III
Turning from such generic characteristics to more specific traits, the first thing to strike an attentive observer is that the doubles differ in width; that they are not mensurably alike in the property they hold in common of being paired. In some the twin lines are obviously farther apart than in others, and the relation persists however repeated the observations. Of two doubles the one will always surpass its fellow. This contrasted individuality first struck me in the Phison and the Euphrates; and from the first moment at which these doubles showed as such. The Phison pair seemed perceptibly the narrower of the two. A like distinction was evident at the next opposition and the next; in fact, at every succeeding one to the present day. Nor was the recognition of the fact confined to me. If we turn to Schiaparelli’s Memoriae we shall find that that master had registered the same idiomatic width for the two canals from first to last throughout his long series of records. The observation thus made proved to apply to each and all of these curious twins.
Diversity in width for different doubles appears plainly in drawings where more than one double is depicted. As an example, two drawings are here given in the text, made, the one on July 13, 1905, λ15°, and the other on July 20, λ313°. In them the Phison, Euphrates, Djihoun, and Thoth appear contrasted as unmistakably as either of them does with the single canals apparent at the same time. That this drawing is typical is borne out by all the best measures of the several doubles as seen at successive oppositions, and marshaled in the subjoined list. How truly individual the quality is stands proved by the relative values in different years which are even more accordant than the absolute ones.
The canals were:—
| Width | ||||
| 1903 | 1905 | Mean | ||
| 1. | Phison | 3.5 | 3.4 | 3.4 |
| 2. | Euphrates | 4.0 | 4.2 | 4.1 |
| 3. | [*]Protonilus | 2.8 | 2.0 | 2.4 |
| 4. | Deuteronilus | 2.2 | 2.4 | 2.3 |
| 5. | Pierius | 2.5 | — | 2.5 |
| 6. | Callirrhoe | 2.5 | [*]2.1 | 2.3 |
| 7. | [*]Hiddekel | 3.8 | 4.9 | 4.3 |
| 8. | [*]Gihon | 3.9 | 4.9 | 4.4 |
| 9. | Djihoun | 2.0 | 1.9 | 1.9 |
| 10. | Sitacus | 3.8 | [*]3.3 | 3.6 |
| 11. | Jamuna | 4.5 | — | 4.5 |
| 12. | Ganges | 5.0 | 5.2 | 5.1 |
| 13. | Nilokeras I and II | 11.0 | 11.7 | 11.3 |
| 14. | Nilokeras I | 2.3 | — | 2.3 |
| 15. | Gigas | 3.5 | — | 3.5 |
| 16. | Laestrygon | 2.2 | — | 2.2 |
| 17. | Cerberus N. | 4.0 | — | 4.0 |
| 18. | Cerberus S. | 4.0 | — | 4.0 |
| 19. | Cyclops | 2.9 | [*]2.2 | 2.6 |
| 20. | Nar | 2.6 | 2.0 | 2.3 |
| 21. | Fretum Anian | 2.8 | — | 2.8 |
| 22. | Aethiops | 3.3 | — | 3.3 |
| 23. | Eunostos | 2.8 | — | 2.8 |
| 24. | Lethes | 2.9 | — | 2.9 |
| 25. | Marsias | 3.2 | — | 3.2 |
| 26. | Hyblaeus | 3.0 | — | 3.0 |
| 27. | Amenthes | 3.2 | 3.5 | 3.3 |
| 28. | Thoth | 2.8 | 2.3 | 2.5 |
| 29. | Nepenthes | 2.8 | 2.3 | 2.5 |
| 30. | Triton | 2.7 | [*]2.3 | 2.5 |
| 31. | Pyramus | 2.9 | [*]2.0 | 2.5 |
| 32. | Astaboras S. | 3.2 | 3.1 | 3.1 |
| 33. | Tithonius | 2.6 | 2.2 | 2.4 |
| 34. | Vexillum | 3.5 | 2.9 | 3.2 |
| 35. | Tartarus | — | 2.7 | 2.7 |
[*] Poor.
Here we have widths ranging from eleven degrees to two. The widths given are those when the canal was at or sufficiently near its full strength, and are measured from the centres of the constituents. We notice two points: the agreement of the same canal with itself and its systematic disagreement with others. But what is especially to the point, if we compare the values found at successive oppositions, we find that for different canals the values agree in their difference. This shows that each of these values is, in most cases if not in all, a norm for that particular canal; a value distinctive of it and to which it either absolutely or relatively conforms. In other words, the width of the gemination is a personal peculiarity of the particular canal, as much an idiosyncrasy of it as its position on the planet.
Two general classes may be distinguished; those up to about five degrees in width apart and those above this figure. Whether such very widely separated lines as go to make up the second class, such as the Nilokeras I and II, constitutes a double is a debatable point. Schiaparelli thought they did, and so classed them. To me it did not at first occur so to consider them, and in some instances, such as the Helicon I and II, later observations seem to justify the omission. With the Nilokeras I and II the outcome seems the other way. The reasons for distrust of a physical relation between the constituents is not so much the distance separating them, nor any lack of parallelism, as the self-sufficient manner in which they show alone. Even this, however, tends to be recognized in the narrower pairs as they come to be better seen. It may be that width alone is wholly competent to selective showing. For the farther apart two lines are on the planet, the more opportunity is afforded the air waves to disclose the one without the other, a relative revelation which is constantly happening to detail in different parts of the disk. As long as any doubt exists of a physical community of interest, it seems best to distinguish such possibly merely parallel canals by suffixed numerals.
Of this class of doubles is the Nilokeras I and II. So wide is it that Mr. Lampland succeeded in photographing it as such, the two constituents showing well separated, and if it prove a true double it will be the first Martian double to leave its impress on a sensitive plate. Although separated by four hundred miles of territory, the two lines are parallel so far as observation can detect, which, of course, is not so very easy with the lines so far apart. In the country between one crosswise canal certainly lies, the Phryxus, and much shading thus far unaccounted for. Recent discoveries, however, point to the cause of such shading as lines imperfectly seen. For in some cases the lines have actually disclosed themselves, and warrant us in believing that it is only imperfect seeing that keeps the others hid. Of the pair the Nilokeras I is itself double, curiously reproducing what sometimes is seen in the case of double stars, one of whose components turns out to be itself a binary. The second line of the Nilokeras I lies close to its primary on the north, and was on the only occasion of its detection the merest of gossamers, while the Nilokeras I itself stood out strong and dark. Thus do these Martian details increase and multiply in intricacy the better the seeing brings them out.
In the case of the other doubles, the doubles proper so to speak, there is every indication of a physical bond between the pair. What that bond may be is another matter and seems to be of different divulging, according to the particular instance. At one end of the subject, both as the widest of these doubles and one of the most important, stands the Ganges. The components of the canal are 5°.1 apart. This great width, joined to the fact of scant extension, gives the canal a stocky aspect, its breadth being but one sixth of its length. Its width draws attention to it while the phenomena it exhibits intrigue curiosity.