Contrast was the open sesame to their detection. When the _maria_ show dark, the lines are lost in the sombreness of the background. As the maria lighten the lines come out. Such was amply witnessed by the effect in 1894 and 1896. In 1894 I found it impossible to perceive them, except where the Padargus crossed Atlantis, for the hue of the _maria_ themselves was then very dark. In 1896, on the other hand, I saw them without difficulty. What is also of interest: so soon as seen they appeared small, without haziness or distention. As the oppositions succeeded one another the northern regions rose into view, and with their appearance came the detection in them of the same phenomena. No large dark areas like the diaphragm exist there, but the smaller patches of blue-green which bestrew them proved to be similarly meshed. At first canals were evident upon their peripheries, contouring them about; then the bodies themselves of the patches showed grid-ironed by lines. The Mare Acidalium with its adjuncts, the Lucus Niliacus on the south and the Lacus Hyperboreus on the north, thus stood out in 1901. On a particularly good evening of definition at the end of May, the Mare suddenly made background for a sunburst of dark rays, six of them in all radiating from a point between it and the Lacus Hyperboreus. Considering how sombre the Mare was at the time, this was as remarkable a vision as it was striking to see. Although at the moment the sight was of the nature of a revelation, these lines have been amply verified since, as the Martian season has proved more propitious. Similar decipherment has befallen all the other patches of blue-green in the northern hemisphere; these having shown themselves first circumscribed and then traversed by canals. Interesting instances were the Wedge of Casius and the Propontis. These markings, first perceived years ago as mere patches of shading, then partially resolved by Schiaparelli, now stood revealed as a perfect network of lines and spots. So many of both kinds of their detail occupied the ground that to identify them all was matter of exceeding difficulty. The outcome is shown in the diagrammatic representations opposite and on page 256. These drawings disclose better than any description the mass of detail of which the patches are in reality composed, and serve to convey an idea of the complexity involved. If the general canal system seems intricate, here is something which exceeds that as much again. [Illustration: The Propontis, 1905.]