detail, and itself instantly suggested its character, namely, that it was water at the edge of the cap due to the melting of the polar snow.

Not only did the band conform with the cap in position; it did so in size. As the snows dwindled, the blue band about them shrank in width to correspond. By August it was a barely discernible thread drawn round the tiny white patch which was all that remained of the enormous snow-fields of some months before.

Finally, on October 13, when the snow entirely disappeared, as we shall presently see, the spot where it and its girdle, long since grown too small for detection, had been became one yellow stretch.

That the blue was water at the edge of the melting snow seems unquestionable. That it was the color of water; that it so persistently bordered the melting snow; and that it subsequently vanished, are three facts mutually confirmatory to this deduction. But a fourth bit of proof, due to the ingenuity of Professor W. H. Pickering, adds its weight to the other three. For he made the polariscope tell the same tale. On scrutinizing the great bay through an Arago polariscope, he found the light coming from the bay to be polarized. Now, to polarize the light it reflects is a property, as we know, of a smooth surface such as that of water is.

Before going further we will take up here at the outset the question of the constitution of these polar caps, which in their general behavior so strikingly suggest our own ice-caps as they would appear could they be seen from a distance of forty millions of miles. That they so instantly suggest snow has suggested, to that class of mind which likes to make of molehills of question mountains of doubt, the possibility that instead of ice we have here snow-caps of solid carbonic acid gas (carbon dioxide). The occasion of the suggestion is the fact that carbonic dioxide under certain conditions becomes a colorless liquid, and then a solid of a floccular, snow-like character. It assumes, in short, under proper conditions of pressure and cold, the various appearances presented by water under higher temperatures, although it does so with very different degrees of ease. Superficially, therefore, the idea seems plausible. Let us see if it still seems so when critically examined.

Faraday made experiments on the relation of the congealing point of carbonic acid gas to the pressure, and found that at 0 degrees C. it took a pressure of 36 atmospheres, that is, 540 pounds to the square inch, to solidify the gas, and that at -99 degrees C., the lowest temperature with which he experimented, it took 1.14 atmospheres. At this point the curve representing the relation was becoming apparently asymptotic, that is, a slight decrease in pressure involved a great falling off of temperature. Under a pressure of one atmosphere, therefore, the temperature would be about -170 degrees F., that is, on the surface of the Earth this would be the congealing point of the gas.

He found further that the curve for the liquefaction point lay very close to that for the congealing point, and approached yet closer as the pressure decreased. In other words, the gas passed almost immediately from the gaseous to the solid state.

In the light of these facts let us consider the condition of Mars. Three points arise which we will take in the inverse order of their importance. First: the appearance of the planet shows conclusively that, if the polar caps be composed of solid carbonic acid gas, then either there is no water at all on Mars in any form whatsoever, or what there is is ice so overlaid with detritus as to be invisible. For if the two substances were there together, and the cold at the surface of the planet of so extreme a character as to congeal the carbon dioxide, the water must a fortiori be frozen, and would continue so long after the temperature rose above the melting point of the former substance. We should therefore still have snow- fields of snow after the melting of those formed of carbonic acid gas, either visible as white patches or so covered up with dirt as to pass for land. Now there are no such additional white patches to be seen, nor, so far as we can judge, does any part of the planet behave as if it were glacier-bound.

Second: carbonic dioxide passes, as we saw, almost simultaneously into the liquid and solid states, especially under slight pressure. Now, the pressure is certainly very slight on the surface of Mars; not probably