Close treading on the heels of the detection of lines athwart the seas came the recognition of spots there likewise. At the opposition of 1896-1897 the number was added to; and so the tale has been steadily increased. Their number as found at Flagstaff up to the present time, that is, to the close of the opposition of 1905, is 186; of which 121 lie in the light regions, 42 in the dark areas of the southern hemisphere, and 23 in the smaller sombre patches of the northern zones. From their relationships and behavior it became apparent that the spots were not lakes but something which answered much more nearly to oases. Of the spots three kinds may be distinguished: the large, the little, and the less, if by the latter term it may be permitted to denote what has but collateral claim to be included and yet demands a certain recognition. For though not spots like the others, the members of the third class have certain traits in common with them while differing radically in others. To the kind called large belong the greater number of spots so far found upon the disk. They are large only by comparison with the little. For they measure according to my latest determinations but seventy-five or one hundred miles in diameter; on the planet some two degrees across. Sizable black pin-heads, it is their tone that chiefly catches the eye, for they are commonly the darkest markings on the disk. Against the ochre stretches they appear black, and even in the midst of the dark areas they stand out almost as much contrasted with their surroundings as these do with the light regions themselves. About a hundred and forty are now known. Those in the light areas were discovered first; those in the dark regions being harder to see. Of this first kind are such spots as the Pseboas Lucus, the Aquae Calidae, the Lacus Phoenicis, and the Novem Viae; or, in English, the Grove of Pseboas, the Hot Springs, the Phœnix Lake, and the Nine Ways, to mention no more. That they bear dissimilar names implies no dissimilarity in structure. The phenomena are all remarkably alike, and clearly betoken one and the same class of objects; differing between themselves at most in size and importance. In form they all seem to be round. They certainly appear so, and were it not that retinal images of small areas tend to assume this shape might implicitly be credited with being what they seem. The reason for optical circularity probably resides in the shape of the retinal cones and in their patterning into a mosaic floor. So that unless a sufficient number of cones be struck the image takes on to consciousness a roughly circular figure—whether it be so in fact or not. In the present case, however, they seem to be too well seen for self-deception of the sort.