Aristotle · a new plain-English translation from the original language
1| These, then, are the accounts others give of thunder and lightning: some say lightning is a reflection, others that lightning is a flashing-through of fire and thunder its quenching, the fire not coming to be at each occurrence but already present within. We, however, say that the same nature is, on the earth, wind, in the earth, earthquake, and in the clouds, thunder. For all these are, in substance, the same thing: a dry exhalation, which, when it flows in one way, is wind, and when it acts in another way, produces earthquakes, and when it changes within the clouds and is expelled, as the clouds come together and are condensed into water, produces thunder and lightning and, besides these, the other phenomena that are of the same nature as these. So much, then, has been said about thunder and lightning. 3.
2| Let us now speak of the remaining effects of this expulsion, describing them in the way already laid out. For this current of air, when it is expelled in small amounts, scattered abroad, occurring often, breathing through, and finer in its parts, produces thunder and lightning; but if it is expelled all at once and denser, and less of it comes out fine, it becomes a squall-wind. That is also why it is violent: for the speed of the expulsion produces the force. Now when a great and continuous expulsion follows along, the same thing happens as when the process again rushes to the opposite; for then there comes rain and an abundance of water. Both of these, then, exist potentially in the matter; and whichever of the two gets the start as an active power, that is what follows, being expelled together out of the matter, whichever kind is present in the greater quantity, and the result is, on the one hand, a rainstorm, and on the other, from the other exhalation, a squall. But when the current of air being expelled within the cloud meets resistance from another current in just the way that wind, forced from a wide space into a narrow one, does at gateways or roadways—for it often happens in such cases that, the first portion of the flowing body being pushed back because it cannot yield, either through narrowness or because of an opposing current, a circle and a whirl of the current of air comes about—
3| for what lies ahead prevents it from advancing, while what is behind pushes it on, so that it is forced to move sideways, where it is not hindered, and so continually what is next to it, until it becomes a single thing, and this is a circle; for that of which the movement follows a single shape must itself be a circle. On the earth, then, whirlwinds come about for these reasons, and likewise in the clouds, as regards their origin, except that, just as when a squall-wind comes about, the cloud is continually being expelled and a continuous wind results, so here too the continuous portion of the cloud continually follows. But because of density, the current of air, unable to be expelled from the cloud, at first turns in a circle for the reason stated, but is carried downward because the clouds are continually being condensed, at the point where the heat falls out. This affection, when it is without color, is called a typhoon, being a wind, as it were an undigested squall. In north winds no typhoon occurs, nor, when there is snow, does a squall-wind occur, because all these are currents of air, and a current of air is a dry and hot exhalation. So frost and cold, because they prevail, quench the beginning of the process as soon as it comes to be.
4| That cold does prevail is clear: for otherwise there would be no snow, nor would north winds bring moisture; for these come about when coldness prevails. A typhoon, then, comes about when what is becoming a squall-wind is unable to be expelled from the cloud. It occurs through the resistance of the whirl, when the coil, being carried onto the earth, drags the cloud down with it, unable to be released from it. In the direction in which it breathes out in a straight line, it moves things by its current, and by its circular motion it turns and lifts up whatever it forcibly strikes upon. When what is drawn down is set ablaze (this happens if the current becomes finer), it is called a prester; for it sets the air ablaze along with itself, coloring it by its burning. But if within the cloud itself a great deal of fine current is squeezed out, this becomes a thunderbolt: if it is very fine, not burning because of its fineness, this is what the poets call the bright bolt; but if less fine, it does burn, and this is what they call the smoky bolt. For the one, because of its fineness, is carried along, and because of its speed passes through before it can set things ablaze, though by lingering it blackens them; the slower one colors but does not burn, but rather it passes through too quickly. That is why some things that offer resistance are affected in some way, while others are not at all—for instance, on a shield, the bronze has already melted while the wood has suffered nothing at all;
5| for because of its porousness the current passed through and got through before it could act. Likewise through garments it does not burn them up but makes them, as it were, into rags; so that it is clear from cases like these too that all these things are currents of air. It is possible sometimes even to see this with one's own eyes, as indeed happened just now in the case of the temple at Ephesus when it was burning; for in many places the flame moved along continuously, then broke off separately. That smoke is a current of air, and that smoke burns, is evident, and has been said before elsewhere; and when it moves along all at once, then it clearly seems to be a current of air. What appears, then, in small fires, was happening then too, but much more powerfully, because a great deal of material was burning. As the timbers split apart, then, from which the current of air originated, a great deal moved along all at once, at the point where it was breathing out, and was carried upward set on fire, so that the flame appeared to be moving and falling upon the houses. For one must always suppose that a current of air accompanies thunderbolts and goes ahead of them; but it is not seen, because it is without color. That is also why, in the place it is about to strike, there is movement before the blow lands, since the beginning of the current of air falls upon it first. And thunder and lightning likewise split things apart, not by the noise, but because the current of air that produces both the blow and the noise is expelled at the same moment;
6| whatever it strikes, it splits apart, but it does not burn it. So much, then, has been said about thunder and lightning and the squall-wind, and further about presters and typhoons and thunderbolts, and that they are all the same thing, and what the difference between them is. Let us now speak about the halo and the rainbow, what each is and through what cause it comes about, and about mock suns and streaks; for all of these come about through the same causes as one another. First we must grasp the affections and what happens in connection with each of them. Of the halo, then, a whole circle is often seen, and it occurs around the sun and moon and the bright stars, and no less by night than by day, and around midday or afternoon; but less often at dawn and around sunset. Of the rainbow, a full circle never occurs, nor a segment greater than a semicircle; and at sunset and sunrise the circle is smallest but the arc greatest, while as the sun rises higher the circle is greater but the arc smaller; and after the autumn equinox, on the shorter days, it occurs at every hour of the day, but in summer it does not occur around midday. Nor do more than two rainbows occur at once. Of these, each has three colors, and the colors are the same and equal in number to one another, but fainter in the outer one, and placed opposite in their position. For the inner one has, as its first and outermost band, the largest, a crimson; the outer one has this same color as its smallest band but nearest to it, and the others correspondingly.
7| These colors are about the only ones that painters cannot produce; for some they can mix themselves, but crimson, green, and purple do not come about by mixing — and the rainbow has just these colors. The color between the crimson and the green often appears yellow. Mock suns and light-shafts always occur to the side, and never overhead, never close to the ground, never directly opposite the sun, and never at night — always around the sun, and further, while it is rising or setting; most often toward the west. When the sun is at mid-heaven it is a rare occurrence — as it once happened at the Bosporus: two mock suns rose with the sun and continued the whole day through until it set. These, then, are the things that happen with respect to each of them, and the cause of all of them is the same: all of these are reflection. They differ in their manner and in what they occur from, and in how the reflection happens to occur toward the sun or toward some other bright body. A rainbow occurs by day; by night, one from the moon, as the ancients thought, did not occur — they were mistaken about this because of its rarity, for it escaped their notice: it does occur, but only rarely.
8| The cause is that colors escape notice in the dark, and many other conditions must coincide as well, and all of these on a single day of the month; for it is necessary that a moon-rainbow, if one is to occur, occur at the full moon, and then only while the moon is rising or setting. That is why in more than fifty years we have come across it only twice. That sight is reflected — just as from water, so too from air and from everything whose surface is smooth — this conviction must be drawn from what is demonstrated in the works on sight, and likewise the fact that in some mirrors shapes as well as colors appear, while in others only colors appear. Mirrors of this latter kind are those that are small and have no perceptible division of parts; for in these it is impossible for a shape to appear (since it would have to appear as divisible, for every shape seems at once to be a shape and to have division), and since something must appear, and this is impossible, it remains that only the color appears. And the color of a bright object sometimes appears bright, and sometimes, either because it mixes with the color of the mirror or because of a weakness of sight, produces the appearance of a different color.
9| Let these matters be taken as already examined by us in the demonstrations in the works on the senses; hence we shall state some points and use others of them as established. Let us speak first about the shape of the halo — why it is a circle, and why it forms around the sun or the moon, and likewise around some of the other stars; for the same account will fit all cases. The reflection of sight occurs when the air and the vapor condense into a cloud, provided the condensation happens to be uniform and made of fine parts. Hence the condensation is a sign of rain, while the breakings-up or the fadings-away — the fadings being signs of fair weather, the breakings-up being signs of wind. For if the halo neither fades away nor breaks up, but is left to receive its own nature, it is reasonably a sign of rain; for it shows that a condensation of just this kind is already occurring, from which, as the thickening continues without interruption, it must necessarily result in rain. That is why these halos, more than the others, become black in color. When it breaks up, it is a sign of wind; for the division comes about through wind that already exists but is not yet present at that spot. A sign of this is that the wind comes from just that quarter from which the main break occurs. A halo fading away is a sign of fair weather; for if the air is not in such a condition as to master the heat trapped within it, nor to come to a watery condensation, it is clear that the vapor has not yet separated out from the dry and fiery exhalation —
10| and this is the cause of fair weather. We have now stated, then, in what condition the air must be for the reflection to occur. Sight is reflected from the mist that forms around the sun or the moon; that is why the halo does not appear opposite the sun, as the rainbow does. Since the reflection happens alike from every side, the result must necessarily be a circle or part of a circle; for from the same point to the same point, equal broken lines will always fall on the line of a circle. For let there be, from the point A to the point B, the broken lines A-C-B and A-Z-B and A-D-B; let AC, AZ, AD be equal to one another, and let the segments to B, namely CB, ZB, DB, also be equal to one another; and let AEB be joined, so that the triangles are equal (since they stand on the equal line AEB); then let perpendiculars be drawn to AEB from the angles — from C the line CE, from Z the line ZE, from D the line DE. These, then, are equal;
11| for they belong to equal triangles, and all lie in one plane; for they are all at right angles to AEB, and they all meet at the one point E. The figure traced out will therefore be a circle, with E as its center. Let B be the sun, A the eye, and the arc around C, Z, A the cloud from which sight is reflected toward the sun. One must think of the mirror-points as continuous; but because of their smallness each one is invisible, while the whole made up of all of them seems to be one, because of their being in unbroken succession. The white appears — the sun — continuously in a circle, appearing in each of the mirror-points and having no perceptible division, and more so near the ground because it is calmer there; for when there is wind, it is evident there is no stillness. Next to this the adjoining arc is black, seeming blacker because of the whiteness of the other. Halos occur more often around the moon, because the sun, being hotter, dissolves the condensations of air more quickly. Around the stars they occur through the same causes, but are not equally significant as signs, because they indicate condensations that are altogether small and not yet productive of anything.
12| That the rainbow is a reflection has been said before; what kind of reflection it is, and how and for what cause each of the things that occur in connection with it comes about, let us now state. Sight, then, when reflected, is found to occur from all smooth things, and among these are both air and water. It occurs from air whenever air happens to be condensing. But because of a weakness of sight, reflection often occurs even without condensation, as once happened as an affection to a certain man who saw only dimly and not sharply; for it always seemed to him, as he walked, that an image went ahead of him, looking toward him from directly opposite. He experienced this because his sight was reflected back to him; for it was so weak and altogether thin because of his infirmity that even the nearby air became a mirror for it, and it could not push through the air the way strong sight pushes through air that is far off and dense. That is why headlands appear lifted up at sea, and why all sizes look greater when east winds are blowing, and why things in mist look greater too — as the sun and stars do when rising and setting, more than when at mid-heaven. Reflection occurs most of all from water, and even more from water just beginning to form than from air; for each of the droplets, out of whose condensing the fine rain is composed, must necessarily be a better mirror than mist is.
13| Since it is both clear, and has already been said, that in mirrors of this sort only color appears and not shape, it is necessary that whenever it begins to rain, and the air in the clouds is already condensing into raindrops but it has not yet rained, if the sun is opposite, or something else so bright that the cloud becomes a mirror, and the reflection occurs toward the bright thing opposite, an appearance of color comes about, not of shape. And since each of the mirrors is small and invisible, but the continuity of magnitude arising from all of them together is visible, a continuous magnitude of the same color necessarily appears; for each of the mirrors renders the same color to what is continuous with it. So since these things can happen whenever the sun and the cloud stand in this relation and we are between them, there will be some appearance because of the reflection. And indeed the rainbow is in fact seen occurring at just such times and not otherwise. That the rainbow, then, is a reflection of sight toward the sun is clear; that is why it always occurs opposite the sun, while the halo occurs around it; and yet both are reflections; but the variety of their colors differs.
14| for the one reflection occurs from water and from something dark, even from far off, the other from nearby and from air that is by nature whiter. The bright thing appears crimson when seen through the dark, or in the dark (it makes no difference which). One can see this in the fire of green wood: its flame appears red because the fire, though bright and white, is mixed with a great deal of smoke. And through mist and smoke the sun appears crimson. This is why in the reflection of the rainbow, the first band appears to have this color (since the reflection comes from small droplets), whereas that of the halo does not. About the other colors we will speak later. Further, around the sun itself no such lingering formation occurs; instead it either rains or the cloud dissolves. But out of opposite conditions, in the interval before the water actually forms, some time elapses; for if this did not happen, the haloes would be colored just like the rainbow. As it is, they do not occur as wholes with such an appearance, nor in a full circle, but small and only in part -- these are called "rods" -- since if such a mist formed as would come from water or some other dark thing, as we were saying, the whole rainbow would appear, just as it does around lamps.
15| For around lamps, since conditions in winter are for the most part moist, a rainbow occurs, and it is most evident to people whose eyes are moist; for their sight, on account of its weakness, is quickly reflected. It comes about both from the moisture of the air and from the soot flowing off from the flame and mixing in; for then it becomes a mirror, and this is also because of the blackness, since soot is smoky. The light of a lamp appears not white but purple in a circle, and rainbow-colored, but not crimson; for the sight that is reflected is small, and the mirror is dark. The rainbow that appears from oars lifted up out of the sea comes about, in its position, in the same way as the one in the sky, but its color is more like the one around lamps; for it appears with a hue that is not crimson but purple. This reflection occurs from droplets that are extremely small but continuous; these are already water fully separated out. It also occurs if someone sprinkles fine droplets onto a place positioned so as to face the sun, where on one side the sun shines and on the other there is shadow;
16| for in such a place, if a person standing outside sprinkles inward, a rainbow appears where the rays overlap and produce the shadow. The manner of it, and the color, are similar, and the cause is the same as with the one from oars; for the person sprinkling uses his hand as an oar. That the color is of this kind, and at the same time the appearance of the other colors, will be clear from the following. For one must, having thought this through as has been said, lay down first that the bright thing in the dark, or seen through the dark, produces a crimson color; second, that sight, when stretched out over a distance, becomes weaker and lesser; and third, that black is like a negation, since a thing appears black through a failure of sight, which is why all distant things appear blacker, because sight does not reach all the way to them. Let these points, then, be examined on the basis of what happens in connection with the senses generally, for the accounts of those matters belong properly to that inquiry; here let us say only as much about them as is necessary. It appears, then, for this reason that distant things are blacker and smaller and smoother, both the things seen in mirrors and clouds -- people see clouds as blacker when looking at them in water than when looking at the clouds themselves.
17| And this is quite evident, for because of the reflection they are viewed with a diminished sight. It makes no difference whether the thing seen changes or the sight does; either way the result will be the same. Besides this, one must not fail to notice the following: it happens that when a cloud is near the sun, a person looking straight at it sees nothing colored, but it is white, while the same person looking at this same cloud in water finds it has some color of the rainbow. It is clear, then, that sight, like black, when broken by weakness, makes things appear blacker, and white less white, and draws it toward black. So the stronger sight changed into a crimson color, the next into green, and the still weaker into violet. Beyond that it no longer appears, but stops within these three, as is also true of most other cases, and here too it reached its limit; the change in the rest is imperceptible. That is why the rainbow appears three-colored, in each band alike, but in reverse order. The first band, then, has crimson on the outside; for from the greatest circumference the most sight strikes the sun, and the outer band is the greatest; the second band, and the third, follow the same proportion.
18| So if what has been said about the appearance of the colors is correct, the rainbow must be three-colored, and colored with only these colors. Yellow appears because of things appearing alongside one another. For the crimson appears white when set beside the green. Here is a sign of this: in the darkest cloud the rainbow becomes purest; and it happens that at that time the crimson seems most yellow. Now yellow is the color in the rainbow that lies between the crimson and the green. Because of the blackness of the cloud all around it, then, the whole of the crimson in it appears white; for it is white relative to those dark surroundings. And again, when the rainbow nearest the sun fades, this happens as the crimson is dissolving; for the cloud, being white, when set beside the green, changes into yellow. The greatest sign of all this is the rainbow from the moon; for it appears entirely white. This happens because it appears both in a cloud that is dark and at night. So just as fire beside fire, black beside black, makes something faintly white appear entirely white, so here; and this white is the crimson. This effect becomes evident also in the case of flowers;
19| For in woven fabrics and embroideries, some colors differ enormously in appearance depending on what is set beside what — for instance purple set among white or black wools, and again in one kind of light or another; this is why embroiderers say they often go wrong when working by lamplight among their patterns, taking one color for another. We have said, then, why the rainbow has three colors, and why it appears to be made up of only these colors. The enclosing rainbow is double and fainter in its colors, and by its position has its colors set in reverse order, for the same cause. For sight, stretching farther, sees the more distant thing as it were less clearly, and here it works in the same way. The reflection from the outer bow is therefore weaker, because the reflection takes place at a greater distance, so that striking the eye less, it makes the colors appear fainter. And they are reversed for the corresponding reason: more light strikes toward the sun from the smaller, inner circumference, since that circumference, being closer to the sight, is reflected from the part of the first rainbow nearest to it. In the outer rainbow the nearest circumference is the smallest, so that this one will have the color red; the next in order, and the third, follow the same proportion.
20| Let the outer rainbow be marked B, the inner one marked A; and of the colors, let the one marked G be red, the one marked D green, the one marked E purple; and yellow appears at the point marked Z. Three rainbows, or more, no longer occur, because the second is already fainter, so that a third reflection becomes altogether weak and is unable to reach the sun.
21| That the rainbow cannot form a full circle, nor a segment greater than a semicircle, and the other facts about it, will be clear to anyone who studies the diagram. For let there be a hemisphere on the circle of the horizon, the circle marked A, with K as its center, and some other point rising, marked H. If the lines falling from K in the shape of a cone form, as it were, an axis along the line HK, and the lines joined from K to M are reflected from the hemisphere to H, then the lines from K will strike the circumference of the circle at the greater angle. And if the reflection occurs when the star is rising or setting, a semicircle will be cut off from the circle by the horizon — namely the part that lies above the earth; but if the star is higher up, the part cut off will always be less than a semicircle. It is smallest when the star is on the meridian. Suppose first that it is at the rising, where H is; let KM be reflected to H, and let the plane in which A lies be produced from the triangle in which HKM lies.
22| The section will then be a circle of the sphere. Let the greatest circle be the one marked A; for it will make no difference through which of the circles on HK the plane through the triangle KMH is produced. Now the lines drawn from H and K will not, by this ratio, be constructed to the semicircle A at one point and then at another; for since the points H and K are given, KH will be given, and so will MH, so that the ratio of MH to MK will also be given. Given this ratio, then, M will touch a determinate circumference. Let this be the circumference on which N and M lie; so the intersection of the circumferences is given. But on any other circumference besides the one through M and N, the same ratio from the same points cannot hold in the same plane. Let some line DB, then, be set out, and let it be cut so that D is to B as MH is to MK. Now MH is greater than MK, since the reflection of the cone occurs at the greater angle; for it subtends the greater angle of the triangle KMH.
23| D, therefore, is also greater than B. Let a length, marked Z, be added to B, so that as D is to B, so BZ is to D. Then let it be arranged that as Z is to KH, so B is to another line, KP; and let MP be joined from P to M. P will then be the pole of the circle on which the lines from K fall; for as Z is to KH, so will B be to KP, and D to PM. For suppose it is not so, but the ratio holds instead to something less than PM or greater than it — it will make no difference. Let it be to PR. HK, KP, and PR will then stand in the same ratio to one another as D, B, Z do. But D, B, Z were proportional, such that as D is to B, so ZB is to D; so that as PH is to PR, so PR is to PK.
24| If, then, HR and KR are joined from H and K to the point R, these joined lines will have the same ratio as PH has to PR; for about the same angle P, the sides of triangle HPR and of triangle KRP are proportional. So PR will also have the same ratio to KR that HP has to PR. And MH has to MK the same ratio that D has to B — both ratios being equal. So the lines from the points H and K, having the same ratio, would be constructed not only at the circumference through M and N, but elsewhere as well; which is impossible. Since, then, D stands in this ratio neither to something less than PM nor to something greater than it (for we shall show this in the same way), it is clear that it must stand in this ratio to PM itself. So it will be that as MP is to PK, so PH is to MP, and, by the remaining term, MH is to MK.
25| If, then, using the point marked P as pole, and the distance MP as radius, a circle is drawn, it will touch all the angles that the lines from H and K make when reflected to the circle through M; but if not, it will be shown in the same way that lines having the same ratio are constructed at one point of the semicircle and then at another, which was impossible. If, then, you rotate the semicircle A about the diameter HKP, the lines from H and K, reflected to the point marked M, will behave in the same way in every plane, and will make the angle KMH equal in each. And the angle that HP and MP make at HP will always be equal. Triangles equal to triangle HMP and triangle KMP are therefore constructed on HP and KP. The perpendiculars of these will fall on the same point of HP, and will be equal. Let them fall on the point O. O, then, is the center of the circle, and the semicircle around MN has been cut off from the horizon; for the sun does not overpower the drops that are above, but does overpower those settled near the earth, and dissolves the moisture in the air.
26| And because of this the circle of the rainbow does not close completely. It occurs at night too, from the moon, though rarely; for the moon is not always full, and it is by nature weaker, so that it cannot master the air. The rainbow stands out most where the sun is most mastered, for there the most moisture remains in it. Again, let the horizon be the line on which A B Γ lies, and let H have risen, and let the axis now be the line on which H Π lies. Now everything else will be shown in the same way as before, but the pole of the circle, the point on which Π lies, will be below the horizon on which AΓ lies, once the point H has been raised. And on the same line, whenever the pole and the center of the circle coincide with the point of the horizon that now marks the rising; for this is the line on which HΠ lies. And since K H lies above the diameter AΓ, the center will be below the horizon AΓ, as before, on the line KΠ, at the point O. So the segment above will be smaller than a semicircle, the one on which Ψ Υ Ω lies;
27| for ΨΥΩ was a semicircle, but now it has been cut off from the horizon AΓ. ΥΩ will be invisible once the sun itself has risen, and least invisible at midday; for the higher H is, the lower both the pole and the center of the circle will be. As for why, during the shorter days after the autumn equinox, a rainbow can always occur, but during the longer days that run from one equinox to the other, a rainbow does not occur around midday, the cause is this: the segments toward the north are all greater than a semicircle, and always tend toward a still greater semicircle, while the invisible part is small; but of the segments on the south side of the equinoctial circle, the part above is small and the part under the earth is great. And the farther out segments are always the greater. So on the days near the summer solstice, because of the size of the segment, before H reaches the middle of the segment and the meridian, Π has already gone completely below, because midday stands far off from the earth on account of the size of the segment. But on the days near the winter solstice, because the segments of the circles do not rise much above the earth, the opposite necessarily happens;
28| for once H has risen only a little, the sun is already at midday. We must suppose the same causes hold for the sun-dog and for the streaks as for the phenomena already discussed. A sun-dog occurs when the sight is reflected toward the sun, while streaks occur because the sight, being of the kind we said always arises when clouds are near the sun, is reflected from some moist surface toward the cloud. For the clouds themselves appear colorless when looked at directly, but in the water the cloud appears full of streaks — except that in that case the color seems to belong to the cloud as seen in the water, whereas in the case of the streaks it seems to belong to the cloud itself. This happens whenever the composition of the cloud is uneven, dense in one part and thin in another, more watery in one part and less in another. For when the sight is reflected toward the sun, the shape of the sun is not seen, because the mirrors are too small, but the color is, because the sun — bright and white — appears in an uneven medium; and depending on which part the sight was reflected toward, one part appears crimson, another green or yellow. For it makes no difference whether one sees through such media or sees a reflection from such media; in both cases the color appears the same, so that if it is crimson seen the one way, it is crimson seen the other way too.
29| So the streaks arise from unevenness in the mirror, not in shape but in color, while the sun-dog occurs when the air is as even as possible and likewise dense throughout — which is why it appears white. For the evenness of the mirror produces a single color in the reflected image, while the reflection of the sight, coming all at once, since it strikes the sun all together from mist that is dense — not yet water, but close to water — makes the color that belongs to the sun appear, just as when sight is bent back from smooth bronze on account of its density. So, since the color of the sun is white, the sun-dog too appears white. And for this same reason the sun-dog is a stronger sign of coming water than the streaks are, for it happens that the air is more favorably disposed toward the generation of water in that case. And a southern sun-dog is a stronger sign than a northern one, because southern air changes into water more than air toward the north does. They occur, as we said, around sunset and around sunrise, and neither from above nor from below, but from the sides — both the streaks and the sun-dogs; and neither too near the sun nor altogether too far. For if it is near, the sun dissolves the cloud's composition, and if it is far, the sight will not be reflected;
30| for sight, stretched far out from a small mirror, becomes weak. That is also why halos do not occur directly opposite the sun. So if the cloud is high up and near, the sun will dissolve it; but if it is far off, the sight, being too weak to produce a reflection, will not strike it. But to the side, below the sun, the mirror can stand at just such a distance that the sun does not dissolve it, yet the sight arrives all at once, because, being carried toward the earth, it does not travel as if through empty space. But directly under the sun it does not occur, because near the earth the cloud would be dissolved by the sun, while up above, with the sun at mid-heaven, the sight is scattered apart. And in general it does not occur to the side either when the sun is at mid-heaven, for the sight is not carried under the earth, so that only a little of it reaches the mirror, and the reflected sight becomes altogether weak. The effects, then, that the exhalation happens to produce in the regions above the earth are, roughly speaking, this many and of this kind. We must now speak of the effects it produces within the earth itself, shut up inside the earth's own parts. For it produces two differentiae of bodies, because it too is by nature twofold, just as it is in the region above;
31| for there are two exhalations, as we say, the vaporous and the smoky, and there are also two kinds of things that arise in the earth: the mined minerals and the worked metals. The dry exhalation, then, is what, by a kind of firing, produces all the minerals — such as the kinds of stone that do not melt, and realgar, and ochre, and red ochre, and sulfur, and other things of that sort. Most minerals are either colored powder, or stone that has come to be from such a composition, as for example cinnabar. From the vaporous exhalation come the things that are mined and are either fusible or malleable, such as iron, copper, gold. The vaporous exhalation, shut up inside — especially within stones — produces all these by being compressed together into one through dryness and congealed, just as dew or frost is congealed when it separates out. But in this case these things are generated before the exhalation separates out. That is why in one sense these are water, and in another sense they are not: their matter was potentially water, but it is no longer so, nor did it come to be from water that had already come to be, through some affection, as the flavors do. For copper does not come to be that way, nor gold; rather, before coming to be, once the exhalation has congealed, each of these already is what it is.