probable that they were then, as afterward, parallel, and that the slight convergence apparent at the bottom was due simply to foreshortening. The swath ran thus north-northwest all the way from the Gulf of the Dawn to the Lacus Labeatis. By moments of better seeing, its two sides showed darker than its middle; that is, it was already double in embryo, with a dusky middle-ground between the twin lines.

In October the doubling had sensibly progressed. The double visions were more frequent, and the ground between the twin lines had grown lighter. By November the doubling was unmistakable, and the mid- clarification had become nearly complete. It is to be remarked that the doubling did not involve the Fons Juventae and the canal leading to it, both of which lay well to the right of the Ganges. The space included between the East and West Ganges was very wide, some six degrees. The canals themselves were, so far as could be seen, quite similar, and about a degree, or 37 miles, wide. Both started in the Gulf of the Dawn, and ran down to the lower Lake of the Moon, one entering each side of the lake or oasis. Two thirds of the way down, both similarly touched the sides of another oasis, an upper Lacus Lunae; the other I have called the Lacus Labeatis. The length of each canal was 1200 miles.

Except for fleeting suspicions of gemination, and for possible doublings like the parallelism of the two Hades, the next canal to show double was the Nectar, which was so seen by Mr. Douglass on October 4, and under still better seeing, a few minutes later, the doubling was detected by him extending straight across the Solis Lacus. In the Solis Lacus this was evidently a case of mid-clarification. What occurred in the Nectar seems more allied to the second class of manifestations, such as happened later with the Euphrates and the Phison.

Glimpses of a dual state in these canals we caught during the summer and autumn, but it was not till the November presentation of the region that they came out unmistakably twinned. On the 18th of that month, just as the twilight was fading away, the air being very still and the definition exceptional, so soon as the sunset tremors subsided, the Euphrates and its neighbor the Phison I saw beautifully doubled, exactly like two great railroad tracks with bright ground between, each set extending down the disk for a distance of 1600 miles.

After that evening, whenever the seeing was good enough, they continued to present the same appearance. Now, with them no process of midway clarification, such as had taken place in the Ganges, had previously made itself manifest. They had, indeed, not been very well defined before duplication occurred, but apparently sufficiently so not to hide such broadening had it taken place; for, though the twin canals were not as far apart as the two Ganges, they were quite comparably distant, being, instead of six, about four and a quarter degrees from each other. Evidently, the process was, in the case of the Euphrates at least, under way in October, and even earlier, but was not well seen because the twin canals were not yet dark enough.

There seem, I may remark parenthetically, to be two other double canals in the region between the Syrtis Major and the Sabaeus Sinus, one to the east of the Phison, and another between the Phison and the Euphrates, both debouching at the same points as the Phison and the Euphrates themselves. On the 19th of November I suspected duplication in the Typhon, another canal in the same region. It looked to be double, with dusky ground between.

On the 21st I similarly suspected the Jamuna and the Dardanus. Both looked broad and dusky, with very ill-defined condensation at the sides. But the seeing was not good enough On the 22d I brought my observations to an end, in consequence of having to return East.

Exactly what takes place, therefore, in this curious process of doubling, I cannot pretend to say. It has been suggested that a progressive ripening of vegetation from the centre to the edges might cause a broad swath of green to become seemingly two. There are facts, however, that do not tally with this view. For example, the Ganges was always broad, but fainter, not narrower, earlier in the season. The Phison, on the other hand, went through no such process. Indeed, we are here very much in the dark, certainly very far off from what does take place in Martian canal gemination. Perhaps we may learn