important for bringing out the characteristics that give the stamp of artificiality, the straightness and fineness of the lines-would ever have any doubt as to their seeming artificial, however he might choose to blind himself to the consequences. An element akin to the comic enters criticism based, not upon what the critics have seen, but upon what they have not. Books are reviewed without being read, to prevent prejudice; but it is rash to carry the same admirable broad-mindedness into scientific subjects.

In detail the doubles vary, chiefly, it would seem, in the distance the twin lines lie apart. In the widest I have seen, the Ganges, six degrees separate the two; in the narrowest,the Phison, four degrees and a quarter,--not a very great difference between the extremes. Four degrees and a quarter on Mars amount to 156 miles; six degrees, to 220. These, then, are the distances between the centres of the twin canals. Each canal seems a little less than a degree wide, or about 30 miles in the narrower instances; in the broader, a little more than a degree, or about 45 miles. Between the two lines, in the cases where the gemination, as it is called, is complete, lies reddish-ochre ground similar to the rest of the surface of the bright regions. Deducting the two half-widths of the bordering canals, we have, therefore, from 120 to 175 miles of clear country between the paralleling lines.

This gemination of a canal is certainly a passing strange phenomenon. Although, in steady air, the observation is not a difficult one, to see the region where it occurs minutely enough for a sufficient length of time to mark the details of the process is another matter. I shall here give what I have been able to gather at the last opposition, and shall hope to add to it at the next. One element of mystery may be eliminated at the outset. The process is not so sudden as it seems. It is perceived of a sudden by the observer because of some specially favorable night. But it has been for some time developing. So much is apparent from my observations. Suggestions of duality occurred weeks before the thing stood definitely revealed. Furthermore, the gemination may lie concealed from the observer some time after it is quite complete, owing to lack of favorable atmospheric conditions. For it takes emphatically steady air to see it unmistakably.

The next point is, that the phenomenon is individual to the particular canal. Each canal differs from its neighbor not only in the distance the lines lie apart, but in the time at which duplication occurs. The event seems to depend both upon general seasonal laws governing all the duplications, and upon causes intrinsic to the canal itself. Within limits, each canal doubles at its own good time and after its own fashion. For example, although it seems to be a rule that north and south canals double before east and west ones, nevertheless, of two north and south lines, one will double, the other will not, synchronously with a doubling running east and west; the same is true of those running at any other inclination.

Now this shows that the duplication is not an optical illusion at this end of the line; for, by any double refraction here, all the lines running in the same direction over the disk should be similarly affected, which they are not. On the contrary, there will be, say, two cases of doubling in quite different directions coexistent with several single canals that run the same way.

Nor is there any probability of its being a case of double refraction at the other end of the line,--that is, in the atmosphere of Mars; for in that case it is hard to see why all the lines should not be affected, to say nothing of the fact that, to render such double refraction possible, we must call upon a noumenon to help us out, as we know of no substance capable of the quality upon so huge a scale. Furthermore, what is cogent to the observer, though of no particular weight with his hearers, the phenomenon has no look of double refraction. It looks to be, what it undoubtedly is, a double existence.

Strengthening this conclusion is the mode of development of the doubling. This appears to take place in two ways, although it is possible that the two are but different instances of one and the same process. Of the first kind, during this last opposition, the Ganges was an example.

The Ganges was in an interesting protoplasmic condition during the whole of last summer. About to multiply by fission, it was not at first evident how this would take place. Hints of gemination were visible when I first looked at it in August. It showed then as a very broad but not dark swath of dusky color, of nearly uniform width from one extremity to the other, with sides suggestively even throughout. It is