To the large spots, those of the first class, fall the places of intersection of the largest and most numerous canals, while the little spots make termini to fainter lines, ones that bear to them a like ratio of unimportance. Spots and lines are thus connected not simply in position but in size. The one is clearly dependent on the other, the importance of the centre being gauged by the magnitude of its communications. From the fact of association we now pass to the manner of it, which is quite as remarkable. The position of the spot relative to its tributary canals depends upon the character of the connecting lines. If the canal be single it runs, so far as may be judged, straight into the middle of the oasis, or, in other words, the oasis is symmetrically disposed about its end. This is true of the greater number of the large spots and of all the little ones, since the latter have as connections only single canals. In the case of a double canal arriving at a spot, a different and most curious dependence is observable. This fact I first noticed in a general way at the opposition of 1896-1897, the initial appearance of it being presented on September 30, 1896, by the Coloe Palus and the Phison. It was again visible in the case of the spots in the Trivium at the time the canals leading to that place doubled in March, 1897. But the exact nature of the phenomenon was not fully appreciated till 1903, when the thing was seen so well as to appear cut on copper plate. It was this: the spot is exactly embraced between the two arms of the double canal. It is, moreover, seemingly perfectly round and just fits in between the parallel lines. The Ascraeus Lucus was the first spot that showed thus in association with the double Gigas. Others followed suit in so showing, several specimens presenting themselves so well as to leave no doubt of the precise connection. The sight presented by such a spot and its incasing double is a beautiful bit of detail, perhaps the most beautiful so far to be seen upon the Martian disk. The distinctness with which it stands out on occasion suggests a steel engraving, and shows how clear-cut the Martian features really are when our own air ceases from troubling and allows them to be at rest. Incidentally, we may note that this phenomenon alone serves to disprove the diplopic theory of the production of the double canals. For if a double were a single line seen out of focus, any spot upon it should be doubled too. [Illustration: _Ascraeus Lucus and Gigas.—March. 2. 1903._]