The return of the part of the planet where it had been seen was eagerly awaited the night after by both observers, to see if it would bring the projection with it. For only once a day is the same region of Mars similarly presented. But in order not to miss the projection, should it be ahead of time, observations were begun before it was due. Shortly after they were started, there appeared higher up the terminator and therefore farther north than had been the case the night before, a small projection. It was with difficulty made out and its position measured. Without careful watching it must have been missed altogether. As it was, it differed in every respect from that of the preceding day. It was not nearly so high, not nearly so large, and lay in a different place on the planet, being now in longitude 31°.7, latitude 25°.5. Either the two, therefore, were totally different things or the projection had moved in the elapsed interval of time over seven degrees of latitude and eight degrees of longitude, a distance of three hundred and ninety miles in twenty-four hours. Where the previous projection had been nothing showed. On the following night, May 28, no trace of anything unusual could be seen anywhere. We are now concerned to inquire to what this series of appearances could have been due. The first observers of projections on Mars had unhesitatingly attributed them to the same cause that produces projections on the Moon, to wit, mountains. Such they were held to be in France and at the Lick. This view, however, was in 1892 disputed by W. H. Pickering who considered them to be not mountains, but cloud. And this view was strongly supported by A. E. Douglass in a discussion of a large number of them observed in 1894 at Flagstaff. The mountain theory of their generation was finally shown to be untenable and their ascription to cloud conclusively proved to be the correct solution by the observations of a remarkable one made in December, 1900, and the careful study to which by the writer they were subjected. We shall now explain how this was done and we will begin by pointing out that the fact that only a single specimen of the phenomenon was visible at each of the three oppositions of 1900, 1903, and 1905 was itself conclusive, rightly viewed, of their non-mountainous character. This conclusion follows at once from the isolateness of the phenomenon. For a mountain cannot change its place. Now, the shift in the aspect presented by the planet’s disk from one night to the next is not sufficient to alter perceptibly the appearance shown by anything upon its edge on the two occasions. If, then, a peak stood out upon it one evening, the peak should again show salient when the region reached the same position upon the succeeding night. That nothing then was seen where something had previously been visible proved the phenomenon not that of a mountain peak, since what produced the projection was clearly not fixed