The limb-light, the apparent evidence of a twilight arc and the planet’s low albedo indicate that this atmosphere is thin. The appearance of the surface, (35), suggests cold, indicative again of a thin air. Such tenuity is in accord with what _a priori_ principles would lead us to expect, and tends to show that reliance on general principles is here not misplaced, a point of some interest. Lastly, the occurrence of clouds, (34), visibly floating and traveling over the surface, and haze at times, (38), proves in another way the existence of the medium in which alone this could be possible. Water is the next substance vital to planetary life. As to its actual presence the polar caps, (4)-(12), have most to say; as to its relative absence, the rest of the disk, (17)-(22). The forthright conception of the polar caps as composed of snow and ice is borne out by further investigation into what could cause the observed phenomenon. Carbonic acid, the only other substance we know capable in any way of resembling what we see, turns out not capable of producing one important detail of the caps’ appearance, the blue band, (5), which accompanies them in their retreat. Water alone could do this. The melting of the caps shows that water vapor must be a constituent of the Martian atmosphere. Moreover, as the molecular weight of water vapor is less than that of oxygen or nitrogen or carbon dioxide, if the former can exist in the atmosphere of the planet, _a fortiori_ must these other gases. So that from this we have knowledge of the possibility of the presence of oxygen, nitrogen, and carbon dioxide there. From (28) we saw that their actual existence is demonstrated. The next step is the ascertainment that the water is in very small amount. The extensive melting of the caps, (7), shows their quantity to be inconsiderable, which is the first fact pointing to a dearth of water. The second comes from the aspect and behavior of the reddish-ochre regions which proclaim them deserts, (18) and (19); the third from the detection of the character of the blue-green areas as not seas, (20), (21), and (22). In several different ways, study of these regions asserts their non-aquatic constitution, the easiest to appreciate being that they are traversed by permanent dark lines and other equally sedentary markings, (22). No bodies of water, therefore, are to be seen outside of the ephemeral polar seas, immediately surrounding the caps as they melt.