change, therefore, was not due to clouds or mist.
What was suggestive about the occurrence was that it was unaccompanied by a corresponding increase of blue-green elsewhere. It was not simply that portions of the planet's surface changed tint, but that, taking the disk in its entirety, the whole amount of the blue-green upon it had diminished, and that of the orange-yellow had proportionally increased. Mars looked more Martian than he had in June. The canals, indeed, began at the same time to darken; but, highly important as this was for other reasons, the whole area of their fine lines and associated patches did not begin to make up for what the dark regions lost.
If the blue-green color was due to water, where had all the water gone? Nowhere on the visible parts of the planet; that is certain. Nor could it very well have gone to those north circumpolar regions hid from view by the tilt of the disk; for there was no sign of a growing north polar cap, and, furthermore, Schiaparelli's observations upon that cap show that there should not have been. At the opposition of 1881, he found that it developed late, apparently one month or so after the vernal equinox of its hemisphere, whereas at the time the above change occurred it was not long after that hemisphere's winter solstice.