Thus is the opposed action upon the two planets accounted for, and we are led to the conclusion that the canals are strips of vegetation fed by water from the polar caps, and that the floral seasons there as affecting the canals are conditioned, not as they would be with us, directly upon the return of the sun, but indirectly so through its direct effect upon the polar snows.
Once adventured on the idea of vegetation, we find that it explains much more than the time taken by the wave of canal-development down the disk. It accounts at once for the behavior of the canals in the three northern zones: the polar, arctic, and sub-arctic. The mean cartouches of these three zones dip down at their latter end instead of rising there, as is the case with the cartouches of the mean canals farther south. This dip denotes that the most northern canals were waning already by the middle of their August, though the others showed no such tendency; while the date of the deposition of the frost in these northern latitudes shows that they were started upon their course toward extinction before the snow itself had covered them. In other words, they were not obliterated but snuffed out. That their decline was thus preparatory to the coming of the first snowfall or frost-fall, sufficiently severe to whiten the ground so that it did not melt the next day, is suggestive of their constitution. It is clear that they were not abruptly cut off by the frost, but were timed by nature to such extinction. Vegetation would behave in just this way, since evolution would accommodate the career of a plant to its environment.
The first question to present itself chronologically in the canals’ annual history is connected with the size of the cap. Unfortunately for the simplicity of the phenomena, the cap is not an extensionless source of flow, but an extended surface melting from the outer edge in. It would seem, therefore, that water liberated from the outer parts should have an effect before the main body of it were ready to begin its general march down the disk. There should be, one would think, at least a partial action, locally, before the main action got under way. Now, there are certain canals that show cartouches increasing apparently from the time observations began, and the most pronounced is the Jaxartes, which lies of all the canals observed the farthest north. Now, the cartouches were founded on canals quickened from the north polar cap. The farther north the canal, therefore, the greater the likelihood of its showing the phenomena.