What is significant, the deposition of the frost took place simultaneously over large areas. The very first patch of it, in about longitude 320°, extended at one stroke down to latitude 55°. For it actually crossed the Pierius somewhat to the south. A second patch stretched to the east of the cap. Two wings these made to the kernel of cap itself. Through the wings could be marked the line of the canal: the Pierius upon the one side, the Enipeus upon the other. Such visibility of the canals through the white stretches proved the white not to be due to cloud suspended between us and them, but a surface deposit which found no lodgment upon the canals themselves. The same avoidance of dark markings was evidenced by the showing of the dark rim round the cap’s kernel. Now, if the deposit were indeed hoarfrost, this failure to find permanent foothold on the dark markings is what we should expect to witness. For whether they were vegetation or water, equally in either case the frost would melt from them first. Probably they were both vegetal, though some doubt might exist about the latter, the band around the kernel. It was then August 20 in that hemisphere. Such deposition over great stretches of country is perhaps not so surprising as it appears at first sight when seen from without in its totality. After all, something not unlike it occurs in our snow-storms when hundreds of square miles are whitened at once. Furthermore, with an atmosphere as thin as Mars seems to possess the temperature must be perilously near the freezing-point in the arctic and subarctic regions at the close of summer. Steadily, with intermissions, the white sheet increased until even the dark border to the cap became obliterate, the kernel showing at first through the veil like the ghost of what it had been, and then ceasing to be visible at all, its delimitations being buried under deeper and deeper depositions of frost. [Illustration: First northern snow.] The perennial portion of the cap was thus merged in the new-fallen snow. This marked the on-coming of the arctic winter in full force and happened even before the polar sun had wholly set. For the pole did not enter into the shadow till two of our months later, the autumnal equinox occurring 183 days after the summer solstice or 55 days after the first fall of frost. Then the pole passed into its star-strewn arctic night, a polar night of twice the duration of our own and the circumpolar regions entered upon their long hibernation of ten of our months. CHAPTER VII WHITE SPOTS