, 1766, the corona presented four luminous expansions, and seems to have presented a greater expansion than we should expect in a year of minimum solar disturbance. Such, however, the year 1766 certainly was. The evidence in this case is unfavourable to our theory—not quite decisively so, but strongly. For we should expect that in the year of actual minimum disturbance the corona would be even narrower than in the year 1878, which was the year following that of least disturbance. And again, a strongly distinctive feature in the corona of July, 1878, was the absence of wide expansions, such as were seen in 1870 and 1871. Now if this peculiarity should really be attributed to the relation existing between the corona and the sun-spots, we should infer that in 1766 the corona would have been still more markedly uniform in shape. The existence of four well marked expansions on that occasion forces us to assume that either the relation referred to has no real existence, or else that the corona may change from week to week as the condition of the sun's surface changes, and that in February, 1766, the sun was temporarily disturbed, though the year, as a whole, was one of minimum disturbance. But as the epoch of actual minimum was the middle of 1766, February 1766 should have been a time of very slight disturbance. I do not know of any observations of the sun recorded for the month of February, 1766. On the whole, the eclipse of 1766 must be regarded as throwing grave doubt on the relation assumed by our theory as existing between the corona and the sun-spots; and as tending to suggest that some wider law must be in question than the one we have been considering—if any association really exists.

The account given by Don Antonio d'Ulloa of the appearance presented by the corona during the total eclipse of 1778, is rendered doubtful by his reference to an apparent rotatory motion of the normal rays. He says that about five or six seconds after totality had begun, a brilliant luminous ring was seen around the dark body of the moon. The ring became brighter as the middle of totality approached. 'About the middle of the eclipse, the breadth of the ring was equal to about a sixth of the moon's diameter. There seemed to issue from it a great number of rays of unequal length, which could be discerned to a distance equal to the moon's diameter.' Then comes the part of d'Ulloa's description which seems difficult to accept. He says that the corona 'seemed to be endued with a rapid rotatory motion, which caused it to resemble a firework turning round its centre.' The colour of the light, he proceeds, 'was not uniform throughout the whole breadth of the ring. Towards the margin of the moon's disc it appeared of a reddish hue; then it changed to a pale yellow, and from the middle to the outer border the yellow gradually became fainter, until at length it seemed almost quite white.' Setting aside the rays and their rotation, d'Ulloa's account of the inner corona may be accepted as satisfactory. The height of this ring was, it seems, about 140,000 miles, or twice that of the ring seen in July 1878. As the year 1779 was one of maximum solar disturbance, there were doubtless many spots in 1778; and the aspect of the corona accorded well with the theory that the corona expands as the number of sun-spots increases.

We come now to three eclipses which are especially interesting as having been all carefully observed, some observers having seen all three,—the eclipses, namely, of 1842, 1851, and 1860. Unfortunately the eclipses of 1842 and 1851 occurred when the sun-spots were neither at their greatest nor at their least degree of frequency. For a maximum of sun-spots occurred in 1837, and a minimum in 1844, so that 1842 was on what may be called the descending slope of a sun-spot wave, nearer the hollow than the crest, but not very near either: again, a maximum occurred in 1848, and a minimum in 1856, so that 1851 was also on the descending slope of a sun-spot wave, rather nearer the crest than the hollow, but one may fairly say about midway between them. Still it is essential in an inquiry of this sort to consider intermediate cases. We must not only apply the comparentia ad intellectum instantiarum convenientium, but also the comparentia instantiarum secundum magis ac minus. If the existence of great solar disturbances causes the corona to be greatly enlarged, as compared with the corona seen when the sun shows no spots, we should expect to find the corona moderately enlarged only when the sun shows a considerable but not the maximum number of spots. And again, it is conceivable that we may find some noteworthy difference between the aspect of the corona when sun-spots are diminishing in number, and its aspect when they are increasing. This point seems the more to need investigation when we note that the evidence derived from eclipses occurring near the time either of maximum or of minimum solar disturbance has not been altogether satisfactory. It may be that we may find an explanation of the discrepancies we have recognised, in some distinction between the state of the corona when spots are increasing and when they are diminishing in number.

It is noteworthy that several careful observers of the corona in 1842 believed that they could recognise motion in the coronal rays. Francis Baily compared the appearance of the corona to the flickering light of a gas illumination. O. Struve also was much struck by the appearance of violent agitation in the light of the ring. It seems probable that the appearance was due to movements in that part of our atmosphere through which the corona was observed. The extent of the corona was variously estimated by different observers. Petit, at Montpelier, assigned to it a breadth corresponding to a height of about 200,000 miles; Baily a height of about 500,000 miles; and O. Struve a height of more than 800,000 miles. The last-named observer also recognised luminous expansions extending fully four degrees (corresponding to nearly seven million miles) from the sun. Picozzi, at Milan, noticed two jets of light, which were seen also by observers in France. Rays also were seen by Mauvais at Perpignan, and by Baily at Paria. But Airy, observing the corona from the Superga, could see no radiation; he says 'although a slight radiation might have been perceptible, it was not sufficiently intense to affect in a sensible degree the annular structure by which the luminous appearance was plainly distinguished.' These varieties in the aspect of the corona were doubtless due to varieties in the condition of the atmosphere through which the corona was seen. Now it cannot be questioned that, so far as extension is concerned, the corona seen in 1842 was one which, if the theory we are considering were sound, we should expect to see near the time of maximum rather than of minimum solar disturbance. On the other hand, in brightness the corona of 1842 resembled, if it did not surpass, that of July 1878.

'I had imagined,' says Baily, 'that the corona, as to its brilliant or luminous appearance, would not be greater than that faint crepuscular light which sometimes takes place (sic) in a summer evening, and that it would encircle the moon like a ring. I was therefore somewhat surprised and astonished at the splendid scene which now so suddenly burst upon my view.'

The light of the corona was so bright, O. Struve states, that the naked eye could scarcely endure it; 'many could not believe, indeed, that the eclipse was total, so strongly did the corona's light resemble direct sunlight.' Thus while as to extent the corona in 1842 presented the appearance to be expected at the time of maximum solar disturbance, if our theory is sound, its brightness was that corresponding to a time of minimum disturbance. Its structure corresponded with the former condition. The light of the corona was not uniform, nor merely marked by radiations, but in several places interlacing lines of light could be seen. Arago, at Perpignan, observed with the unaided eye a region of the corona where the structure was as of intertwined jets giving an appearance resembling a hank of thread in disorder.

Certainly, for an eclipse occurring two years from the time of minimum, and five years from the time of maximum disturbance, that of July, 1842,[1] has not supplied evidence favouring the theory with which we started. Whether any other theory of association between the corona and the sun-spots will better accord with the evidence hitherto collected remains to be seen.

Turn we now to the eclipse of 1851, occurring nearly midway between the epochs of maximum solar disturbance (1848) and minimum solar disturbance (1856). I take the account given by Airy, our Government astronomer, as he was one of the observers of the eclipse of 1842.

'The corona was far broader,' he says, 'than that which I saw in 1842. Roughly speaking, the breadth was little less than the moon's diameter, but its outline was very irregular. I did not notice any beams projecting from it which deserved notice as much more conspicuous than the others; but the whole was beamy, radiated in structure, and terminated—though very indefinitely—in a way which reminded me of the ornament frequently placed round a mariner's compass. Its colour was white, or resembling that of Venus. I saw no flickering or unsteadiness of light. It was not separated from the moon by any interval, nor had it any annular structure. It looked like a radiated luminous cloud behind the moon.'

The corona thus described belongs to that which our theory associates with the period of maximum rather than of minimum solar disturbance. Definite peculiarities of structure seem to have been more numerous and better marked than in 1842. It accords with our theory that 1851 was a year of greater solar disturbance than was observed in 1842, as the following numbers show:—

Days of observation Days without spots New groups observed

1842 307 64 68

1851 308 0 141

1860 332 0 211

I have included the year 1860, as we now proceed to consider the corona then seen by Airy. The year 1860 did not differ very markedly, it will be observed, from 1851, as regards the number of new groups of spots observed by Schwabe, especially when account is taken of the number of days in which the sun was observed in these two years. But 1860 was a year of maximum solar disturbance, whereas 1851 was not.[2]

Airy remarks of the corona in 1860:—'It gave a considerable body, but I did not remark either by eye-view or by telescope-view anything annular in its structure; it appeared to me to resemble, with some irregularities (as I stated in 1851), the ornament round a compass-card.'

Bruhns of Leipsic noted that the corona shone with an intense white light, so lustrous as to dim the protuberances. He noticed that a ray shot out to a distance of about one degree indicating a distance of at least 1,500,000 miles from the sun's surface. This was unquestionably a coronal appendage as neither the direction nor the length of the ray varied for ten seconds, during which Bruhns directed his attention to it. Its light was considerably feebler than that of the corona, which was of a glowing white, and seemed to coruscate or twinkle. Bruhns assigned to the inner corona a height varying from about 40,000 to about 80,000 miles. But this was unquestionably far short of the true height. In fact, Secchi's photographs show the corona extending to a distance of at least 175,000 miles from the surface of the sun. Therefore probably what Bruhns calls the base of the corona was in reality only the prominence region, and the inner corona was that which he describes as varying in breadth or height from nearly one-half to a quarter of a degree—that is from about 800,000 to about 400,000 miles. De la Rue gives a somewhat similar general description of the corona seen in 1860. He remarks that it was extremely bright near the moon's body, and of a silvery whiteness. The picture of the corona by Feilitsch (given at p. 343 of my book on the Sun) accords with these descriptions.

On the whole, the eclipse of 1860 affords evidence according well with the theory we have been considering, except as regards the brightness and the colour of the corona, which correspond more closely with what was observed in July, 1878, with the lustre and colour of the corona in 1870 and 1871. In this respect, it is singular that the eclipse of 1867, which occurred (see preceding note) when the sun spots were fewer in number, presented a decided contrast to that of 1860,—the contrast being, however, precisely the reverse of that which our theory would require, if the colour and brightness of the corona be considered essential features of any law of association.

Herr Grosch, describing the corona of 1867, says, 'There appeared around the moon a reddish glimmering light similar to that of the aurora, and almost simultaneously with this (I mean very shortly after it) the corona.' It is clear, however, from what follows, that the reddish light was what is now commonly called the inner corona, which last July, when the sun was in almost exactly the same condition as regards the spots, was pearly white and intensely bright. 'This reddish glimmer,' he proceeds, 'which surrounded the moon with a border of the breadth of at most five minutes' (about 140,000 miles) 'was not sharply bounded in any part, but was extremely diffused and less distinct in the neighbourhood of the poles.' Of the outer corona he remarks that 'its apparent height amounted to about 280,000 miles opposite the solar poles, but opposite the polar equator to about 670,000 miles. Its light was white. This white light was not in the least radiated itself, but it had the appearance of rays penetrating through it; or rather as if rays ran over it, forming symmetrical pencils diverging outwards, and passing far beyond the boundary of the white light. These rays had a more bluish appearance, and might best be compared to those produced by a great electro-magnetic light. Their similarity to these, indeed, was so striking, that under other circumstances I should have taken them for such, shining at a great distance. The view of the corona I have described is that seen with the naked eye.... In the white light of the corona, close upon the moon's edge, there appeared several dark curves. They were symmetrically arched towards the east and west, sharply drawn, and resembling in tint lines drawn with a lead pencil upon white paper.... Beginning at a distance of one minute (about 26,000 miles), they could be traced up to a distance of about nine minutes (some 236,000 miles) from the moon's edge.'

Almost all the features observed in this case correspond closely with those noted and photographed during the eclipse of December, 1871. In other words the corona seen in 1867, when the sun was passing through the period of least solar disturbance, closely resembled the corona seen in 1871, when the sun was nearly in its stage of greatest disturbance. Even the spectroscopic evidence obtained in 1871 and July, 1878, may be so extended as to show with extreme probability what would have been seen in 1867 if spectroscopic analysis had then been applied. We cannot doubt that the reddish inner corona, extending to a height of about 140,000 miles, would have been found under spectroscopic analysis to shine in part with the light of glowing hydrogen, as the reddish corona of 1871 did. The white corona of July, 1878, on the contrary, shone only with such light as comes from glowing solid or liquid matter. Here then, again, the evidence is unfavourable to our theory; for the corona in 1867 should have closely resembled the corona of 1878, if this theory were sound.

It would be idle, I think, to seek for farther evidence either in favour of the theory we originally proposed to discuss, or against it: for the evidence of the eclipse of 1867 disposes finally of the theory in that form. I may note in passing that the eclipse of 1868 gave evidence almost equally unfavourable to the theory, while the evidence given by the eclipse of 1869 was neutral. It will be desirable, however, to consider, before concluding our inquiry, the evidence obtained in 1871 and last July, in order that we may see what, after all, that evidence may be regarded as fairly proving with regard to coronal variations.

First, however, as I have considered two eclipses which occurred when the sun spots were decreasing in number—namely, those of 1842 and 1851, midway (roughly speaking) between the crest and hollow of the sun-spot wave on its descending slope, it may be well to consider an eclipse which was similarly situated with respect to the ascending slope of a sun-spot wave. I take, then, the eclipse of 1858, as seen in Brazil by Liais. The picture drawn by this observer is one of the most remarkable views of the corona ever obtained. It is given at p. 339 of my book on the Sun. Formerly it was the custom to deride this drawing, but since the eclipse of 1871, when the corona was photographed, it has been admitted that Liais's drawing may be accepted as thoroughly trustworthy. It shows a wonderfully complex corona, like that of 1871, extending some 700,000 miles from the sun, and corresponding in all respects with such a corona as our theory (if established) would have associated with the stage of maximum solar disturbance. As in this respect the eclipse of 1858, when sun-spots were increasing, resembled those of 1842 and 1851, when sun-spots were diminishing in number, we find no trace of any law of association depending on the rate of increase or diminution of solar disturbance.

If we limited our attention to the eclipses of 1871 and of July, 1878, we should unquestionably be led to adopt the belief that the corona during a year of many spots differs markedly from the corona when the sun shows few spots, or none. So far as the aspect of the corona is concerned, I take the description given by the same observer in both cases, as the comparison is thus freed as far as possible from the effect of personal differences.

Mr. Lockyer recognised in 1871 a corona resembling a star-like decoration, with its rays arranged almost symmetrically—three above and three below two dark spaces or rifts at the extremity of a horizontal diameter. The rays were built up of innumerable bright lines of different length, with more or less dark spaces between them. Near the sun this structure was lost in the brightness of the central ring, or inner corona. In the telescope he saw thousands of interlacing filaments, varying in intensity. The rays so definite to the eye were not seen in the telescope. The complex structure of interlacing filaments could be traced only to a height of some five or six minutes (from 135,000 to 165,000 miles) from the sun, there dying out suddenly. The spectroscope showed that the inner corona, to this height at least (but Respighi's spectroscopic observations prove the same for a much greater distance from the sun), was formed in part of glowing gas—hydrogen—and the vapour of some as yet undetermined substance, shining with light of a green tint, corresponding to 1474 of Kirchhoff's scale. But also a part of the coronal light came from matter which reflected sunlight; for its spectrum was the rainbow-tinted streak crossed by dark lines, which we obtain from any object illuminated by the sun's rays. It should be added that the photographs of the corona in 1871 show the three great rays above and three below, forming the appearance as of a star-like decoration, described by Mr. Lockyer; insomuch as it is rather strange to find Mr. Lockyer remarking that 'the difference between the photographic and the visible corona came out strongly, ... and the non solar origin of the radial structure was conclusively established.' The resemblance is, indeed, not indicated in the rough copy of the photographs which illustrates Mr. Lockyer's paper; but it is clearly seen in the photographs themselves, and in the fine engraving which has been formed from them for the illustration of the volume which the Astronomical Society proposes to issue (some time in the present century, perhaps).

Now, in July, 1878, the corona presented an entirely different appearance. Mr. Lockyer, in a telegram sent to the Daily News, describes it as small, of pearly lustre, and having indications of definite structure in two places only. Several long rays were seen; but the inner corona was estimated as extending to a height of about 70,000 miles from the sun's surface. The most remarkable change, however, was that which had taken place in the character of the corona's spectrum—or, in other words, in the physical structure of the corona. The bright lines or bright images of the inner corona (according as it was examined through a slit or without one) were not seen in July, 1878, showing that no part, or at least no appreciable part, of its light came from glowing gaseous matter. But also the dark lines seen by Janssen in 1871 were wanting on this occasion, showing that the corona did not shine appreciably by reflecting sunlight. The spectrum was, in fine, a continuous rainbow-tinted streak, such as that given by glowing solid or liquid matter.

The inference clearly is: 1. That in July, 1878, the gaseous matter which had been present in the corona in 1871 was either entirely absent or greatly reduced in quantity; 2. The particles of solid or liquid (but probably solid) matter which, by reflecting sunlight, produced a considerable portion of the corona's light in 1871, were glowing with heat in July, 1878, and shone in the main with this inherent light; and 3. The entire corona was greatly reduced in size in July, 1878, as compared with that which formed the 'star-like decoration' around the black body of the moon in December, 1871.

We cannot, however, accept the theory that such a corona as was seen in 1871 invariably surrounds the sun in years of great disturbance, while the corona of last month is the typical corona for years of small solar disturbance. The generalisation is flatly contradicted by the evidence which I have presented in the preceding pages. It may be that such a corona as was seen in 1871 is common in years of great disturbance, just as spots are then more common, though not always present; while such a corona as was seen in July, 1878, is more common in years of small disturbance, just as days when the sun is wholly without spots are then more common, though from time to time several spots, and sometimes very large spots, are seen in such years. On the whole, I think the evidence I have collected favours rather strongly the inference that an association of this sort really exists between the corona and the sun-spots. It would, however, be unsafe at present to generalise even to this extent; while certainly the wide generalisation telegraphed to Europe from America as the great result of the eclipse observations in July, 1878, must unhesitatingly be rejected.

It remains to be considered how science may hope to obtain more trustworthy evidence than we yet have respecting the corona and its changes of form, extent, lustre, and physical constitution. In the case of the prominences, we have the means of making systematic observations on every fine, clear day. It has been, indeed, through observations thus effected by the spectroscopic method that an association has been recognised between the number, size, and brilliancy of the prominences on the one hand, and the number, size, and activity of the sun-spots on the other. But in the case of the corona, we are as yet unable to make any observations except at the time of total solar eclipse. It seems almost impossible to hope that any means can be devised for seeing the corona at any other time. Of course, without the aid of the spectroscope the corona, as ordinarily seen during total eclipses, must be entirely invisible when the sun is shining in full splendour. No one acquainted with even the merest elements of optics could hope to see the corona with an ordinary telescope at such a time. The spectroscope, again, would not help in the slightest degree to show such a corona as was shining in July, 1878. For the power of the spectroscope to show objects which under ordinary conditions are invisible, depends on the separation of rays of certain tints from the rays of all the colours of the rainbow, which make up solar light; and as the corona in July, 1878, shone with all the colours of the rainbow, and not with certain special tints, the power of the spectroscope would be thrown away on a corona of that kind. All that we can ever hope to do is to discern the gaseous corona when, as in 1871, it is well developed, by spectroscopic appliances more effective for that purpose than any which have hitherto been adopted; for all which have as yet been adopted have failed.

Now, the difficulty of the problem will be recognised when we remember that the strongest tints of the corona's light—the green tint classified as 1474 Kirchhoff—has been specially but ineffectually searched for in the sun's neighbourhood with the most powerful spectroscopic appliances yet employed in the study of the coloured prominences. In other words, when the light of our own air over the region occupied by the corona has been diluted as far as possible by spectroscopic contrivances, the strongest of the special coronal tints has yet failed to show through the diluted spectrum of the sky. Again, we have even stronger evidence of the difficulty of the task in the spectroscopic observations made by Respighi during the eclipse of 1871. The instrument, or I should rather, perhaps, say the arrangement, which during mid totality showed the green image of the corona to a height of about 280,000 miles, did not show any green ring at all at the beginning of totality. In other words, so faint is the light of the gaseous corona, even at its brightest part, close to the sun, that the faint residual atmospheric light which illuminates the sky over the eclipsed sun at the beginning of totality sufficed to obliterate this part of the coronal light.

Whether with any combination specially directed to meet the difficulties of this observation, the gaseous corona can be rendered discernible, remains to be seen. I must confess my own hopes that the problem will ever be successfully dealt with are very slight, though not absolutely evanescent. It seems to me barely possible that the problem might be successfully attacked in the following way. Using a telescope of small size, for the larger the telescope the fainter is the image (because of greater loss of light by absorption), let the image of the sun be received in a small, perfectly darkened camera attached to the eye-end of the telescope. Now if the image of the sun were received on a smooth white surface we know that the prominences and the corona would not be visible. And again, if the part of such a surface on which the image of the sun itself fell were exactly removed, we know (the experiment has been tried by Airy) that the prominences would not be seen on the ring of white surface left after such excision. Still less, then, would the much fainter image of the corona be seen. But if this ring of white surface, illuminated in reality by the sky, by the ring of prominences and sierra, and by the corona, were examined through a battery of prisms (used without a slit) adjusted to any one of the known prominence tints, the ring of prominences and sierra would be seen in that special tint. If the battery of prisms were sufficiently effective, and the tint were one of the hydrogen tints—preferably, perhaps, the red—we might possibly be able to trace the faint image of the corona in that tint. But we should have a better chance with the green tint corresponding to the spectral line 1474 Kirchhoff. If the ring of white surface were replaced by a ring of green surface, the tint being as nearly that of 1474 Kirchhoff as possible, the chance of seeing the coronal ring in that tint would be somewhat increased; and, still further, perhaps, if the field of view were examined through green glass of the same tint. It seems just possible that if prisms of triple height were used, through which the rays were carried three times, by an obvious modification of the usual arrangement for altering the level of the rays, thus giving a power of eighteen flint glass prisms of sixty degrees each, evidence, though slight perhaps, might be obtained of the presence of the substance which produces the green line. Thus variations in the condition of the corona might be recognised, and any law affecting such variations might be detected. I must confess, however, that a consideration of the optical relations involved in the problem leads me to regard the attempt to recognise any traces of the corona when the sun is not eclipsed as almost hopeless.

It is clear that until some method for thus observing the corona has been devised, future eclipse observations will acquire a new interest from the light which they may throw on the coronal variations, and their possible association in some way, not as yet detected, with the sun-spot period. Even when a method has been devised for observing the gaseous corona, the corona whose light comes either directly or by reflection from solid or liquid matter will still remain undiscernible save only during total eclipses of the sun. Many years must doubtless pass, then, before the relation of the corona to the prominences and the sun-spots shall be fully recognised. But there can be no question that the solution of this problem will be well worth waiting for, even though it should not lead up (as it most probably will) to the solution of the mystery of the periodic changes which affect the surface of the sun.

FOOTNOTES:

SUN-SPOTS AND COMMERCIAL PANICS.

We are not only, it would seem, to regard the sun as the ultimate source of all forms of terrestrial energy, existent or potential, but as regulating in a much more special manner the progress of mundane events. Many years have passed since Sabine, Wolf, and Gauthier asserted that variations in the daily oscillations of the magnetic needle appear to synchronise with the changes taking place in the sun's condition, the oscillations attaining their maximum average range in years when the sun shows most spots, and their minimum range when there are fewest spots. And although it is well known that the Astronomer Royal in England and the President of the Academy of Sciences in France reject this doctrine, it still remains in vogue. True, the average magnetic period appears to be about 10.45 years, while Wolf obtains for the sun-spot period 11.11 years; but believers in the connection between terrestrial magnetic disturbances and sun-spots consider that among the imperfect records of the past condition of the sun Wolf must have lost sight of one particular wave of sun-spots, so to speak. If there have been 24 such waves between 1611 and 1877, when sun-spots were fewest, we get Wolf's period of 11.11 years; if there have been 25 such waves then, taking an admissible estimate for the earliest epoch, we get 10.45 years, the period required to synchronise with the period of terrestrial magnetic changes. The matter must be regarded as still sub judice. This, however, is only one relation out of many now suggested. Displays of the aurora, being unquestionably dependent on the magnetic condition of the earth, would of course be associated with the sun spot period, if the magnetic period is so; and certainly the most remarkable displays of the aurora in recent times have occurred when the sun has shown many spots. Yet this of itself proves nothing more than had been already known—namely, that the last few magnetic periods have nearly synchronised with the last few sun-spot periods. It is rather strange, too, that no auroras are mentioned in the English records for 80 years preceding the aurora of 1716, and in the records of the Paris Academy of Sciences one only—that of 1666, which occurred when sun-spots were fewest. The great aurora of 1723, seen as far south as Bologna, also occurred at the time of minimum solar activity. Here we are not depending on either Wolf's period of 11 years or Brown's of 10½ years; from records of actual observation it is known that in 1666 and 1713 there were no sun-spots. In fact it is worth mentioning that Cassini, writing in 1671, says, 'It is now about 20 years since astronomers have seen any considerable spots on the sun,' a circumstance which throws grave doubt on the law of sun-spot periodicity itself. It is at least certain that the interval from maximum, spot-frequency to maximum, or from minimum to minimum, has sometimes fallen far short of 9 years, and has at others exceeded 18 years.

It appears again that certain meteorological phenomena show a tendency, more or less marked, to run through a ten-year cycle. Thus, from the records of rainfall kept at Oxford it appears that more rain fell under west and south-west winds when sun-spots were largest and most numerous than under south and south-east winds, these last being the more rainy winds when sun-spots were least in size and fewest in number. This is a somewhat recondite relation, and at least proves that earnest search has been made for such cyclic relations as we are considering. But this is not all. When other records were examined, the striking circumstance was discovered that elsewhere, as at St. Petersburg, the state of things observed at Oxford was precisely reversed. At some intermediate point between Oxford and St. Petersburg, no doubt the rainfall under the winds named was equally distributed throughout the spot period. Moreover, as the conditions thus differ at different places, we may assume that they differ also at different times. Such relations appear then to be not only recondite, but complicated.

When we learn that during nearly two entire sun-spot periods cyclones have been somewhat more numerous in the Indian Seas when spots are most numerous than when the sun is without spots, and vice versâ, we recognise the possible existence of cyclic relations better worth knowing than those heretofore mentioned. The evidence is not absolutely decisive; some, indeed, regard it as scarcely trustworthy. Yet there does seem to have been an excess of cyclonic disturbance during the last two periods of great solar disturbance, precisely as there was also an excess of magnetic disturbance during those periods. The excess was scarcely sufficient, however, to justify the rather daring statement made by one observer, that 'the whole question of cyclones is merely a question of solar activity.' We had records of some very remarkable cyclonic disturbances during the years 1876 and 1877, when the sun showed very few spots, the actual minimum of disturbance having probably been reached late in 1877. A prediction that 1877 would be a year of few and slight storms would have proved disastrous if implicit reliance had been placed on it by seamen and travellers.

Rainfall and atmospheric pressure in India have been found to vary in a cyclic manner, of late years at any rate, the periods being generally about 10 or 11 years. The activity of the sun, as shown by the existence of many spots, apparently makes more rainfall at Madras, Najpore, and some other places; while at Calcutta, Bombay, Mysore, and elsewhere it produces a contrary effect. Yet these effects are produced in a somewhat capricious way: for sometimes the year of actual maximum spot frequency is one in which rainfall is below the average (instead of above) at the former stations, and above the average (instead of below) at the latter. It is only by taking averages—and in a somewhat artificial manner—that the relation seems to be indicated on which stress has been laid.

Since Indian famines are directly dependent on defective rainfall, it is natural that during the years over which observation has hitherto extended the connection apparently existing between sun-spots and Indian rainfall should seem also to extend itself to Indian famines. It was equally to be expected that since cyclones have been rather more numerous, for some time past, in years when sun-spots have been most numerous, shipwrecks should also have been somewhat more frequent in such years. Two years ago Mr. Jeula gave some evidence which, in his opinion, indicated such a connection between sun-spots and shipwrecks. He showed that in the four years of fewest spots the mean percentage of losses was 8.64; in four intermediate years the mean percentage was 9.21; in three remaining years of the eleven-year cycle—that is, in three years of greatest spot frequency the mean percentage was 9.53. Some suggested that possibly such events as the American war, which included two of the three years of greatest spot frequency, may have had more effect than sun-spots in increasing the percentage of ships lost; while perhaps, the depression following the commercial panic of 1866 (at a time of fewest sun-spots) may have been almost as effective in reducing the percentage of losses as the diminished area of solar maculation. But others consider that we ought rather to regard the American war as yet another product of the sun's increased activity in 1860-61, and the great commercial panic of 1866 as directly resulting from diminished sun-spots at that time, thus obtaining fresh evidence of the sun's specific influence on terrestrial phenomena instead of explaining away the evidence derived from Lloyd's list of losses.

This leads us to the last, and, in some respects, the most singular suggestion respecting solar influence on mundane events—the idea, namely, that commercial crises synchronise with the sun-spot period, occurring near the time when spots are least in size and fewest in number; or, as Professor Jevons (to whom the definite enunciation of this theory is due) poetically presents the matter, that from 'the sun, which is truly "of this great world both eye and soul," we derive our strength and our weakness, our success and our failure, our elation in commercial mania, and our despondency and ruin in commercial collapse.' We have better opportunities of dealing with this theory than with the others, for we have records of commercial matters extending as far back as the beginning of the eighteenth century. In fact, we have better evidence than Professor Jevons seems to have supposed, for whereas in his discussion of the matter he considers only the probable average of the sun-spot period, we know approximately the epochs themselves at which the maxima and minima of sun spots have occurred since the year 1700. The evidence as presented by Professor Jevons is very striking, though when examined in detail it is rather disappointing. He presents the whole series of decennial crises as follows:—1701? (such query marks are his own), 1711, 1721, 1731-32, 1742 (?), 1752 (?), 1763, 1772-73, 1783, 1793, 1804-5 (?), 1815, 1825, 1836-9 (1837 in the United States), 1847, 1857, 1866 and 1878. The average interval comes out 10.466 years, showing, as Jevons points out, 'almost perfect coincidence with Brown's estimate of the average sun-spot period.' Let us see, however, whether these dates correspond so closely with the years of minimum spot-frequency as to remove all doubt. Taking 5¼ years as the average interval between maximum and minimum sun-spot frequency, we should like to find every crisis occurring within a year or so on either side of the minimum though we should prefer perhaps to find the crisis always following the time of fewest sun-spots, as this would more directly show the depressing effect of a spotless sun. No crisis ought to occur within a year or so of maximum solar disturbance; for that, it should seem, would be fatal to the suggested theory. Taking the commercial crises in order, and comparing them with the known (or approximately known) epochs of maximum and minimum spot frequency, we obtain the following results (we italicize numbers or results unfavourable to the theory):—The doubtful crisis of 1701 followed a spot minimum by three years; the crisis of 1711 preceded a minimum by one year; that of 1721 preceded a minimum by two years; 1731-32, preceded a minimum by one year; 1742 preceded a minimum by three years; 1752 followed a maximum by two years; 1763 followed a maximum by a year and a half; 1772-73 came midway between a maximum and a minimum; 1783 preceded a minimum by nearly two years; 1793 came nearly midway between a maximum and a minimum; 1804-5 coincided with a maximum; 1815 preceded a maximum by two years; 1825 followed a minimum by two years; 1836-39 included the year 1837 of maximum solar activity (that year being the time also when a commercial crisis occurred in the United States); 1847 preceded a maximum by a year and a half; 1866 preceded a minimum by a year; and 1878 followed a minimum by a year. Four favourable cases out of 17 can hardly be considered convincing. If we include cases lying within two years of a minimum, the favourable cases mount up to seven, leaving ten unfavourable ones. It must be remembered, too, that a single decidedly unfavourable case (as 1804, 1815, 1837) does more to disprove such a theory than 20 favourable cases would do towards establishing it. The American panic of 1873, by the way, which occurred when spots were very numerous, decidedly impairs the evidence derived from the crises of 1866 and 1878.

NEW PLANETS NEAR THE SUN.

Perhaps no scientific achievement during the present century has been deemed more marvellous than the discovery of the outermost member (so far as is known) of the sun's family of planets. In many respects, apart from the great difficulty of the mathematical problem involved, the discovery appealed strongly to the imagination. A planet seventeen hundred millions of miles from the sun had been discovered in March, 1781, by a mere accident, though the accident was not one likely to occur to any one but an astronomer constantly studying the star-depths. Engaged in such observation, but with no idea of enlarging the known domain of the sun, Sir W. Herschel perceived the distant planet Uranus. His experienced eye at once recognised the fact that the stranger was not a fixed star. He judged it to be a comet. It was not until several weeks had elapsed that the newly discovered body was proved to be a planet, travelling nearly twice as far away from the sun as Saturn, the remotest planet before known. A century only had elapsed since the theory of gravitation had been established. Yet it was at once perceived how greatly this theory had increased the power of the astronomer to deal with planetary motions. Before a year had passed more was known about the motions of Uranus than had been learned about the motion of any of the old planets during the two thousand years preceding the time of Copernicus. It was possible to calculate in advance the position of the newly discovered planet, to calculate retrogressively the path along which it had been travelling, unseen and unsuspected, during the century preceding its discovery. And now observations which many might have judged to be of little value, came in most usefully. Astronomers since the discovery of the telescope had formed catalogues of the places of many hundreds of stars invisible to the naked eye. Search among the observations by which such catalogues had been formed, revealed the fact that Uranus had been seen and catalogued as a fixed star twenty-one several times! Flamsteed had seen it five times, each time recording it as a star of the sixth magnitude, so that five of Flamsteed's stars had to be cancelled from his lists. Lemonnier had actually seen Uranus twelve times, and only escaped the honour of discovering the planet (as such) through the most marvellous carelessness, his astronomical papers being, as Arago said, 'a very picture of chaos.' Bradley saw Uranus three times.[3] Mayer saw the planet once only.

It was from the study of the movements of Uranus as thus seen, combined with the planet's progress after its discovery, that mathematicians first began to suspect the existence of some unknown disturbing body. The observations preceding the discovery of the planet range over an interval of ninety years and a few months, the earliest observation used being one made by Flamsteed on December 23, 1690. There is something very strange in the thought that science was able thus to deal with the motions of a planet for nearly a century before the planet was known. Astronomy calculated in the first place where the planet had been during that time; and then, from records made by departed observers, who had had no suspicion of the real nature of the body they were observing, Astronomy corrected her calculations, and deduced more rigorously the true nature of the new planet's motions.

But still stranger and more impressive is the thought that from researches such as these, Astronomy should be able to infer the existence of a planet a thousand million miles further away than Uranus itself. How amazing it would have seemed to Flamsteed, for example, if on that winter evening in 1693, when he first observed Uranus, he had been told that the orb which he was entering in his lists as a star of the sixth magnitude was not a star at all, and that the observation he was then making would help astronomers a century and a half later to discover an orb a hundred times larger than the earth, and travelling thirty times farther away from the sun.

Even more surprising, however, than any of the incidents which preceded the discovery of Neptune was this achievement itself. That a planet so remote as to be quite invisible to the naked eye, never approaching our own earth within less than twenty-six hundred millions of miles, never even approaching Uranus within less than nine hundred and fifty millions of miles, should be detected by means of those particular perturbations (among many others) which it produced upon a planet not yet known for three-quarters of a century, seemed indeed surprising. Yet even this was not all. As if to turn a wonderful achievement into a miracle of combined skill and good fortune, came the announcement that, after all, the planet discovered in the spot to which Adams and Leverrier pointed was not the planet of their calculations, but travelled in an orbit four or five hundred millions of miles nearer to the sun than the orbit which had been assigned to the unknown body. Many were led to suppose that nothing but a most marvellous accident had rewarded with such singular success the calculations of Adams and Leverrier. Others were even more surprised to learn that the new planet departed strangely from the law of distances which all the other planets of the solar system seemed to obey. For according to that law (called Bode's law) the distance of Neptune, instead of being about thirty times, should have been thirty-nine times the earth's distance from the sun.

In some respects the discovery of a planet nearer to the sun than Mercury may seem to many far inferior in interest to the detection of the remote giant Neptune. Between Mercury and the sun there intervenes a mean distance of only thirty-six millions of miles, a distance seeming quite insignificant beside those which have been dealt with in describing the discovery of Uranus and Neptune. Again it is quite certain that any planet between Mercury and the sun must be far inferior to our own earth in size and mass, whereas Neptune exceeds the earth 105 times in size and 17 times in mass. Thus a much smaller region has to be searched over for a much smaller body. Moreover, while mathematical calculation cannot deal nearly so exactly with an intra-Mercurial planet as with Neptune, for there are no perturbations of Mercury which give the slightest information as to the orbital position of his disturber, the part of the heavens occupied by the intra-Mercurial planet is known without calculation, seeing that the planet must always lie within six or seven degrees or so of the sun, and can never be very far from the ecliptic.

Yet in reality the detection of an intra-Mercurial planet is a problem of far greater difficulty than that of such a planet as Neptune, while even now when most astronomers consider that an intra-Mercurial planet has been detected, the determination of its orbit is a problem which seems to present almost insuperable difficulties.

I may remark, indeed, with regard to Neptune, that he might have been successfully searched for without a hundredth part of the labour and thought actually devoted to his detection. It may sound rather daring to assert that any fairly good geometrician could have pointed after less than an hour's calculation, based on the facts known respecting Uranus in 1842, to a region within which the disturbing planet must certainly lie,—a region larger considerably no doubt than that to which Adams and Leverrier pointed, yet a region which a single observer could have swept over adequately in half-a-dozen favourable evenings, two such surveys sufficing to discover the disturbing planet. I believe, however, that no one who examines the evidence will deny the accuracy of this statement. It was manifest, from the nature of the perturbations experienced by Uranus, that between 1820 and 1825 Uranus and the unknown body had been in conjunction. From this it followed that the disturber must be behind Uranus in 1840-1845 by about one-eighth of a revolution round the sun. With the assumptions made by Adams and Leverrier, indeed, the position of the stranger in this respect could have been more closely determined. There could be little doubt that the disturbing planet must be near the ecliptic. It followed that the planet must lie somewhere on a strip of the heavens, certainly not more than ten degrees long and about three degrees broad, but the probable position of the planet would be indicated as within a strip four degrees long and two broad.[4] Such a strip could be searched over effectually in the time I have named above, and the planet would have been found in it. The larger region (ten degrees long and three broad) could have been searched over in the same time by two observers. If indeed the single observer used a telescope powerful enough to detect the difference of aspect between the disc of Neptune and the point-like image of a star (the feature by which Galle, it will be remembered, recognised Neptune), a single night would have sufficed for the search over the smaller of the above-mentioned regions, and two nights for the search over the larger. The search over the smaller, as already stated, would have revealed the disturbing planet.

On the other hand, the astronomer could not determine the direction of an intra-Mercurial planet within a considerably larger space on the heavens, while the search over the space within which such a planet was to be looked for was attended by far more serious difficulties than the search for Neptune. In fact, it seems as though, even when astronomers have learned where to look for such a planet, they cannot expect to see it under ordinary atmospheric conditions when the sun is not eclipsed.

Let us consider the history of the search for an intra-Mercurial planet from the time when first the idea was suggested that such a planet exists until the time of its actual discovery—for so it seems we must regard the observations made during the total eclipse of July, 1878.

On January 2, 1860, M. Leverrier announced, in a paper addressed to the Academy of Sciences, that the observations of Mercury could not be reconciled with the received elements of the planet. According to those elements, the point of Mercury's orbit which lies nearest to the sun undergoes a certain motion which would carry it entirely round in about 230,000 years. But to account for the observed motions of Mercury as determined from twenty-one transits over the sun between the years 1697 and 1848, a slight increase in this motion of the perihelion was required, an increase, in fact, from 581 seconds of arc in a century to nearly 585. The result would involve, he showed, an increase in our estimate of the mass of Venus by a full tenth. But such a change would necessarily lead to difficulties in other directions; for the mass of Venus had been determined from observations of changes in the position of the earth's path, and these changes had been too carefully determined to be readily regarded as erroneous. 'This result naturally filled me with inquietude,' said Leverrier later. 'Had I not allowed some error in the theory to escape me? New researches, in which every circumstance was taken into account by different methods, ended only in the conclusion that the theory was correct, but that it did not agree with the observations.' At last, after long and careful investigation of the matter, he found that a certain slight change would bring observation and theory into agreement. All that was necessary was to assume that matter as yet undiscovered exists in the sun's neighbourhood. 'Does it consist,' he asked, 'of one or more planets, or other more minute asteroids, or only of cosmical dust? The theory tells us nothing on this point.'

Leverrier pointed out that a planet half the size of Mercury between Mercury and the sun would account for the discrepancy between observation and theory. But a planet of that size would be a very conspicuous object at certain times, even when the sun was not eclipsed; and when favourably placed during eclipses would be a resplendent orb which would attract the notice of even the most careless observer. For we must remember that the brightness of a planet depends in part on its size and its distance from the earth, and in part on its distance from the sun. A planet half as large as Mercury would have a diameter about four-fifths of Mercury's, and at equal distance would present a disc about two-thirds of Mercury's in apparent size. But supposing the planet to be half as far from the sun as Mercury (and theory required that the planet should be rather nearer the sun), its surface would be illuminated four times as brightly as that of Mercury. Hence, with a disc two-thirds as large as Mercury's, but illuminated four times as brightly, the planet would shine nearly three times as brilliantly when seen under equally favourable conditions during eclipse. In such an inquiry, the mean distance of the two bodies need not be specially considered. Each planet would be seen most favourably when in the part of its path remotest from the earth, so that the planet nearest to the sun would on the whole have the advantage of any difference due to that cause. For, of course, while Mercury, being farther from the sun, approaches the earth nearer when between the earth and sun, he recedes farther from the sun for the same reason when on the part of his path beyond the sun.

It was perfectly clear that no such planet as Leverrier considered necessary to reconcile theory and observation exists between the sun and Mercury's orbit. It appeared necessary, therefore, to assume that either there must be several smaller planets, or else that a cloud of cosmical dust surrounds the sun. Now it is to be noticed that in either case the entire mass of matter between Mercury and the sun must be greater to produce the observed disturbance than the mass of a single planet travelling at the outside of the region supposed to be occupied either by a group of planets or a cloud of meteorites.

Leverrier considered the existence of a ring of small planets afforded the most probable explanation. He recommended astronomers to search for such bodies. It is noteworthy that it was in reference to this suggestion that M. Faye (following a suggestion of Sir J. Herschel's) proposed that at several observatories, suitably selected, the sun should be photographed several times every day with a powerful telescope. 'I have myself,' he says, 'shown how to give these photographs the value of an astronomical observation by taking two impressions on the same plate after an interval of two minutes. It will be sufficient to superpose the transparent negatives of this size taken at a quarter of an hour's interval, to distinguish immediately the movable projection of a small planet in the middle of the most complex groups of small spots.'

It was while Leverrier and Faye were discussing this matter, that news came of the recognition of an intra-Mercurial planet by Lescarbault, a doctor residing at Orgères, in the department of Eure et Loire. The story has been so often told that I am loth to occupy space with it here. An account is given of the leading incidents in an article called 'The Planets put in Leverrier's Balance,' in my 'Science Byways,' and a somewhat more detailed narrative in my 'Myths and Marvels of Astronomy.' Here, it will suffice to give a very slight sketch of this interesting episode in the history of astronomy.

On January 2, 1860, news reached Leverrier that Lescarbault had on March 26, 1859, seen a round black spot on the sun's face, and had watched it travelling across like a planet in transit. It had remained in view for one hour and a quarter. Leverrier could not understand why three-quarters of a year had been allowed to elapse before so important an observation had been published. He went to Orgères with the idea of exposing a pretender. The interview was a strange one. Leverrier was stern and, to say the truth, exceedingly rude in his demeanour, Lescarbault singularly lamb-like. If our chief official astronomer called uninvited upon some country gentleman who had announced an astronomical discovery, and behaved as Leverrier did to Lescarbault, there would most certainly have been trouble; but Lescarbault seems to have been rather pleased than otherwise. 'So you are the man,' said Leverrier, looking fiercely at the doctor, 'who pretends to have seen an intra-Mercurial planet. You have committed a grave offence in hiding your observation, supposing you really have made it, for nine months. You are either dishonest or deceived. Tell me at once and without equivocation what you have seen.' Lescarbault described his observation. Leverrier asked for his chronometer, and, hearing that the doctor used only his watch, the companion of his professional journeys, asked how he could pretend to estimate seconds with an old watch. Lescarbault showed a silk pendulum 'beating seconds,'—though it would have been more correct to say 'swinging seconds.' Leverrier then examined the doctor's telescope, and presently asked for the record of the observations. Lescarbault produced it, written on a piece of laudanum-stained paper which at the moment was doing service as a marker in the Connaissance des Temps. Leverrier asked Lescarbault what distance he had deduced for the new planet. The doctor replied that he had been unable to deduce any, not being a mathematician: he had made many attempts, however.[5] Hearing this, Leverrier asked for the rough draft of these ineffective calculations. 'My rough draft?' said the doctor. 'Paper is rather scarce with us here. I am a joiner as well as an astronomer' (we can imagine the expression of Leverrier's face at this moment); 'I calculate in my workshop, and I write upon the boards; and when I wish to use them in new calculations, I remove the old ones by planing.' On adjourning to the carpenter's shop, however, they found the board with its lines and its numbers in chalk still unobliterated.

This last piece of evidence, though convincing Leverrier that Lescarbault was no mathematician, and therefore probably in his eyes no astronomer, yet satisfied him as to the good faith of the doctor of Orgères. With a grace and dignity full of kindness, which must have afforded a singular contrast to his previous manner, he congratulated Lescarbault on his important discovery. He made some inquiry also at Orgères, concerning the private character of Lescarbault, and learning from the village curé, the juge de paix, and other functionaries, that he was a skilful physician, he determined to secure some reward for his labours. At Leverrier's request M. Rouland, the Minister of Public Instruction, communicated to Napoleon III. the result of Leverrier's visit, and on January 25 the Emperor bestowed on the village doctor the decoration of the Legion of Honour.

To return to astronomical facts.

It appears from Lescarbault's observation, that on March 26, 1859, at about four in the afternoon, a round black spot entered on the sun's disc. It had a diameter less than one-fourth that of Mercury (which he had seen in transit with the same telescope and the same magnifying power on May 8, 1845). The time occupied in the transit of this spot was about one hour seventeen minutes, and, the chord of transit being somewhat more than a quarter of the sun's diameter in length, Lescarbault calculated that the time necessary to describe the sun's diameter would have been nearly four and a half hours. The inclination of the body's path to the ecliptic seemed to be rather more than 6 degrees, and was probably comprised between 5-1/3 and 7-1/3 degrees.

From Leverrier's calculations, it appeared that the time of revolution of the new planet would be 19 days 17 hours, its distance from the sun about 147, the earth's being taken as 1,000; giving for Mars, the earth, Venus, Mercury, and Vulcan (as the new planet was named), the respective distances 1, 524, 1,000, 723, 387, and 147. Leverrier assigned 12-1/5 degrees as Vulcan's inclination, and the places where it crosses the ecliptic he considered to be in line with those occupied by the earth on or about April 3 and October 6. Judging from Lescarbault's statement respecting the apparent size of the dark spot, Leverrier concluded that the volume of the stranger must be about one-seventeenth of Mercury's, the masses being presumably in the same proportion. Hence he inferred that the new planet would be quite incompetent to produce the observed change in the orbit of Mercury.

Leverrier further found that the brilliancy of Vulcan when the planet was furthest from the sun on the sky (about eight degrees) would be less than that of Mercury when similarly placed in his orbit, and he hence inferred that Vulcan might readily remain unseen, even during total eclipse. Here, as it seems to me, Leverrier's reasoning was erroneous. If Vulcan really has a volume equal to one-seventeenth of Mercury's, the diameter of Vulcan would be rather less than two fifths of Mercury's and the disc of Vulcan at the same distance about two-thirteenths of Mercury's. But Vulcan, being nearer the sun than Mercury in the ratio of 147 to 387, or say 15 to 39, would be more brightly illuminated in the ratio of 39 times 39 to 15 times 15, or nearly as 20 to 3. Hence if we first diminish Mercury's lustre when at his greatest apparent distance from the sun in the ratio of 2 to 13, and increase the result in the ratio of 20 to 3, we get Vulcan's lustre when he is at his greatest apparent distance from the sun. The result is that his lustre should exceed Mercury's in the same degree that 40 exceeds 39. Or practically, for all the numbers used have been mere approximations, the inference is that Vulcan and Mercury, if both seen when at their greatest distance from the sun during eclipse, would probably shine with equal lustre. But in that case Vulcan would be a very conspicuous object indeed, at such a time; for Mercury when at his greatest distance from the sun, or greatest elongation, is a bright star even on a strongly illuminated twilight sky; moreover, Vulcan, when at either of his greatest elongations, ought to be visible in full daylight in a suitably adjusted telescope. For Mercury is well seen when similarly placed, and even when much nearer to the sun and on the nearer part of his path where he turns much more of his darkened than of his illuminated hemisphere towards us. Venus has been seen when so near the sun that the illuminated portion of her disc is a mere thread-like sickle of light. Nay, Professor Lyman, of Yale College, in America, has seen her when so near the sun that she appeared to be a mere circular thread of light, the completion of the circle being the best possible proof how exceedingly fine the thread must have been, and also how small its intrinsic lustre.

This is indeed the chief difficulty in Lescarbault's supposed observation. If he really saw a body in transit across the sun, moving at the observed rate, and having anything like the observed diameter, that body ought to have been seen repeatedly during total eclipses of the sun, and ought not to have escaped the search which has been made over and over again near the sun for intra-Mercurial planets. Either we must reject Lescarbault's narrative absolutely, or we must suppose that he greatly over-estimated the size of the body he observed.

Another difficulty almost equally important is found to exist when we consider the circumstances of Lescarbault's supposed discovery. Suppose the path of Vulcan to be inclined about twelve degrees or thereabouts to the ecliptic or to the plane in which the earth travels. Then, as seen from the earth on April 3, and October 6, this path, if it were a material ring, would appear as a straight line across the sun's centre, and extending on either side of the sun to a distance of about 16 sun-breadths. As seen on January 3 and July 5, when it would have its greatest opening, Vulcan's path would appear as an oval whose longest axis would be about 32 sun-breadths, while its shortest would be little more than 6 sun-breadths, the sun of course occupying the centre of the ellipse, which, where closest to him, would lie but about 2½ sun-breadths only from the outline of his disc. Now it is easily seen that the path of Vulcan, changing in this way from apparent straightness to a long oval (whose breadth is about one-fifth its length), back to straightness but differently inclined, then to the same oval as before but opened out the other way, and so back to its original straightness and inclination, must, for no inconsiderable portion of the year on either side of April 3 and October 6, intersect the outline of the sun's disc. From a rough but sufficiently accurate calculation which I have made, I find that the interval would last about 36 days at each season, that is, from about March 16 to April 21 in spring, and from about September 18 to about October 24 in autumn. But during a period of 36 days there would generally be two passages of Vulcan between the earth and sun, and there would always be one (in any long period of time two such passages would be five times as common an event during one of these intervals as a single passage). Consequently there would be at least two transits of Vulcan every year, and there would generally be four transits; the average number of transits would be about eleven in three years. With a wider orbit and a greater inclination transits would be fewer; but even with the widest orbit and the greatest inclination that can possibly be allowed, there would be at least one transit a year on the average.

Now when we remember that, so far as the northern hemisphere is concerned, the sun is observed on every fine day in almost every country in Europe and in half the States of the American Union, to say nothing of observations in Asia, where England and Russia have several observatories, while in the southern hemisphere there are many observatories, in Australia, South Africa, and South America (on both side of the Andes), we see how exceedingly small must be the chance that Vulcan could escape detection even for a single year. Far less could Vulcan have escaped all the years which have elapsed since Lescarbault announced his discovery, to say nothing of all the observations made by Carrington, Schwabe, and many others, before the year 1860. If Vulcan really exists, and really has the dimensions and motions described by Lescarbault, the planet must long ere this have been repeatedly seen upon the sun's disc by experienced observers.

As a matter of fact, Wolf has collected nineteen observations of dark bodies unlike spots on the sun, during the interval between 1761 and 1865. But as Professor Newcomb justly points out, with two or three exceptions, the observers are almost unknown as astronomers. In one case at least the object seen was certainly not a planet, since it was described as a cloud-like appearance. 'On the other hand,' says Newcomb, 'for fifty years past the sun has been constantly and assiduously observed by such men as Schwabe, Carrington, Secchi, and Spörer, none of whom have ever recorded anything of the sort. That planets in such numbers should pass over the solar disc, and be seen by amateur astronomers, and yet escape all these skilled astronomers, is beyond all moral probability.'

It must be remembered that an inexperienced observer of the sun might readily mistake a spot of unusual roundness and darkness for a planet's disc. The practised observer would perceive peculiarities at once indicating the object as a spot on the sun; but these peculiarities would escape the notice of a beginner, or of one using a telescope of small power. Again, an inexperienced observer is apt to mistake the change of position which a spot on the sun undergoes on account of the diurnal motion, for a change of place on the sun's disc. At noon, for instance, the uppermost point of the sun's disc is the north point; but in the afternoon the uppermost point is east of the true north point. Thus a spot which at noon was a short distance below the highest point of the sun's disc would at two or three be considerably to the west of the highest point, though it had undergone in the interval no appreciable change of position on the solar disc. Suppose now that at two or three in the afternoon clouds come over the sun's face, and he is not seen again that day. On the morrow the spot may have disappeared, as solar spots are apt enough to do. The observer, then (assuming him to be inexperienced like most of those who have described such spots), would say, I saw at noon a small round spot which in the course of the next three hours moved over an appreciable arc towards the west (the right direction, be it remembered, for a planet to cross the sun's face). An experienced observer would not make such a mistake. But let one point be carefully noted. An experienced astronomer would be very apt to forget that such a mistake could be made. He would take it for granted that the observer who described such a change in a spot's position meant a real change, not a change due to the diurnal motion.

Therefore, although Leverrier, Moigno, Hind, and other men of science, have adopted Lescarbault's account, I hold it to be absolutely certain that that account is in some respect or other erroneous. Newcomb goes even farther. He says, it is very certain that if the disturbance of Mercury is due to a group of planets, 'they are each so small as to be invisible in transits across the sun. They must also,' he proceeds, 'be so small as to be invisible during total eclipses of the sun, because they have always failed to show themselves then.' This remark relates, of course, to naked-eye vision. As no intra-Mercurial planet had ever been searched for systematically with the telescope, before the recent eclipse, there was nothing to prevent astronomers from believing that a group of planets, visible in the telescope during total eclipse, may travel between the sun and the path of Mercury.

I proceed at once to consider the evidence afforded during the eclipse of July, 1878, not discussing further the question of Lescarbault's Vulcan, because it appears to me so clear that there must have been some mistake, and because later observations seem to throw clearer evidence on the matter than any which had been before obtained. Yet it must be admitted that even now the evidence is not all that could be desired.

Professor Watson, of Ann Arbor, the discoverer of more than a score of the small planets which travel between the paths of Mars and Jupiter, had been searching for an extra-Neptunian planet, when the approach of the eclipse of July, 1878, suggested the idea that he should return for a while from those dismal depths which lie beyond the path of Neptune to seek for a new planet within the glowing region between the sun and the path of Mercury. The occasion was exceptionally favourable because of the great height above the sea-level from which the eclipse could be observed. Accordingly he betook himself to Rawlins, Wyoming, and prepared for the search by providing his telescope with card circles in such sort that the place of any observed star could be recorded by a pencil-mark on these circles, instead of being read off (with the possibility of error) in the usual way. It is unnecessary to explain further, because every one who has ever used an equatorial telescope, or is acquainted with the nature of the instrument, will at once understand Professor Watson's plan, whereas those unfamiliar with the instrument, would not gain any insight into the nature of his plan without much more explanatory matter than could be conveniently given here, even if any explanation without illustrations could make the matter clear. Let it suffice to note that, having brought any star centrally into the telescopic field of view, Professor Watson marked in pencil where the ends of certain pointers came; and that these marks served to indicate, after the eclipse was over, the position of the observed star.