substantialize by darkening in tint throughout. Now, to deepen thus in color with one consent all over would be a peculiar thing for a lake to do. For had the lake appreciable depth to start with, it should always be visible; and had it not, its bed would have to be phenomenally level to permit of its being all flooded at once. If, however, the spots be not bodies of water, but areas of verdure, their deepening in tint throughout is perfectly explicable, since the darkening would be the natural result of a simultaneous growth of vegetation. This inference is further borne out by the fact that to the spot class belong unquestionably those larger oval markings of which the Lake of the Sun is the most conspicuous example. For both are associated in precisely the same manner with the canal system. Each spot is a centre of canal connections in exactly the way in which the Solis Lacus or the Phoenix Lake itself is. But the light coming from the Solis Lacus and the Phoenix Lake showed, in Professor W. H. Pickering's observations, no sign of polarization such as a sheet of water should show, and such as the polar sea actually did show.
When we put all these phenomena together,—the presence of the spots at the junctions of the canals, their strangely systematic shapes, their seasonal darkening, and, last but not least,