seas viewed from some remote planet would appear unchanged. For they would not be covered with snow and broken ice, nor therefore white; but would consist of pure ice throughout, retaining the partial transparency and greenish colour of deep-sea water. No winds would disturb the surface of the frozen seas, for winds have their origin in heat, and with the death of the solar heat the winds would utterly die out also. If we are to choose between these two explanations,--one that the snows and ice have not the great range we should expect, because the temperature is somehow raised despite Mars's greater distance to the same temperature which we experience, and the other that it is not heat but cold which diminishes the quantity of Martian snow, I conceive that there is every reason the case admits of for accepting the latter instead of the former explanation. As extreme cold would certainly prevent glacial masses from being very large and deep, simply because the stores whence the ice was gathered would be less, the snow caps of a very cold planet would vary as readily with varying seasons as those of a planet like our earth. For though less heat would be poured upon them with the returning summer, less heat would be required to melt away their outskirts. I think we may fairly regard Mars as in all probability a somewhat old and decrepit planet. He is not absolutely dead, like our own moon, where we see neither seas nor clouds, neither snow nor ice, no effects, in fine, of either heat or cold. But I think he has passed far on the road towards planetary death,--that is, towards that stage of a planet's existence when at least the higher forms of life can no longer exist upon the planet's surface.