Very strange, then, at first must the statement seem, that we are as sure of the existence of water in all these forms on Mars as if we had sent some messenger to the planet who had brought back for study by our chemists a block of Martian ice, a vessel full of Martian water, and a flask of Martian air saturated with aqueous vapour. Indeed, I do not know of any discovery effected by man which more strikingly displays the power of human ingenuity in mastering difficulties which, at a first view, seem altogether insuperable. When we know that a mass of ice as large as Great Britain would appear at the distance of Mars a mere bright point; that a sea as large as the Mediterranean would appear like a faint, greenish-blue, streak; and that cloud masses such as would cover the whole of Europe would only present the appearance of a whitish glare, how hopeless seems the task of attempting to determine what is the real chemical constitution of objects thus seen! It might well be thought that no possible explanation of the method used by astronomers could serve to establish its validity. Yet nothing can be simpler than the principle of the method, or more satisfactory than its application in this special case. First, let the reader rid his mind of the difficulty arising from the enormous distance of the celestial bodies. To do this let him note that there are some things which a body close by can tell us no more certainly than a remote body. For instance, we are just as certain that Mars is a body capable of reflecting sunlight as we are that a cricket-ball is. We know as certainly, too, that the quality of Mars is such that more of the red of the sun's light is sent to us than of the other colours. For we perceive that Mars is a ruddy planet. Since distance in no way interferes with our perception of these general facts, and others like them, we need not necessarily find in mere distance any difficulty in the way of recognising some other facts. All that we require to be shown before admitting the validity of the evidence is, that it is of such a kind that distance does not affect its _quality_, however much distance may and must affect the quantity of evidence.