much the appearance of a bright comet with a long double tail. Schröter, of Lilienthal, discovered this object, and drew it carefully, making the western formation of the two -or Messier itself-slightly the larger. He also suspected that its appearance was variable. In consequence of the observation of Schröter, Beer and Mädler paid particular attention to these formations, examining them most carefully on more than three hundred distinct occasions, between 1829 and 1837. They were thus enabled to declare with certainty that the two formations, Messier and Messier A, were exactly alike in every manner, not the slightest difference in any particulars being detectable. Both were circular crater plains, 9 miles in diameter, with 7° bright greyish white walls, surrounding a yellowish grey interior only 3° bright. On the walls, which were of the same height, were several wall peaks, in both formations situated in the same position with regard to the formation. In fact, as Beer and Mädler draw particular attention to, their own observations are decisive proof that the two formations were exactly equal in every respect, and that in diameter, form, height of their walls above the surrounding surface, and depth of the interior beneath the crest of the walls, colour of the interior and of the walls, and the position of the surrounding wall peaks, Messier and its twin formation Messier A were completely alike. And it is impossible that any difference in these respects would have escaped the attention of the greatest selenographer of our day, during a course of over three hundred observations in a space of nine years.

Some years after this, Gruithuisen-a most zealous and keen-sighted, though fanciful, observer-detected a slight dissimilarity between the form of the two craters, but this fact received little, if any, attention. In November, 1855, the Rev. T. W. Webb, one of the best living iunar observers. who was examining this region with a telescope of similar dimensions to that of Beer and Mädler's, observed that the eastern crater plain appeared the larger of the two. On again looking at the formations on March 11th, 1856, he at once detected that not only was the western crater plain, or Messier the smaller of the two, but that it was elliptical, with its greatest diameter extending from east to west. This fact was confirmed by subsequent observations and from drawings made by him in 1857, whilst Messier A, the eastern formation, appears to have remained unchanged, still being a circular crater plain, with a diameter of 9 miles, the western crater plain, Messier itself, had an elliptical

form, with a long diameter of about 10 to II miles and a short diameter of about 7 to 8 miles. The most rigid equality between the form and dimensions of these two formations, so strongly asserted to exist by Beer and Mädler, then no longer existed in 1857, according to the observations of the Rev. T. W. Webb.

This fact was of very great selenographical importance, and should have attracted general attention, because whatever uncertainty may be attached to the observations of Linné made by Beer and Mädler, it cannot be seriously urged that Beer and Mädler could have failed to recognise the dissimilarity between Messier and Messier A discovered by the Rev. T. W. Webb, when we consider that for over three hundred times they most carefully examined the two formations for the express purpose of detecting any difference. Nor can that potent agent, the lunar librations, be invoked to account for this difference, for in this particular instance it is entirely without sensible influence. For Messier lying very close to the moon's equator, and in longitude 47° W., no variation in the libration of the moon can sensibly affect the meridianal apparent diameter of the formation, and it is in this last that the diminution has been observed. If it was established, therefore, that the two formations were no longer alike, but differed markedly in any respect, the very strongest evidence of actual physical change would be the consequence.

Although therefore it is thus important that further observations of these formations should be obtained, the matter attracted little attention until the period 1870 to 1875. During these years Messier and Messier A have been repeatedly examined with the aid of most powerful telescopes, and the present dissimilarity between Messier and Messier A has been placed beyond the possibility of doubt. From some measures during the past year the long diameter of Messier appears to be 12'2 miles, and the short diameterwhich is nearly, though not quite meridianal--is 6'9 miles. The difference between the form and dimensions of the two formations is now obvious in the smallest astronomical telescope, for it is unmistakable with a power of 150 on a telescope with an aperture of a little over 2 inches. With the fine Frauenhofer equatorial of Beer and Mädler, with an aperture of nearly 4 inches and a power of nearly 300, the difference now existing between the two formations would have instantly arrested the attention of any selenographer. It is inconceivable that the two formations could have possessed their present form in Beer and Mädler's time, and

have been on hundreds of different occasions carefully scrutinised to detect any difference that might exist between them, without the most marked dissimilarity which now exists being detected.

The most difficult portion of the question remains, however, still untouched; for even were it granted that in the case of Messier there exists an instance of real physical change, the process by which that change could have been brought about seems to defy explanation. In the case of Linné we have a plausible and easy method of accounting for the observed phenomenon, but in Messier what it seems must have occurred is a slow squeezing of an immense crater plain out of shape. The nature of the forces and processes at work upon the surface of the moon, which could compress a circular crater 9 miles in diameter into an elliptical formation some 12 miles by 7, is such with which there at present exists nothing analogous on the surface of the earth; and until some satisfactory theory could be devised which would show by what means this change in Messier could have been brought about, in a manner consistent with other selenographical facts, it is not surprising that a strong reluctance should exist to admitting that a change had occurred in the form and dimensions of Messier, even when backed by the extremely strong evidence in its favour that has been adduced.

It is possible, however, that this difficulty may be avoided, for careful examination of the crater plain Messier and its neighbourhood suggests that, instead of a bodily compression of the entire crater, the observed change may have occurred from the gradual sliding of the north and south walls into. the interior, and in so doing pushing the entire western wall, outwards and westwards, down an incline existing there. If this explanation is found to be consistent with the present condition of the formations, it at once removes all the difficulties, because the process it involves-namely, the gradual slipping of the wall of a formation into the interior-is one of which numerous traces exist upon the moon. There would be little difficulty on the part of selenographers in pointing to a hundred instances where a like circumstance had occurred,-one within Gassendi being a notable case, and one where a massive wall has been pushed back nearly 10 miles. As far as is at present known, this explanation is in accordance with the condition of the surface immediately around Messier, but further observations with powerful telescopes are indispensable. Unfortunately the region. of the moon containing Messier is one which cannot often

be observed at the proper time for this purpose; so that only one or two days during the year are likely to be found when Messier is in a proper position to be observed, and the weather will permit a high power and a powerful telescope to be used to advantage, for these require very steady and favourable atmospheric conditions.

It is not, however, in these two instances alone that selenographers have detected instances of changes in the lunar surface; but these two instances are cases where there exist the very strongest evidence that real physical changes. have occurred. There are various other remarkable cases which render it very probable that different small portions of the moon's surface have undergone striking variations; but though they rest on evidence that in other branches of astronomy have been usually accepted as sufficient to establish the existence of such variations, they do not rest on evidence of so overpowering a weight as is now alone considered worthy of credit on these selenographical questions. Thus towards the central portions of the moon, on the wide open plain called the Mare Nubium, according to Beer and Mädler, when they drew their map and whilst they were making their observations, there existed a fine, deep, glittering white crater, called by them Alpetragius d. It had a diameter of 5 miles, and close to it, on the south-west, was a smaller crater, about 1 mile in diameter. The large crater Alpetragius d has entirely disappeared, leaving in its place a slightly less bright, perfectly round, greyish white spot, with a diameter of 72 miles. The smaller crater, however, remains unaltered, and corresponds to Mädler's description. We have here, in fact, an exactly similar case to the wellknown instance of Linné. About 240 miles south of this last region, on the border of the grey Mare Nubium, is a fine though ruined walled plain, named Hesiodus, and consisting of a level plain nearly 20 miles in diameter, surrounded by the ruins of a mountain border. Beer and Mädler describe and draw this formation as containing a perfectly level interior, grey in colour, and becoming slightly lighter towards the centre. Within it now, nearly in the centre, is a fine deep crater, 4 miles in diameter, and very distinct long after sunrise, so that it seems difficult to understand how it could have escaped Beer and Mädler's notice, had it then existed like it is now. Many other instances of a similar nature might be easily quoted.

Selenographers have also been long acquainted with variations of a different nature, occurring in lunar regions, and consisting in variation in colour or brightness. Beer and

Mädler, who were the first to notice the existence of definite variations in tint, were struck by the analogy they bore to the changes which would result from processes of vegetation on the lunar surface, though, as they pointed out, this seemed incompatible with the absence of masses of water upon the moon. It is remarkable that, despite this apparently fatal objection, other experienced selenographers have been led to the same conclusion, from evidence of widely diverse nature. Other instances of colour changes on the moon, also periodical in character, appear to indicate the effects of the solar rays continuously poured down on the lunar surface for nearly fifteen days at a time; and to this class appears to belong the variation in the floor of Plato, which has been discussed by Mr. Proctor in the "Quarterly Journal of Science" for January and October, 1873. There are other instances which, unlike the last, are not periodical in nature, and which indicate in a very forcible manner the effect of the slow, but energetic and increasing, disintegrative agents which have destroyed all the older lunar formations, and the principal of which appears to be the extensive and all-powerful lunar atmosphere. The capabilities and power of this last member of the forces moulding the surface of the moon are generally overlooked, on the supposition that its small density, when compared with the extremely dense atmosphere of the earth, must destroy its energy and capabilities, forgetting that what it may want in density it may make up in volume.*

The variation in the tint of the floor of Plato is one of the

It is to be remembered that the influence of the atmosphere upon the surface of the moon depends almost entirely upon its mass, and not upon its surface density. It is not through its pressure that the lunar atmosphere exerts its influence upon the surface of the moon, but through its chemical action upon the constituents of the crust of the moon, or through modifying the great variations in temperature to which the lunar surface is exposed. In its chemical action the effects of the atmosphere will depend almost entirely upon its mass, because, through diffusion and similarly acting causes, all portions of the atmosphere will successively come in contact with the surface. Were the action of the atmosphere very rapid, so that the entire chemical power of any portion was liable to be exhausted before its place could be taken by another portion through the agency of diffusion, then it is true a decrease in the density of the atmosphere might produce a decreased action in a given time. But the action of the atmosphere is so excessively slow that any effect of the kind stated is out of the question. The effect of the atmosphere in modifying the variation in temperature to which the lunar surface is exposed likewise depends almost entirely upon the mass of the atmosphere, and not upon its surface density. Thus, in the retardation of the radiation of heat from the moon, and in the absorption of the solar heat, it is not the surface density of the atmosphere on which these effects depend, but, as both Tyndall and Magnus have shown, upon its mass. Yet the strange idea that the powers of the lunar atmosphere lay entirely in its surface density seems to be entertained by a hostile reviewer of my work on the Moon.

VOL. VII. (N.S.)

C