At the canal’s maximum and minimum the equality of its two constituents is chiefly to be remarked, though it occurs on other occasions as well. But, what is significant, when the two differ it is always the same one that outdoes its fellow. It may be the right-hand twin in one pair, the left-hand one in another; but whichever it be, for the particular canal its preëminence is invariable. It is this canal which, except for adventitious help or hindrance from the air-waves, alone shows when the double assumes the seemingly single state. We may therefore call it the original canal, the other being dubbed the duplicate. In some cases it has been possible to decide which is which. It might seem at first sight as if this point should always be ascertainable. But the determination is more dilemmic than appears, not from any difficulty in seeing the canal, but from the absence of distinguishing earmark at its end. In a long stretch of commonplace coast, the precise point of embouchure of a solitary canal cannot be so certainly fixed as to be decisive later between two which show close together in the same locality. It is only where some landmark points the canal’s terminal that the problem admits of definite solution. This telltale tag may be a bay like the Margaritifer Sinus, or double gulfs like the Sabaeus Sinus, or portions of a marking not too large to permit of partitive location like the Mare Acidalium, or a canal connection like the Tacazze which prolongs the one line and not the other. In these and similar instances the two lines become capable of identification, and in such manner have been found those comprised in the following list:—
| Double Canal | Original Line | Date of Ascertainment |
| Phison | The Eastern | 1894 |
| Euphrates | The Western | 1894 |
| Titan | The Western | 1896 |
| Hiddekel | The Eastern | 1896 |
| Gihon | The Western | 1896 |
| Gigas | The Northwestern | 1896 |
| Djihoun | The Western | 1901 |
| Laestrygon | The Eastern | 1903 |
| Nilokeras I and II | The Northern | 1903 |
| Astaboras | The Southern | 1903 |
| Jamuna | The Eastern | 1905 |
| Ganges | The Western | 1905 |
In this list of originals the canals stand chronologically marshaled according to date of detection. The Phison and Euphrates were the first to permit of intertwin identification in 1894, while the Jamuna and Ganges were the last to be added to the column in 1905. The list is not long, though the time taken to compile it was. In the case of the Ganges and the Jamuna, for example, although suspected for some time on theoretic grounds, it was only at the opposition just passed that the fact was observationally established. In his Memoria V, Schiaparelli has a list of similar detection, and if the present list be compared with his, the two having been independently made, the concordance of the result will prove striking, corroborative as it is of both. For the necessary observations are very difficult.
Having thus realized the original by means of its superior showing, and then identified it by its position, it is suggestive to discover that the duplicate betrays its subordinate character, not only by its relative insignificance, but by its secondary position as well. The original always takes its departure from some well-marked bay, seemingly designated by nature as a departure-point, or from a caret belonging clearly to itself; the adjunct, on the other hand, leaves from some neighboring undistinguished spot, as in the case of the additional Djihoun, or makes use of a neighbor’s caret, as in the case of the second Phison and the supplementary Euphrates. In either case it plays something of the part of an afterthought; and yet the postscript when finished reads as an integral part of the letter. An example will serve to make the connection evident while leaving the character of the connection as cryptic as ever.
In the long stretch of Aerial coastline bounding the Mare Icarium, which sweeps with the curve of a foretime beach from the Hammonis Cornu to the tip of the Edom Promontory, there stand halfway down its far-away seeming sea-front two little nicks or indentations. Even in poor seeing they serve to darken this part of the coast while in good definition they come out as miniature caret-like bays. They are the Portus Sigaei, and mark the spots where the Phison and the Euphrates respectively leave the coast. About four degrees apart, the eastern makes embouchure to the original Phison, the western to the original Euphrates, and each in some mysterious manner is associated not only in position but in action with the canal itself. In the single state each canal leaves the Mare from this its own caret, the Phison proceeding thence northeast down the disk, the Euphrates nearly due north, so that starting four degrees apart at the south they are forty degrees asunder at their northern termini. Clearly at these latter points they are not even neighbors, and except for the accident of close approach at their other ends have nothing in common anywhere. And yet when gemination takes place a curious thing occurs: each borrows its neighbor’s terminal as departure-point for its own duplicate canal. Having thus got its base the replica proceeds to parallel its own original canal without the least reference to the other canal whose own caret it has so cuckoo-wise appropriated. What the Phison thus does to the Euphrates, the Euphrates returns the compliment by doing to the Phison. In this manner is produced an interrelation which suggests, without necessarily being, an original community of interest; suggests it on its face and yet appears to be rather of the nature of an adaptation to subsequent purposes of a something aboriginally there.
Mouths of Euphrates and Phison.
June. 1903.
That such latter-day appropriation is the fact is clearly hinted by the behavior of another understudy of an original canal, in this case the duplicate of the Djihoun, which in consequence of the position of its original finds no neighboring embouchure already convenient to its use. The single or original Djihoun leaves the tip of the needle-pointed Margaritifer Sinus, which serves a like end to the Oxus and the Indus, both single canals. The Sinus is itself a single bay, and so large that for many degrees its shores on both sides converge smoothly to their sharp apex. Because of this probably, the coast in the immediate neighborhood is without canal connection, no canal being known along either side till one reaches the Hydraotes at the Aromaticum Promontorium, which marks the western limit of the gulf. The consequence is that when the Djihoun doubles, the duplicate canal, not having any terminus ready to its hand, has to make one for itself by simply running into the Margaritifer Sinus, some distance up its eastern side. It thus advertises its adjunctival character, and at the same time the general fact that a neighbor’s terminus, though used from preference, when convenient, is not an essential in the process. Gemination occurs of its own initiative, but is conditioned by convenience.
Whether one canal shows thus to the exclusion of the other, or whether both stand so confused as not to be told apart, the fact remains that the double is not always recognizable as such. If we turn to the list of the doubles on page [222], we shall note that the same canals were not always seen in the dual condition at successive oppositions. Some, indeed, are so emphatically of the habit as to appear year after year in a paired state, but others are not so constant to their possibilities. Now, when it is remembered that at different oppositions we view Mars at diverse seasons of its tropical year, we see that this means that the phenomenon is seasonal; and furthermore that its exhibition depends upon the canal’s position. Gemination, like the showing or non-showing of the single canal, is conditioned by the place of the canal upon the planet.