There is one peculiarity of the planet's appearance which seems strikingly to accord with this view that Mars holds a position intermediate between that of our earth and the moon,--as indeed we might fairly expect from his intermediate proportions. The seas of our earth cover nearly three-quarters of her entire globe. The moon has no visible water on her surface. If we examine the chart of Mars at page 167, we see that the seas and oceans of the planet are much smaller (relatively as well as actually) than are the seas of our own earth. I have carefully estimated their relative extent in the following simple but effective way. I drew a chart such as the above-mentioned, but on a projection of my own invention, in which equal surfaces on a globe are represented by equal surfaces on the planisphere. Then I cut out with a pair of scissors the parts representing land and the parts representing water (leaving the polar parts as doubtful), and carefully weighed these in a delicate balance. I found that they were almost exactly equal: whatever preponderance there was seemed to be in favour of the land. Thus, if we assume that, when in the same stage of planetary existence, Mars had as great a relative extent of water surface as our earth, or that about 72/100 of the surface of Mars were originally water, we should have to admit that the water had so far been withdrawn into the planet's interior as to diminish the water-surface by 22/100 (for there are now barely 50/100). At a very fair assumption as to the slopes of the Martian sea-bottoms, it would follow that more than half the Martian water originally existing above the surface had been withdrawn into the interior, as the planet's mass gradually cooled. I am aware the assumption above mentioned is in itself somewhat daring, and is not supported by direct evidence. But, since we have very strong reasons for considering that the moon once had seas, which have been withdrawn in the way suggested, and since Mars unquestionably holds a position midway between the earth and moon as to size and presumably as to age,[13] it seems not unreasonable to find in the character of her seas,--less extended relatively than the earth's, but, unlike the moon's, still existing,--the evidence that she has gone partially through the process through which the moon has long since passed completely.