Now, this curious evolution in depiction points, rightly viewed, to an absence of design. It shows that Schiaparelli started with no preconceived idea on the subject. On the contrary, it is clear that he shared to begin with the prevailing hesitancy to accept anything out of the ordinary. Nor did he overcome his reluctance except as by degrees he was compelled. For the canals did not change their characteristics from one opposition to another; the eye it was that learned to distinguish what it saw, and the brain made better report as it grew familiar with the messages sent it. In other words, it is patent from these successive maps that the geometrical character of the ‘canals’ was forced upon Schiaparelli by the things themselves, instead of being, as his critics took for granted, foisted on them by him. We have since seen the regularity of the canals so undeniably that we are not now in need of such inferential support to help us to the truth; but too late, as it is, to be of controversial moment the deduction is none the less of some corroboratory force. With the third period enters what has been done since Schiaparelli’s time. For that master was obliged, from failing sight, to close his work with the opposition of 1890. In 1892 W. H. Pickering at Arequipa was the chief observer of the planet and made two important discoveries: one was the detection of small round spots scattered over the surface of the planet and connected with the canal system; the other the perception of what seemed to him more or less irregular lines traversing the Mare Erythraeum. Both were notable detections. The first set of phenomena he called lakes, the second river-systems, sometimes schematically ‘canals,’ but without committing himself to canaliform characteristics as his drawings make clear. The same phenomena were seen at that opposition at the Lick, by Schaeberle, Barnard and others, and called streaks. These discoveries took from the _maria_ their supposed character of seas—a most important event in knowledge of Mars. [Illustration: Map IX. Lowell, 1894.] [Illustration: Map X. Lowell, 1896.] [Illustration: Map XI. Lowell, 1901.] [Illustration: Map XII. Lowell, 1905.]