The Atlantic Monthly/Volume 53/Number 318/The Red Sunsets

way in which men take the unexpected is an excellent gauge of their state of mind. Sudden changes in the face of nature bring the man out of his burrow and exhibit his hidden motives. Fifty years ago the meteoric shower of 1833 showed that the less cultivated class, even in America, still looked upon the accidents of the skies as signs of heavenly wrath and portents of coming ills. Now even the least educated no longer ask, What does this presage? but What is its cause? Naturalism has advanced fast and far in the last century.

The autumn of 1883 will always have a large place in scientific history on account of the strange aspect of its heavens, as well as on account of the preceding eruptions of volcanoes in the Straits of Sunda, which in their grandeur and effects much exceeded any disturbance recorded in history. Although the volcanic outburst changed the geography of a large district, destroyed somewhere near one hundred thousand lives, and sent the ocean waves and the throbs of the air produced by the convulsion over the whole circumference of the earth, its nature was not unusual; it differed from a thousand similar accidents of this troubled world only in degree, — only as the discharge of a twenty-inch cannon differs from that of a small field-piece. But the strange heavens of the later autumn, the fiery glow of sunrise and sunset, the brownish haze that girdles the sun all day, are phenomena so out of the range of common experience that at first all the experts in meteorology were at sea in their explanations. At the outset, many of these students of the atmosphere turned naturally to the conjecture that some of the vagrant matter of space, such as we see in the comets or dust-like meteors, had been drawn down upon our atmosphere, and so enveloped the earth with a meteoric mist. Others looked upon these movements as a mere intensification of the afterglow, or second sunset, which is not an unfamiliar phenomenon in all extra-tropical regions at certain seasons of the year, particularly in the autumn, and which is probably due to the condensation of vapor in the upper regions of the atmosphere. Gradually, as the facts have been gathered in from all parts of the world, these explanations have been overthrown, and the sunsets have been proven to be in some way connected with the Javanese convulsion. At several points in Europe the new-fallen snow contains particles of volcanic dust essentially like those that fell upon the decks of ships near the point of eruption, and which presumably are the heavier bits that have descended from the dust-cloud in the upper air.

Still further, it has been shown that these curious appearances of the sky occurred more quickly in the district near the volcano than in regions remote from it. It is not easy to determine the precise times when the sunset and sunrise became so brilliant; for at first the phenomenon might seem accidental in its nature, and so not become recorded. Yet it is clear that at Rodriguez, Mauritius, and Seychelles, points from three thousand to three thousand five hundred miles west of Krakatoa, the red sunsets were seen on the 28th of August, within thirty-six hours after the eruption occurred. In Brazil, which is over ten thousand miles away, they appeared on the 30th of September. In Florida, thirteen thousand miles distant, on September 8th. It was noticed in England on the 9th of September, but in Sweden not until the 30th of November; each of these countries being about seven thousand five hundred miles from the point of eruption. The volcanic mist spread more rapidly in the tropical belt between the parallels of latitude in which Java lies than in the regions to the north and south of this line. Sweeping swiftly about the earth in this tropical belt, it seems to have been carried thence by some slower motion to higher latitudes.

These successions of occurrence, first near the point of disturbance, then in regions more remote, would of themselves be sufficient to establish some connection between the Java convulsion and the brilliant sunsets; but any doubt that might remain is removed by the fact that we have at least one instance of a similar convulsion, in the last century, which we can in the same way connect with a great eruption in Iceland. In 1783, Skapta Jokul, one of the greatest of our volcanoes, passed through a period of eruption which, for its energy, was the most violent ever known in any but a Javanese volcano. Shortly after this eruption occurred, the English skies put on the fiery aspect that our own have at present. In those days men still looked to the heavens for portents, and deep alarm took possession of the people. Mr. James Macaulay has noted the fact that the poet Cowper refers to these sunsets in his letters, as well as in the Task, Book II. line 58: —


 * “Fires from beneath, and meteors from above,
 * Portentous, unexampled, unexplained,
 * Have kindled beacons in the skies;
 * And Nature with a dim and sickly eye
 * To wait the close of all;”

and Mrs. Somerville, in her Physical Geography, called attention to the probable relation between the vapor and ashes thrown out by the Iceland volcano and the brilliant sunsets of Western Europe. Gilbert White, the well-known author of the Natural History of Selborne, also perceived the connection between these lurid skies and the great eruption in Iceland; though he did not perceive the nature of the facts so clearly as did his able countrywoman.

In regard to the connection between volcanic eruptions and these skies as proven, we have next to consider the nature of the material that conveys the light down to us, the singular method in which it became diffused over the earth, and the reason for its long continuance. Here we are on more uncertain ground than in the first inquiry, yet with care we can find our way to the truth.

If the reader has examined these luminous skies with care, he will have observed that at midday, with an otherwise clear sky, the sun seems to be in a vast tract of thin whitish-brown vapor, looking like a thin mist, which is most evident a few degrees from the sun, and fades away insensibly, until at twenty degrees or less from the sun it imperceptibly melts into the apparently clear sky. Watching this faint cloud, we see that it is constantly changing its shape; dim streamers extend from it from time to time, and then fade away. Sometimes it is much stronger on one side of the sun than on the other. On several occasions it has appeared to be rapidly drifting to the northeast, with a speed comparable to the scud in a gale. These appearances, which I have not seen noted in any of the accounts of the sunsets, vary from day to day and hour to hour. They are explicable only on the supposition that there is a constant drifting of a very thin veil of this misty matter across the heavens near the sun. This matter, being intensely illuminated, is made visible in the region near the sun; elsewhere it is not dense enough to alter the blue of the sky. If we follow the descending sun, we find that when it begins to get into the mists of the horizon it no longer shows this ash-colored fringe, which melts into the dim, vaporous color that seems to encircle the horizon, but which is in fact due to the greater thickness and humidity of the air through which we then look. Nor do we see much of anything of these strange vapors in the first stages of the sunset, for there the glowing lower vapors still mask the upper light, it is after the normal sunset has fairly gone that this higher level of very faint cloud becomes illuminated. The long time that elapses after the sun goes below the horizon before these upper vapors find themselves at the right angle to reflect the light to us, and the long duration of this glow, show us that the volcanic vapor is much further above the earth than any common clouds. Computations based on the duration of this sunset light on the mists in question indicate that they must be somewhere near fifty thousand feet above the surface, or between nine and ten miles high. As the lightest ordinary clouds probably do not rise more than about thirty thousand feet above the earth, in northern regions, in the winter season, it is evident that the great height of these volcanic clouds is a part of the problem with which we have to deal.

There is one other important point to be described in order to have the whole matter before us. This is the color of the vapors. It is clear that these colors differ somewhat, but not notably, from the hues reflected from the usual clouds. The mist about the noonday sun is a little more brown than it would be if it were watery vapor alone, and the sunset glow of the cloud appears to be from a less lustrous surface than clouds of pure watery mist would afford. Moreover, the banding or stratification of the mist, though tolerably evident, is not so clear as it is in the case of ordinary cirrus clouds. It has been noticed in other countries, where these volcanic emanations were thicker than they are in the region about the North Atlantic, that the sun at morning and evening had a greenish color, which is never given it by the usual vapors of the atmosphere. Mr. Lockyer calls attention to the fact that on one occasion he observed such a color in the sun when it was seen through the steam of a steamship ; but this effect cannot be had through pure steam, though it is perhaps obtainable through such a mixture of steam and smoke as comes from the locomotive engine.

If these clouds were composed of dust alone, it is reasonable to suppose that they could not be banded or stratified, as cirrus cloud is; yet they exhibit some distinct traces of this banding. As their phenomena of color show that they are not water vapor alone, it is a fair conclusion that they are made up of a mixture of dust and water vapor, such as occurs in our chimney smoke. Our ordinary coal smoke is always composed in large part of steam, in which the little bits of carbon are mingled, as the soot is in the London fog. When dust of any kind becomes entangled in water vapor, the union is of a singularly permanent nature, the two being unwilling to separate until they fall as rain.

But it has often been asked of the present writer, How is it that these particles of mingled water and dust can remain so long at such a height above the earth? Why do they not fall at once to the earth, instead of floating to and fro, miles above its surface, for some months? To this there is a simple and apparently a sufficient answer, though it may not seem at first as evident as could be desired: the rate at which particles fall through the air is determined by the ratio that their superficies bear to their weight. Now the smaller any bits of matter are, the larger in proportion is their surface to their weight. A certain descending force is required to push the resisting atoms of air apart, and so permit the descent of the gravitating particle. It is this resistance that keeps the upper clouds floating so long and so high above the earth. The particles of water are constantly falling through the air, but owing to their fineness they may fall only a few inches each day. The same principle is shown in the settling of mud in water. A tumbler of Mississippi water will require days to deposit its mud. We have only to suppose that the particles of mingled dust and water that constitute these volcanic clouds are extremely small, to account for months, or even years, of suspension in the air.

Having now examined that part of the sunset phenomena that is evident to the eye, let us inquire how the dust and vapor was driven to such a height into the atmosphere, and so rapidly distributed over the earth. If we consider what takes place in any violent eruption of a volcano, we will see the explanation of these facts. In the case of this Krakatoa eruption, as in that of Skapta Jokul, indeed in all great eruptions, we easily see that the principal thing that occurs is a furious uprush of steam from the crater, bearing with it a vast quantity of pulverized rock, called dust or ashes. If it were the purpose of this article to explain the phenomena of volcanoes, it would be shown that the volcanic steam is the water that in old ages was inclosed in the small interstices of the rocks as they were formed on the ancient sea floors, and which became heated from the thick coating, or blanket, of other rocks deposited above. When by some chance fracture these gases of the buried water escape, they force quantities of the heated rock before them, as they rush into the air. As the imprisoned water completely penetrates the rock, on expanding it sends its walls into extremely minute fragments, which are borne upward in the rush of steam.

Some years ago, a skillful inventor devised an ingenious machine to reduce the ordinary Southern cane into a state of paper pulp, which consists of very finely divided woody fibre. He prepared a large cannon-shaped vessel, of great strength ; into this the cane was placed, along with some water; a strong lid closed the aperture; heat then being applied to the vessel, the imprisoned mass was brought to a very high temperature, say to twice the heat of boiling water; then, the lid being suddenly removed, all the water in the fibres of the cane was instantly converted into steam, and the mass, reduced to the finest shreds, was blown out of the muzzle of the gun-like vessel. The invention was never profitable, except to the geologist, who finds in it a capital illustration of the action that takes place in highly heated rocks when, by the rents at the volcano’s base, they are suddenly permitted to escape to the air. He knows that every crystal has water disseminated all through its structure, and this will cause it, when heated, to be reduced to an exceedingly fine powder as soon as the retaining pressure is removed.

The speed of this uprush from the crater of a great volcano is extremely great. Even from a volcano like Vesuvius, the vast, straight column of steam, blackened with ashes, rises to the height of twenty or thirty thousand feet above the base. When the force of the ascending column is overcome by the friction of the air, the steam spreads out like the top of a great Italian pine, and sails away before the wind. Those who have seen a large cannon at the moment of discharge have doubtless noticed the cylinder of smoke that is projected for a hundred feet or so beyond the mouth, and is then broken into swift-circling clouds. Now imagine a gun standing vertically, with its mouth a mile or more in diameter, and discharging its gases into the atmosphere with several times the speed with which they escape from a piece of artillery, and we will have the essential conditions of a volcanic explosion; only in place of the momentary outrush of the cannon we must imagine the explosion to endure for hours, or perhaps for days.

We have no very good data by which to determine the height to which the materials ejected from volcanoes are thrown. The strongest piece of modern artillery will, however, drive a ball straight upwards to the height of about four miles. it may easily be seen, even in small volcanoes such as Vesuvius, that more than this distance is attained by the substances which the eruption throws out. In great volcanoes, such as Krakatoa and many of those of Java and elsewhere, it may be that eruptions eject their matter to several times this height. Masses of considerable size, thrown out of volcanoes, have been known to fall four or five miles away from the crater. Allowing all that we can for wind carriage, it seems necessary to believe that these fragments must have had at least five or six times the speed of motion that we can impress on a cannon-ball, arid must have gone upward with nearly enough velocity to carry them beyond the sphere of the earth’s attraction.

If the observer could view the spectacle of such an eruption from a point well above the surface of the earth, he would see much that is hidden from those below by the wrap of clouds that quickly gather about the volcano. From such a vantage-point, say in a balloon, at the impossible height of sixty thousand feet above the earth, he would see the swift-moving column of steam and gas rising far above the level of our summer clouds, — ascending possibly, in such an eruption as that of Krakatoa, to the height of one hundred thousand feet above the sea. As this mass of mingled dust and steam rushed upwards, it would lean over to the westward, because of the greater eastward movement in the upper regions pf the atmosphere; but the most remarkable effect would be the very rapid horizontal diffusion of the gases in the thin upper air. In the nearly perfect vacuum which would exist around the upper part of the ejection column, these gases would hurry away in all directions with exceeding speed. This swiftly diffusing sheet of vaporous matter would, we may presume, quickly settle down upon the denser atmosphere below. The thicker the atmosphere the more slowly the matter would fall; the mist would be frozen, as is the water in all the higher-lying clouds, even on a summer day, and, entangling the volcanic dust in its meshes, would fall into the region of the air currents, and so journey over all the lands and seas.

If our imaginary observer from his lofty perch beheld an eruption that rose from the surface of the land, the ejection column before him would contain only the steam that came from the deep-buried rocks which are the seat of the volcanic impulse. But when, as was probably the case at Krakatoa, and is certainly so in many outbreaks, the eruption ascended from the sea floor, then, besides the steam that makes the eruption, there would be a large amount of sea water blown up with the ascending gas, which would enhance the mass of the material that found its way into the upper air.

It is not easy to conceive how vast is the volume of the gas thrown out by a great volcanic eruption. If we assume the area of the crater to he a mile square, the column to move upward with the speed of a mile a second, and the gas to have only the density of gunpowder gases within the chamber of a cannon at the moment of firing, as given by Rodman, then we have an amount about equal to all the atmosphere that lies on ten thousand square miles of the earth’s surface thrown out in a second of time. If we reduce the rate of the movement to that of a shot when it leaves a gun, we will still have about one third of this quantity. If all the gas discharged from a volcano stayed in the form of highly heated gas, then the pressure of the earth’s atmosphere would be doubled in about a fortnight, and even a day of eruption should add something like a pound to the pressure of the atmosphere on a square foot of surface.

The sudden movements of the barometer at points near the volcano of Krakatoa during the last eruption, amounting to an inch or so in height, show that a strong local effect on the atmospheric pressure is produced by the out-rush of gases. That no widespread or continuous effect upon the weight of the air is brought about is doubtless due to the fact that by far the greater part of the gas is steam, that is quickly condensed and falls back upon the earth in the form of rain, which in all such great eruptions deluges the region about the active volcano. The most of the dust — all the coarser grains of it, at least — that is thrown up by these eruptions returns also by gravity, or is borne down by the torrential rain, to the region about the base of the volcano. It is only the remnant of water and of powdered rock that remains high in the air, like the wrack of a thunderstorm, to float far away from the point where it was hurled into the air. Although the foregoing calculations have little definite value, they serve to show the reader how vast is the vaporous discharge in such an eruption, and how, even from its mere shreds and flying waste, the whole atmosphere of the earth may for a while put on a strange aspect.

So far we have been considering only the outward appearances given to our atmosphere by the last Java convulsion. Let us see if there are any other effects of it than the changing variety that the mornings and evenings have gained by the eruption. It is a familiar fact that the earth’s atmosphere is a singular, delicate mechanism, that moves with trifling impulses in the most varied ways. From its behavior during the Krakatoa eruption we may find one evidence of its sensitiveness to disturbing actions. Mr. Scott has shown that the barometric spasms that caused, during the eruption, the before-mentioned leaps of the mercury about Krakatoa were passed on through the atmosphere with the speed of some hundred miles or more an hour, until they encircled the earth; so that in about fifteen hours the remotest point on the earth had felt the shock of the explosions. But for reasons already given is is not likely that any permanent effect on the weight of the atmosphere can be produced by the volcanic gases. It is otherwise with the dust clouds that cause our golden sunsets. The fact that these particles of vapor and dust send us back the sunlight is proof that they cut off a share of the sun’s rays from their proper access to the earth’s surface. For months the earth has been wrapped in a veil that denies admission to a small part of the sun’s light, and presumably to a portion of his heat as well. Upon this heat all the machinery of the earth’s physical and organic life most intimately depends. Take but the hundredth part of it away, and all the life of the earth would feel the loss of power. The air currents would become feebler, the ocean streams less strong; thousands of animals and plants would find their conditions changed, so that the boundaries of the provinces they occupy would be altered, or even life itself abandoned after a few years of struggle. The first command which life would put upon the physical forces of the earth, if it were happily in its power to command, would be to “get out of my sunshine."

As yet we have no data on which to base any reckoning concerning the effect of this thin veil that enshrouds the world. At the end of January, the midday sun, in an otherwise cloudless sky, shines through a veil that considerably diminishes its light. May it not be that the remarkably steady cold of the past thirty days in this country is in some way due to this interference? Whether this be so or no can be settled only when we have the records of the temperature stations within the tropics. But every physical consideration leads us to believe that, though slight, there is some result from this action.

If the Krakatoa eruption could be assumed to represent to us the maximum of volcanic energy, the climatal influences of volcanic eruptions might not properly command our attention; but when we consider that the geological record makes it probable that there have been times in the earth’s history when disturbances of this class have been more frequent and on a far larger scale than at present, we are disposed to take a suggestion from this veiled sun, and ask ourselves whether some of those strange changes of climate in the past may not perhaps have had something to do with periods of intense volcanic activity. This supposition may not prove to have any great value, but such is the difficulty we have in explaining the changes that the climate of the earth has undergone in the geologic history that it is worth while to examine any events that promise to help us to understand how such alterations can take place. It appears possible that volcanoes may operate to change the climate of the earth in at least two ways. In the first place, the great amount of water in the form of steam that they hurl into the air tends to increase the rainfall over a wide region; the greater part of this water will fall near the volcano, but the effect will doubtless be of importance over a wide area. If this rain falls from skies that have any considerable amount of their sunlight fended off by the dust wrap that may be formed over them, this water will be apt to fall in the form of snow. If the dust wrap remained for any considerable time in the air, — and as far as we can see it might remain for several years, for this dust that our air now bears has been afloat for nearly half a year, with no sign of diminution, — then the chapter of accidents might lay the foundations of a glacial period which might endure long after the cause that led to its beginning had ceased to exist. It is not likely that any such great and enduring ice time as that which has just passed away from the earth could be due to volcanic dust and vapor, but it seems possible that to such accidents we may owe climatic changes of much consequence.

There is yet another interesting field of inquiry opened to us by the consideration of the Krakatoa convulsion. The first scientific observers of the red sunsets generally inclined to the opinion that they were produced by the falling upon the earth of some clouds of finely divided matter, such as are thought to give rise to the zodiacal light, — cosmic vapor, as it has been unhappily called. There can be no doubt that the celestial spaces — at least within the region that the earth traverses — are crowded with angular bits of stony matter, ranging in weight from thousands of pounds down to particles as light as the finest dust. Every night millions of the smaller bits fall swiftly upon our earth’s atmosphere, sparkle for a moment as shooting stars, and are burnt into vapor by the heat engendered from their friction in the atmosphere. It is a matter of difficulty to account for the origin of these angular fragments in space. Though they are found in every part of the space our earth passes through, they are most thickly gathered on certain belts, through one of which our earth passes in July, through another in October. The most likely conjecture as to the origin of these meteors that can be made is that in certain periods of particularly intense eruptions the ejection of volcanoes — those it may be of other planets, as well as of the earth — attain such an extreme velocity that they fly clear beyond the control of the orb from which they are projected, and are left to swing through space in orbits determined by the control of the sun. At times these bodies would perhaps come sufficiently into the sphere of the gravitation of a planet to be precipitated upon its surface; but the chance is that they would move on for ages before they neared any sphere with attraction strong enough to draw them to its surface.

To project stones beyond the earth’s power to recall them requires a velocity that need not exceed seven miles a second. We have no proof of such extreme speed of uprising in any volcanic eruptions, but there are many reasons for believing that it is not altogether beyond the power of the greater eruptions to accomplish this work.

There is yet another lesson that the Krakatoa convulsion has for us: that is, a lesson in favor of a little more humility on the part of those semi-scientific men who fancy that they know the mechanism of the world as a watchmaker knows the wheels of a watch. Despite all we have known of volcanoes, this Javanese explosion has shown us that they possess powers over the air which were unknown six months ago. This may fairly serve as a warning to those who suppose that we know all the change-bringing agents of the world. There are doubtless many forces that may have had their share in the ancient history of our earth that are as yet undreamt of in our philosophies.

There remains the question as to how long these dust clouds are to endure in the atmosphere. On that point we can make no answer. For three months they have been drifting over this sky, and to-day they appear to be as high above the earth as they were when they first came. If they own their high elevation to electrical repulsion, as is conjectured by Mr. Crookes and others, there seems no reason why they may not stay in the upper air for years; but if, as is more likely, they are slowly settling towards the lower levels of the atmosphere, they will before long come within the limits wherein the rain clouds gather, when they will quickly be dragged down by the action of the falling drops.

This view is rendered the more probable by the fact that while in November and December the red sunsets and the mist-encircled noonday sun were very constant phenomena, they are now, in the first days of February, scarcely perceptible after a heavy rain or snow storm, though they gradually return with less brilliancy after a few days of good weather. This seems to show that the volcanic matter has in good part fallen into the lower zone of our atmosphere, where it may become entangled in the descending rain or snow. My observations on the height of the sunset glow and the duration of the light show that the remaining dust floats at a lessened elevation above the earth; so it is likely that a few months more will bring the last of it to the ground.

When this volcanic dust ceases to glorify our skies at dawn and eve, we shall part with what has probably been the most remarkable and picturesque accident to the earth’s physical life that has been known with the limits of recorded history.