After the melting has been for some time under way and the cap has become permanently bordered by its dark blue band a peculiar phenomenon makes its appearance in the cap itself. This is its fission into one or more parts. The process begins by the appearance of dark rifts which, starting in from the cap’s exterior, penetrate into its heart until at last they cleave it in two. Rifts have been seen by several observers and in both caps; and what is most suggestive they always appear in the same places, year after year. Sometimes oppositions elapse between their several detections for they are not the least difficult of detail; but when they are caught, they prove to lie just where they did before. The permanency in place of the rifts, a characteristic true of them all, shows them to be of local habit. Thus the rift of 1884 and 1897 reappeared again to another observer in the same position in 1901. They are, therefore, features of, or directly dependent on, the surface of the planet. But it will not do from this fact to infer that they are expressive of depressions there. The evidence is conclusive that great irregularities of surface do not exist on Mars. As we shall see when we come to consider the orology of the planet it is certain that elevations there of over two or three thousand feet in altitude are absent. Differences of temperature, able to explain a melting of the ice in one locality coincidentally with its retention in an adjacent one, must in consequence be unknown. And this much more conclusively than at first appears, for the reason that the smaller the planet’s mass the less rapidly does its blanket of air thin out in ascent above the surface. This is in consequence of the greater pull the larger body exerts and the greater density it imparts to a compressible gas like our atmosphere. Gravity acts like any force producing pressure and by it the envelope of air is squeezed into a smaller compass. But as this is done throughout the atmospheric layer it means a more rapid rarefaction as one leaves the body. The action is such that the height necessary to reach an analogic density varies inversely as the gravity of the mass. In consequence of this, to compass a relative thermometric fall for which a moderate difference of elevation would suffice on Earth, an immoderate one must be made on Mars. For gravity there being but three eighths what it is here, eight thirds the rise must be made to attain a proportionate lowering of temperature. This fact renders the above argument against elevation and depression being the cause of the phenomenon three times as cogent as it otherwise would be. With so gradual a gradient in barometric pressure there and so low a set of contour lines, altitude must be a negligible factor in Martian surface meteorologic phenomena. Both density and temperature can be but little affected by such cause, and we must search elsewhere for explanation of what surface peculiarities we detect.