In 1894 (Map IX.) Douglass at Flagstaff discovered a fact which did for the dark regions what Schiaparelli's canals had done for the light ones. He found on scrutinizing the southern dark areas that Pickering's river-systems, which he too had seen in 1892, came out
again and proved to be but part and parcel of a systematic set of lines networking those areas. He noted that these lines were not only simple streaks, but that they had nothing irregular about them, and could not possibly, therefore, be the river-systems supposed by Pickering. The lines were straight, equable in width and connected with one another at
certain determinate points. They followed apparently arcs of great circles and covered all the dark regions which could be well seen with a singularly symmetric mesh. In other words, they presented all those strange, peculiar and enigmatic characteristics which distinguished the
