So far in our account of the phenomena we have regarded the lines, the spots, and everything that is theirs solely from the point of view of their appearance at any one time. In other words, we have viewed them only from a static standpoint. In this we have followed the course of the facts, since in this way were the canals first observed. We now come to a different phase of the matter,—the important disclosure, with continued looking, that these strange things show themselves to be subject to change. That is, they take on a kinematic character. This at once opens a fresh field of inquiry concerning them and widens the horizon of research. It increases the complexity of the problem, but at the same time makes it more determinate. For while it greatly augments the number of facts which must be collected toward an explanation of what the things are, these once acquired, it narrows the solution which can apply to them.

The fact of change in the Martian markings forces itself upon any one who will diligently study the planet. He will be inclined at first to attribute it to observational mistakes of his own or his predecessor’s making, preferably the latter. But eventually his own delineations will prove irreconcilable with one another, and he will then realize the injustice of his inference and will put the cause, where indeed it rightly belongs, on the things themselves. Confronted by this fact he will the more fully appreciate how long and systematic must be the study of him who would penetrate the planet’s peculiarity. Just as the recognition of something akin to seasonal change came to Schiaparelli, because of his attending to the planet with an assiduity unknown to his predecessors; so it became evident that to learn the laws of these changes and from them the meaning of the markings, there was necessary as full and as continuous a record of them as it was possible to obtain. For this end it was not enough to get observations from time to time, however good these might be, but to secure as nearly as might be a complete succession of such, day after day, month after month, and opposition after opposition. The outcome justified the deduction. And it is specially gratifying to realize that to no one have the method and the results thus obtained appealed with more force than to Schiaparelli himself.

Perseverance in scanning the disk long after the casual observer had considered it too far away for observational purposes, resulted in Schiaparelli’s detection of the canals, and this through a characteristic of theirs destined to play a great part in their history, their susceptibility to change. He tells us in his Memoria I how Aeria and the adjoining regions showed blank of any markings while the planet was near in 1877 and the disk large and well shown, and then how, to his surprise, as the planet got farther away and the disk shrank, lines began to come out in the region with unmistakable certainty. Thus to the very variability which had hidden them to others was due in Schiaparelli’s hands their initial recognition.

Flux affecting the canals was apparent from the outset of my own observations. No less the subject of transformation than the large dark regions was the network of tenuous lines that overspread them. At times they were very hard to make out, and then again they were comparatively easy. Distance, instead of rendering them more difficult, frequently did the reverse. Nor was the matter one of veiling. Neither our own atmosphere nor that of Mars showed itself in any way responsible for their temporary disappearance. It was not always when our atmospheric conditions were best that the lines stood out most clearly, and as to Martian meteorology there was no sign that it had anything whatever to do with the obliteration. Long before the canals were dreamt of, veiling by Martian clouds or mist had been considered the cause of those changes in the planet’s general features, which are too extensive and deep-toned wholly to escape observation even though none too clearly seen. It was early evident to me that they were not the cause of general topographic change, and equally clearly as inoperative in those that affected the canals. In short, nothing extrinsic to the canal caused its disappearance; whatever the change was, its action lay intrinsic to the canal itself.

On occasion canals in whole regions appeared to be blotted out. The most careful scrutiny would fail to disclose them, where some time before they had been perfectly clear. And this though distance was at its minimum and definition at its best. Even the strongest marked of the strange pencil lines would show at times only as ghosts of their former selves, while for their more delicate companions it taxed one’s faith to believe that they could ever really have existed. Illumination was invoked to account for this, and plays a part in the effect undoubtedly. For at plumb opposition the centre of the disk for two or three years has shown less detail than before and after that event. This is probably due not, as with the moon, to the withdrawal of shadows, but to the greater glare to which the disk is then subjected. But this is not the chief cause of the change.

Still more striking and unaccountable was the fact that each canal had its own times and seasons for showing or remaining hid. Each had its entrances upon the scene and its exits from it. What dated the one left another unaffected. The Nilokeras was to be seen when the Chrysorrhoas was invisible, and the Jamuna perfectly evident when the Indus could scarcely be made out.

So much shows in the two drawings here reproduced. The increase of the Ganges and the advent of the Chrysorrhoas are noticeable in the second over the first.

Seasonal changes seemed the only thing to account for the phenomena. And in a general sense this was undoubtedly the explanation. To learn more about the matter, to verify it if it existed, and to particularize it if possible, I determined to undertake an investigation permitting of quantitative precision in the case. A method of doing this occurred to me which would yield results deserving of consideration from the amount of data upon which each was based and capable of being compared with one another upon an equal footing from which relative information could be derived. It seemed wise to determine from the drawings the degree of visibility of a given canal at different seasons of the Martian year, and then to do this for every important canal during the same period of time. The great number of the drawings suggested this use to which they might be put. For from a great accumulation of data a set of statistics on the subject could be secured in which accident or bias would be largely eliminated and the telling effect of averages make itself felt.

To render this possible it was necessary that the drawings should be alike numerous, consecutive, and extended in time. These conditions were fulfilled by the drawings made by me at the opposition of 1903. Three hundred and seventy-two drawings had then been secured, and they covered between them a period of six months and a half. They were also as consecutive as it was possible to secure. During a part of the period the planet was seen and drawn at every twenty-four hours, from April 5, namely, to May 26, or for forty-six consecutive days. Though the rest of the time did not equal this perfection, no great gap occurred, and one hundred and forty-three nights were utilized in all. Furthermore, as these drawings were all made by one man, the personal equation of the observer—a very important source of deviation where drawings are to be compared—was eliminated.

But even this does not give an idea of the mass of the data. For by the method employed about 100 drawings were used in the case of each canal, and as 109 canals were examined this gave 10,900 separate determinations upon which the ultimate result depended. That each of these determinations was independent of the others will appear from a description of the method itself on which the investigation was conducted. To understand that method one must begin a little way back.

As the two planets, Mars and the Earth, turn on their axes the parts of their surfaces they present to each other are constantly changing. For a feature on Mars to be visible from a given post on earth, observer and observed must confront each other, and, furthermore, it must be day there when it is night here. But, as Mars takes about forty minutes longer to turn than the Earth, such confronting occurs later and later each night by about forty minutes, until finally it does not occur at all while Mars is suitably above the horizon; then the feature passes from sight to remain hidden till the difference of the rotations brings it round into view again. There are thus times when a given region is visible, times when it is not, and these succeed each other in from five to six weeks, and are called presentations. For about a fortnight at each presentation a region is centrally enough placed to be well seen; for the rest of the period either ill-placed or on the other side of the planet.

If a marking were always salient enough it would appear in every drawing made of the disk during the recurrent fortnights of its display. If it were weaker than this, it might appear on some drawings and not on others, dependent upon its own strength and upon the definition at the moment, and we should have a certain percentage of visibility for it at that presentation. While if it changed in strength between one presentation and the next, the percentage of its recording would change likewise. Definition of course is always varying, but if its value be noted at the time of each drawing this factor may be allowed for more or less successfully. Making such allowance, together with other corrections to produce extrinsic equality, such as the planet’s distance, which we need not enter upon here, we are left with only the marking’s intrinsic visibility to affect the percentages; that is, the percentages tell of the changes it has successively undergone and give us a history of its wax and wane.

From drawings accurately made it is possible to add to the accuracy of the percentage by noting in each, not only the presence or absence of the marking, but the degree of strength with which it is represented. This was done on the final investigation in the present case, and it was interesting to note how little difference it made in the result.

The longitude of each canal was known, and the longitude of the central meridian of each drawing was always calculated and tabulated with the drawing, so that it was possible to tell which drawings might have shown the canal. Only when the position of the canal was within a certain number of degrees of the centre of the drawing (60°) was the drawing used in the result, allowance being duly made for the loss upon the phase side. Each drawing, it should be remembered, was as nearly an instantaneous picture of the disk as possible. It covered only a few minutes of observation, and was made practically as if the observer had never seen the planet before. In other words, the man was sunk in the manner. Such mental effacement is as vital to good observation as mental assertion is afterward to pregnant reasoning. For a man should be a machine in collecting his data, a mind in coÖrdinating them. To reverse the process, as is sometimes done, is not conducive to science.

When the successive true percentages of visibility of a given canal had thus been found, they were plotted vertically at points along a horizontal line corresponding in distance from the origin to the number of days after (or before) the summer solstice of the Martian northern hemisphere. The horizontal distance thus measured the time while the vertical height gave the relative visibility. The points so plotted were then joined by a smooth curve. This curve reproduced the continuous change in visibility undergone by the canal during the period under observation. It gave a graphic picture of the canal’s change of state. It seemed, therefore, proper to call it the canal’s cartouche or sign manual.

In this manner were obtained the cartouches of 109 canals. Now, as the presence or absence of any canal in any drawing was entirely irrespective of the presence or absence of another, each such datum spoke only for itself, and was an entirely independent observation. The whole investigation thus rested on 10,900 completely separate determinations, each as unconditioned by the others as if it existed alone.

As every factor outside of the canal itself which could affect the latter’s visibility was taken account of, and the correction due to it as nearly as possible applied before the cartouches were deduced, the latter represent the visibility of the canal due to intrinsic change alone. In other words, they give not the apparent only but the real history of the canal for the period concerned.

Important disclosures result from inspection of the cartouches. This we shall perceive by considering what different curves mean in the case. If the canal were an unchangeable phenomenon, for any reason whatever, its cartouche would be a straight line parallel to the horizon of the diagram. This is evident from the fact that the visibility would then never vary. If, on the other hand, it were waxing and waning, and the wax or wane were uniform, the cartouche would be a straight line inclined to the horizontal; rising if the canal were increasing, falling when it decreased. Lastly, if the rate of change itself varied, the cartouche would be a curve concave or convex to the line denoting the time, according as the rate of change of the growth or decay grew greater or less.

To see this the more clearly, we may set over against the cartouche the canal character it signalizes:—

If the cartouche first falls and then rises, this shows the canal to have passed through a minimum state at the time denoted by the point of inflection; if it rises first and falls afterward, this betokens in the same way a maximum. Thus the cartouches reveal to us the complete history of the canals,—what changes they underwent and the times at which these occurred. The cartouche, then, is the graphic portrayal of the canal’s behavior. It not only distinguishes at once between the dead and the living, as we may call the effect of intrinsic change, but it tells the exact character of this change,—the way it varied from time to time, the epoch at which the development was at its minimum or its maximum for any given canal, and lastly, its actual strength at any time, thus giving its relative importance in the canal system. For the height of the curve above the diagrammatic horizon marks the absolute as well as the relative visibility and enables us to rank the canals between themselves.

Now, the first point it furnishes a criterion for is the real or illusory character of the canals. If a line be due to illusion, whether optical or physical, it can vary only from extrinsic cause, since it has no intrinsic existence. If, therefore, all extrinsic cause be allowed for, the cartouche of this ghost must needs be a horizontal straight line. Even if the extrinsic factors to its production be imperfectly accounted for, their retention could only cause systematic variations from the straight line in all the lines, which would themselves vary systematically, and these factors could therefore be detected.

This criterion is absolute. Unless all the cartouches were approximately straight lines, no illusion theory of any kind whatever could explain the facts. Even then the lines might all be real; for unchangeable reality would produce the same effect on the cartouches as illusion. In the case therefore of horizontal straight line cartouches, we should have no guarantee on that score of reality or illusion; but, on the other hand, curves or inclined straight lines in them would be instantly fatal to all illusion theories.

Turning now to the 109 cartouches obtained in 1903, the first point to strike one’s notice is that all but three of them are curves and that even these three must be accepted with a caveat. Here, then, the cartouches dispose once and for all of any and every illusion theory. They show conclusively that the canals are real objects which wax and wane from some intrinsic cause.

The second result afforded by the cartouches is not of a destructive, negative character,—however valuable the destruction of bars to knowledge may be,—but of a constructive, positive one. It does not, like the first, follow from mere inspection, but is brought to light only by comparison of all the cartouches. In a positive way, therefore, its testimony is as conclusive as it was in a negative direction. For that 10,900 separate and independent data should result in a general law of development through either conscious or unconscious bias, when those data would have to be combined in so complicated a manner for the result to emerge as is here the case, is impossible. Chance could not do it and consciousness would require a coÖrdinate memory, to which Murphy’s nine games of chess at once would be child’s play.

Of the 109 canals examined 106 showed by their cartouches that they had been during the whole or a part of the period in a state of change. But the change was not the same for all. In some the minimum came early; in others, late. Some decreased to nothing and stayed there; others increased from zero and were increasing still at the time observations closed.

Latitude proved the means of bringing comparative order out of the chaos. When the canals were ranged according to their latitude on the planet, a law in their development came to light. To understand it, the circumstances under which the canals were presented must be considered as regards the then season of the planet’s year. In 1903 the planet passed on February 28 through the point of its orbit where the summer solstice of the northern hemisphere occurs. One hundred and twenty-six days later took place the first snowfall in the arctic and subarctic regions, an event that denoted the beginning of the new polar cap; from which date the snow there gradually increased. Its autumnal equinox the planet did not reach till August 29. Now, the canals were observed from thirty-six days before the summer solstice of the northern hemisphere to one hundred and forty-seven days after that event. We may tabulate the dates as follows:—

The vernal longitude is the longitude of the planet in its orbit reckoned from the vernal equinox. From the table it appears that the cartouches cover the development of the canals from about June 6 to September 1 of the Martian northern hemisphere for the current but to us undated year, ab Marte condita.

The 109 canals included all the more conspicuous canals on the planet at that opposition, all those that lent themselves by the sufficient frequency with which they were seen to a statistical result. They lay spread all the way between the edge of the polar cap in latitude 87° north to the extreme limit south, at which the then tilt of the north pole toward the earth permitted of canal recognition. This southern limit was in about latitude 35° south. Farther south than this vision became too oblique, amounting as it did, with an adverse tilt of twenty-five degrees to start with, to something over sixty degrees, for detection of such fine markings to be possible. Between the two limits thus imposed, by the perpetual snow on the one side and the observational tilt on the other, the 109 canals were distributed by zones as follows:—

As the latitude of a canal in the investigation was taken as that of its mid-point, such being the mean value of its successive parts, the latitudes about which information was obtained lay within the limits given above, the most northern canal, the Jaxartes N having for its mid-latitude 78° north, and the most southern, the Nectar, that of 27° south.

The zones comprised each a belt of territory about thirteen degrees wide, the first being less solely because in part occupied by the permanent polar cap.

The curves of all the canals in a given zone have been combined into a mean curve or cartouche for that zone; and then the cartouches for the several zones have been represented and ranged according to latitude on the accompanying plate. Consideration of these mean canal cartouches is very instructive. In the first place not one of them is a straight line, either horizontal or inclined. All are curves and, with the exception of the top one, all show a minimum or lowest point during the period under observation. From this point they rise with the time, or to the right on the plate. A black star marks this minimum, and is found farther and farther to the right as one goes down the plate; that is, as one travels from the neighborhood of the arctic regions down to the equator and then over into the planet’s southern hemisphere. Drawing now a line approximately through the stars and remembering that the minimum means the date at which the canal started to develop, we see that the canal development began at the border of the north polar cap and thence continued down the disk over the planet’s surface, as far as observation permitted the surface to be seen, which was some thirty-five degrees into the other hemisphere. This is the first broad fact disclosed by the cartouches.

Looking at them again we notice that the three topmost cartouches, those of the north polar, arctic, and sub-arctic canals respectively, dip at the right before the end of the observations, while the other seven were still rising when those observations were brought to a close. A reason for this, or at least a significant coincidence, is to be found in the dotted line pendent from the top of the table and labelled “First Frosts.” This dotted line denotes the date at which the first extensive frost occurred in the polar regions; for even before this time patches of white had appeared north of the Mare Acidalium, denoting the on-coming of the cold. The frost did not last but came and went and came again just as it does on earth, growing more insistent and long-lived at each fresh fall. Its sphere of operation was confined to the three zones in question. Even these zones it by no means covered, merely blotching them in places with fungi-like patches of frost. Beyond them south it never extended during the period of the observations; indeed, it hardly entered the sub-arctic zone at all at this very beginning of the polar winter. For it was only August 20 then. The coincidence of the isotherm as betrayed by the deposition of frost with the dividing line between the canal-development curves that dip down at this season and those that still continue to rise is suggestive.

It becomes all the more so when the three cartouches are considered seriatim. The most polewards, the north polar one, had sunk to zero sometime before the first extensive frost occurred; the second, the arctic, did so later than its northern neighbor, probably just before the epoch in question; while the third, practically outside the zone of deposition, was behind both the others in its descent. Inspection of the drawings upon which the cartouches are based confirms an inference deduced from this: that it was cold that killed, not frost that covered, them, which was responsible for their obliteration. The drawings show that the canals ceased to be seen before the white patches were evident. Now this would be the exact behavior of vegetation. It would be killed, turned brown by freezing, and so rendered invisible to us against its ochre desert background, before the cold had grown intense enough to cover that ground with a solid white carpet of frost. At the opposition of 1905, however, the extreme northern canals were visible after the snow had covered all the country about them, being evident as lines threading the new cap.

These three cartouches furthermore show each a maximum, and what is significant the maximum occurs later in time for each, according as the zone lies remote from the pole. A red star marks this maximum and shows that the time of greatest development for the three zones was respectively:—

We now pass to the other curves, those that were unaffected by cold. Though in these the minima themselves show the law of latitudinal progression, the wavelike character of the advance is even better disclosed by the curves. As the eye follows them down the page, the advance of the wave to the right is plainly apparent. The slope of the wave is much the same for all, implying that a like force was at work successively down the latitudes.

It will be noticed next that in all the mean cartouches the gradient is greater after the minimum than before it. The curves fall gently to their lowest points and rise more steeply from them. Such profile indicates that the effects of a previous force were slowly dying out down to the minimum and that then an impulse started in to act afresh. This explains the attitude of the canals that died out. In them the effect of the old force shows as in the others, but no impulse came in their case to resuscitation.

It seems possible to trace this force to an origin at the south. For beginning with the north sub-tropic zone the gradient on the left shows less and less steep southward to the south sub-tropic zone. Such a dying-down swell is what should be looked for in an impulse which had travelled from the south northward, since the wave would affect the more northern zones last, and less of a calm period would intervene between the two impulses from opposite poles.

The cartouches, then, state that the canals began to develop after the greatest melting of the polar cap had occurred; that this development proceeded down the latitudes to the equator, and then not stopping there advanced up the latitudes of the other hemisphere. In the next place they show that in the arctic region the development was arrested and devolution or decay set in as it began to get cold there, the most northern canals being affected first. Finally, that a similar wave of evolution had occurred from the opposite pole some time before and had then passed away. And this evidence of the cartouches is direct, and independent of any theory.