In addition to the polar caps proper and to the subsidiary polar patches that often in late summer flank them round about, other white spots may from time to time be seen upon the disk. In appearance these differ in no respect, so far as observed, from the arctic subsidiary snow-fields. Of the same pure argent, they sparkle on occasion in like manner with the sheen of ice. Equally with the polar caps they remain permanent in place during the period of their visibility and are themselves long-lived. Though by no means perpetual their duration is reckoned by weeks and even months, and they recur with more or less persistency at successive Martian years. That, when seen, they show in particular positions apparently unaffected by diurnal change precludes their being clouds, and this fact taken in connection with the character of their habitat is the puzzling point about them. For they affect chiefly the north tropic belt. They, or at least their nuclei, are small, about two or three degrees in diameter, and are not particularly easy of detection as a rule, though certain larger ones are at times conspicuous. Chromatic, rather than formal, definition is necessary to their bringing out, as is witnessed by the superb colors the disk presents at the times when they are best seen. It is then that Mars puts on the look of a fire-opal.

The first such spot to be noticed was one which Schiaparelli detected in 1879, at the second opposition in which he studied the planet. He called it the Nix Olympica, showing that he recognized in it a cousinship to the polar snows. Yet it lay in latitude 20° north, longitude^[1] 131°, in the midst of the ochre stretches of that part of the disk. It was a small roundish white speck of not more than two thirds the diameter of the polar cap. Reseen by him in 1881, it failed to appear at subsequent oppositions and was not caught again until 1888. Then once more it vanished, not to be detected anew till many years after at Flagstaff, coming out rather surprisingly in 1903. It showed, however, in the same place as before; so that its position but not its existence is permanent.

A similar but smaller patch was apparent to Schiaparelli at the same opposition of 1879. This one which he styled the Nix Atlantica lay between the Thoth and the Syrtis Major. It was about half the size of the Nix Olympica and has never since been seen, though it should have been had it continued to be what it then was.

On the other hand, phenomena of the sort undetected of Schiaparelli have been remarked at Flagstaff. On May 18, 1901, I was suddenly struck by the singular whiteness of the southeast corner of Elysium where that region bordered the Trivium. Elysium has a way of being bright but not with such startling intensity as this spot presented nor in so restricted an area as was here the case. The spot was so much whiter than anything I had ever previously seen outside the polar caps that it arrested my attention at once. And this the more that I had observed this same part of the planet the day before and perceived nothing out of the ordinary. Once detected, however, the spot continued visible. The next day it was there with equal conspicuousness, and now thrust an arm across the Cerberus, entirely obliterating the canal for the space of several degrees. In this salience it remained day after day till the region passed from sight, to reappear with it six weeks later when the region again rounded into view. The hour of the Martian day seemed to make no difference in its visibility. It was seen from early morning till Martian afternoon, as late as the phase permitted. Clearly there was nothing diurnal about its revealing, and it lasted for at least three months and a half, until the planet got so far away that observations were discontinued.

It was to all appearances and intents snow. But now comes the singular fact about it. It lay within ten degrees of the equator and showed from the end of June to the latter part of August. To our ideas there could be no more inopportune place or time for such an exhibition. For it cannot have been due to a snow-capped peak, as we know for certain that there are no mountains in this, or in any other, part of the planet. Besides, it had not appeared in previous Martian years; which it infallibly would have done had it been a peak. Indeed, it baffles explanation beyond any Martian phenomenon known to me. It seems directly to contradict every other detail presented by the disk.

The phenomenon is thus unique in kind; it is not, however, unique as a specimen of its kind. The eastern coast of Aeria where that region borders the Syrtis Major is prone to a brilliance of the same sort. It is a narrow belt of country that shows thus, nothing but the coastline itself, but this for a considerable distance stretching several hundred miles in length. It has stood out saliently bright now at every opposition which I have observed, beginning with 1894. Sometimes it has been described in the notes as bright simply, sometimes as white, and once, in 1901, as glistening at one point like ice. Yet when upon the terminator it has never stood forth as a mountain range should have done to declare its character.

It has been evident regardless apparently of the Martian season. In 1894 it was bright from October 25 to January 16 (Martian chronology); in 1896, from December 22 to January 7; in 1901, from July 13 to the 15th; in 1903, at about the same date and so in 1905. It was whitest during the latter oppositions, showing that the effect is most marked in its midsummer. All of the above instances of extra-polar white have been located within the tropics. Examples of the same thing, however, occur in the north temperate zone. Tempe, a region just to the west of the Mare Acidalium, is exceedingly given to showing a small white spot close upon the Mare’s border in latitude 50° north. This spot, too, on occasion glitters as it were with ice. It is also at times very small. So that whereas much of Tempe is by nature bright but a small kernel of it is dazzling.

The list might be easily extended from the record book. Thus on March 1 and 2, 1903, the disk showed speckled with minute white spots, one in Arcadia in latitude 41° north, one in Tharsis near the equator, a third just north of the Phoenix Lucus in 10° south, and a fourth, the Nix Olympica, and on April 11, a glittering pin-point starred like a diamond the centre of the Pons Hectoris. On both these occasions the Martian season was summer; July 9 for the latter, June 21 for the former date.

As one approaches the north pole spots of like character become more numerous. Especially are such visible north of the Mare Acidalium in the arctic region thereabout, from 63° to 75° north.

From so widespread a set of instances the only explanation which seems to fit the phenomena is that the mean temperature of Mars is low, not very much above freezing, and that whatever causes a local fall in the temperature results in hoar-frost. Such an explanation accords well with the distance of the planet from the sun and the thinness of its atmosphere. At the same time it shows that the mean temperature over the greater part of the planet the greater part of the time is above the freezing-point and that consequently it is no bar to vegetation of a suitable sort.

That the hoar-frost should be found even at the equator may perhaps be due to the very thinness of the air-covering of Mars, which would tend to make the actual insolation more of a factor than it is with us, and by the great length of the Martian seasons. In midsummer the greatest insolation occurs in the arctic and temperate, not in the tropic regions; on the other hand, an atmosphere tends to accumulate heat for the tropics. With us the latter factor is prepotent; it would be less effective on Mars. Then again the double duration of summer would tend to emphasize insolation as the important factor in the matter. But it is possible that greater deposition plays a part in the matter. On earth the rainfall is greatest near the equator and something of the sort may be true of the zones of moisture on Mars. That the most striking spots are found to the west of large dark areas may in this connection have a meaning inasmuch as, such regions being vegetation-covered, the air over them is probably more moisture-laden.

One point about the position of the spots is of moment: they have all been found in the northern hemisphere or within ten degrees of it in the southern equatorial region. This seems at first a question of hemispheres; but when we consider that the light areas of the surface are chiefly in the boreal hemisphere and in the south tropic belt, we perceive that it may be rather the character of the surface there than the particular hemisphere in the abstract that is decisive in the matter. Nevertheless, the austral hemisphere is the hemisphere of extremes, possessing a shorter, hotter summer and a longer, colder winter than its antipodes. This would not favor sporadic small depositions of frost in summer so much as would a climate of a more mean temperature.

From the relative lack of atmospheric covering over the planet we should expect the nights to prove decidedly cool, while the days were fairly warm. Of this we have perhaps evidence in a singular aspect shown by the Mare Acidalium in June, 1903. The account of it in the Annals reads thus: “On May 22 an interesting and curious phenomenon presented itself. On that day, so soon as the Mare Acidalium had well rounded the terminator on to the disk, at ?352°, the whole of its central part showed white, the edges of the marking alone remaining as a shell to this brilliant core. So striking was the effect that beside appearing in the drawing it found echo in the notes. The next day no mention is made of it, and a drawing made under ?20° shows the Mare as usual and the bright spot in Tempe in its customary place. Neither was anything of the sort noticed on the 24th and 25th. But on the 26th, the day of the projection (upon the terminator), the effect of the 23d reappeared, the longitude of the centre being 332°. Fortunately on that day a further drawing was secured which enabled its subsequent behavior to be followed. Made three hours later than the other, the longitude of the centre being 13°, this drawing shows the Mare well on the disk, its whole area as dark as usual and with Tempe bright to the right of it toward the terminator. The terminator in question was the sunrise one, and we are offered two suppositions in explanation of the phenomenon: either the white was due to a morning deposition of hoar-frost which dissipated as the sun got up, or obliquity rendered some superficial deposit visible which more vertical vision hid. That the former inference is the more probable seems hinted at by the simultaneous appearance from the 19th to the 26th of other areas of white between the Mare and the pole. May 26 was 88 days after the northern summer solstice, and corresponded to July 31 on the earth.” Annals, Volume III, §564.

In this connection mention may pertinently be made of Schiaparelli’s repeated observation of regions that whiten with obliquity, a proclivity to which he particularly noticed Hellas and certain ‘islands’ in the Mare Erythraeum to be prone. Here as with the Mare Acidalium we certainly seem to be envisaging cases of matutinal frost melted by midday under the sun’s rays.