We have so far dwelt on the effects which the particles expelled from the sun and the stars exert on distant celestial bodies. It may be asked whether this dust does not act upon our own earth. We have already recognized the peculiar luminescence which on clear nights is diffused over the sky as a consequence of electrical discharges of this straying dust. This leads to the question whether the magnificent polar lights, which according to modern views are also caused by electric discharges in the higher strata of the atmosphere, are not produced by dust which the sun sends to us. It will, indeed, be seen that we can in this way explain quite a number of the peculiarities of these mysterious phenomena which have always excited man’s imagination.

We know that meteorites and shooting-stars are rendered incandescent by the resistance which they encounter in the air at an average height of 120 km. (75 miles), sometimes of 150 and 200 km. In isolated cases meteorites are supposed to have become visible even at still greater altitudes. It would result that there must be appreciable quantities of air still at relatively high elevation, and that the atmosphere cannot be imperceptible at an altitude of less than 100 km., as was formerly assumed. Bodies smaller than the meteorites as well as the solar dust we have spoken of—which, owing to their minuteness and to the strong cooling by heat radiation and conduction that they undergo in passing through the atmosphere, could never attain incandescence—would be stopped at greater heights. We will assume that they are arrested at a mean height of about 400 km. (250 miles).

The masses of dust which are expelled by the sun are partly uncharged, partly charged with positive or negative electricity. Only the latter can be connected with the polar lights; the former would remain dark and slowly sink through our atmosphere to the surface of the earth. They form the so-called cosmical dust, of whose great importance Nordenskiöld was so firmly convinced. He estimated that the yearly increase in the weight of the earth by the addition of the meteorites was at least ten million tons, or five hundred times more than we stated above (page 108). Like Lockyer and, in more recent days, Chamberlin, he believed that the planets were largely built up of meteorites.

The dust reaching the earth from the sun would not, were it not electrically charged, amount to more than 200 tons in a year. Although this figure may be far too low, yet the supply of matter by these means is certainly very small in comparison with the 20,000 tons which the earth receives in the shape of meteorites and shooting-stars. But owing to its extremely minute distribution, the effect of this dust is very important, and it may constitute a much greater portion of the finely distributed cosmical dust in the highest strata of the atmosphere than the dust introduced by falling meteorites and shooting-stars.

That these particles exert a noticeable influence upon terrestrial conditions, in spite of their relatively insignificant mass, is due to two causes. They are extremely minute and therefore remain suspended in our atmosphere for long periods (for more than a year in the case of the Krakatoa dust), and they are electrically charged.

In order to understand their action upon the earth, we will examine how the terrestrial conditions depend upon the position of the earth with regard to the various active portions of the sun, and upon the change of the sun itself in regard to its emission of dust particles. For this examination we have to avail ourselves of extensive statistical data; for only a long series of observations can give us a clear conception of the action of solar dust.

These particles withdraw from the sun gases which they were able to condense on their surface, and which had originally been in the chromosphere and in the corona of the sun. The most important among these gases is hydrogen; next to it come helium and the other noble gases which Ramsay has discovered in the atmosphere, in which they occur in very small quantities. As regards hydrogen, Liveing and (after him) Mitchell have maintained that it is not produced in the terrestrial atmosphere. Occasionally it is certainly found in volcanic gases. Thus hydrogen escapes, for instance, from the crater of Kilauea, on Hawaii, but it is burned at once in the atmosphere. If hydrogen were present in the atmosphere, it would gradually combine with the oxygen to water vapor; and we have to assume, therefore, that the hydrogen must be introduced into our atmosphere from another source—namely, from the sun. Mitchell finds in this view a strong support for the opinion that solar dust is always trickling down through our atmosphere.

The quantity of solar dust which reaches our atmosphere will naturally vary in proportion with the eruptive activity of the sun. The quantity of dust in the higher strata influences the color of the light of the sun. After the eruption of the volcano Rakata on Krakatoa, in 1883, and again, though to a lesser degree, after the eruption of Mont Pelée on Martinique, red sunsets and sunrises were observed all over the globe. At the same time, another phenomenon was noticed which could be estimated quantitatively. The light of the sky is polarized with the exception of the light coming from a few particular spots. Of these spots, one called Arago’s Point is situated a little above the antipode of the sun, and another, Babinet’s Point, is situated above the sun. If we determine the elevation of these points above the horizon at sunset, we find in accordance with the theoretical deduction that this elevation is greater when the higher strata of the atmosphere are charged with dust (as after the eruption of Rakata) than under normal conditions. Busch, a German scientist, analyzed the mean elevation of these points (stated in degrees of arc) at sunset, and found the following peculiar numbers:

There is a distinct parallelism in these series of figures. Almost simultaneously with the sun-spot maximum the height of the two so-called neutral points above the horizon attains its maximum at sunset, and the same applies to the minimum. That the phenomena in the atmosphere take place a little later than the phenomena on the sun which caused them is perhaps only natural.

When the air is rich in dust, or when it is strongly ionized by kathode rays, conditions are favorable for the formation of clouds. This can be observed, for instance, with auroral lights. They regularly give rise to a characteristic cloud formation, so much so that Adam Paulsen was able to recognize polar lights by the aid of these clouds in full daylight. Klein has compiled a table on the connection between the frequency of the higher clouds, the so-called cirrus clouds, at Cologne, and the number of sun-spots during the period 1850-1900. He demonstrates that during this half-century, which comprises more than four sun-spot periods, the sun-spot maxima fell in the years in which the greatest number of cirrus clouds had been observed. The minima of the two phenomena are likewise in agreement.

A similarly intensified formation of clouds seems also to occur on Jupiter when sun-spots are frequent. Vogel states that Jupiter at such times shines with a whiter light, while at sun-spot minima it appears of a deeper red. The deeper we are able to peep into the atmosphere of Jupiter, the more reddish it appears. During periods of strong solar activity the higher portions of Jupiter’s atmosphere therefore appear to be crowded with clouds.

The discharge of the charged solar dust in our atmosphere calls forth the polar lights.

The polar lights occur, as the name indicates, most frequently in the districts about the poles of the earth. They are, however, not actually more frequent the nearer we come to the poles; but they attain a maximum of frequency in circles which enclose the magnetic and the geographical poles. The northern maximum belt passes, via Cape Tscheljuskin, north of Novaja Semlja, along the northwestern coast of Norway, a few degrees to the south of Iceland and Greenland, right across Hudson Bay and over the northwestern extension of Alaska. When we go to the south of this belt, the auroras, or boreal lights, diminish markedly. They are four times less frequent in Edinburgh, and fifteen times less frequent in London or New York, than in the Orkney Islands or Labrador.

Paulsen divides the auroras into two classes, which behave quite differently in several respects. The great difficulties which the solution of the problems of polar lights has so far offered seem to a large extent to be due to the fact that all polar lights were treated as being of the same kind.

The polar lights of the first class do not display any streamers. They cover a large portion of the sky in a horizontal direction. They are very quiet, and their light is strikingly constant. As a rule, they drift slowly towards the zenith, and they do not give rise to any magnetic disturbances.

These polar lights generally have the shape of an arch whose apex is situated in the direction of the magnetic meridian (Fig. 38). Sometimes several arches are grouped one above another.

Nordenskiöld observed these arches quite regularly during the polar night when he was wintering near Pitlekaj, in the neighborhood of Bering Sound. Adam Paulsen has often seen them on Iceland and Greenland, which are situated within the maximum belt spoken of, where northern lights are very common. Occasionally auroras are also seen farther from the poles, as circular arches of a milky white, which may be quite high in the heavens.

Sometimes we perceive in the arctic regions that large areas of the heavens are covered by a diffused light which might best be compared to a luminous, transparent cloud; the darker portions in it probably appear dark by contrast. This phenomenon was frequently observed during the Swedish expedition of 1882-1883, near Cape Thordsen.

Masses of light at so low a level that the rocks behind them are obscured have frequently been observed to float in the air, especially in the arctic districts. Thus Lemström saw an aurora on the island of Spitzbergen in front of a wall of rock only 300 m. (1000 ft.) in height. In northern Finland he observed the auroral line in the light of the air in front of a black cloth only a few metres distant. Adam Paulsen counts these phenomena also as polar lights of the first class, and he regards them as phosphorescent clouds which have been carried down by convection currents to an unusually low level of our atmosphere.

Polar lights of the second class are distinguished by the characteristic auroral rays or streamers. Sometimes these streamers are quite separated from one another (see Fig. 39); as a rule they melt into one another, especially below, so as to form draperies which are so easily moved and unsteady that they appear to flutter in the wind (Fig. 41.) The streamers run very approximately in the direction of the inclination (magnetic dip) needle, and when they are fully developed around the celestial dome their point of convergence is distinctly discernible in the so-called corona (Fig. 40). When the light is at its greatest intensity the aurora is traversed by numerous waves of light.

The draperies are very thin. Paulsen watched them sometimes drifting over his head in Greenland. The draperies then appeared foreshortened, in the shape of striæ or ribbons of light in convolutions. These polar lights influence the magnetic needle. When they pass the zenith their influence changes sign, so that the deviation of the magnetic needle changes from east to west when the ribbon is moving from north to south. Paulsen therefore concluded that negative electricity (kathode rays) was moving downward in these rays. These polar lights correspond to violent displacements of negative electricity, while polar lights of the first class appear to consist of a phosphorescent matter which is not in strong agitation. The streamers may penetrate down into rather low atmospheric strata, at least in districts which are near the maximum belt of the northern lights. Thus Parry observed at Port Bowen an auroral streamer in front of a cliff only 214 m. (700 ft.) in height.

Polar lights of the first order may pass into those of the second order, and vice versa. We frequently see rays suddenly flash out from the arch of the aurora, mostly downward, but, when the display is very intense, also upward. On the other hand, the violent agitation of a "drapery light" may cease, and may give way to a diffused, steady glow in the sky. The polar light of the first class is chiefly observed in the arctic regions. To it corresponds, in districts farther removed from the pole, the diffused light which appears to be spread uniformly over the heavens and which gives the auroral line.

The usually observed polar lights (speaking not only of those seen on arctic expeditions) belong to the second class, which comprises also all those included in the subjoined statistics, with the exception of the auroral displays reported from Iceland and Greenland. While the streamer lights distinctly conform to the 11.1 years’ period, and become more frequent at times of sun-spot maxima, this is not the case, according to Tromholt, with the auroras of Iceland and Greenland. Their frequency, on the contrary, seems to be rather independent of the sun-spot frequency. Not rarely auroral maxima corresponding to sun-spot maxima are subdivided into two by a secondary minimum. This phenomenon is most evident in the polar regions, but it can also be traced in the statistics from Scandinavia and from other countries.

Better to understand the nature of auroras, we will consider the sun’s corona during the time of a minimum year, taking as an example the year 1900 (compare Fig. 30). The rays of the corona in the neighborhood of the poles of the sun are laterally deflected by the action of the magnetic lines of force of the sun. The small, negatively charged particles have evidently only a low velocity, so that they move quite close to the lines of force in the neighborhood of the solar poles and are concentrated near the equator. There the lines of force are less crowded—that is to say, the magnetic forces are weaker—and the solar dust can therefore be ejected by the radiation pressure and will accumulate to a large disk expanding in the equatorial plane. To us this disk appears like two large streams of rays which project in the direction of the solar equator. Part of this solar dust will come near the earth and be deflected by the magnetic lines of force of the earth; it will hence be divided into two streams which are directed towards the two terrestrial magnetic poles. These poles are situated below the earth’s crust, and therefore not all the rays will be concentrated towards the apparent position of the magnetic poles upon the surface of the earth. It is to be expected that the negatively charged particles coming from the sun will chiefly drift towards that district which is situated somewhat to the south of the magnetic north pole, when it is noon at this pole. When it is midnight at the magnetic pole, most of the negatively charged particles will be caught by the lines of force before they pass the geographical north pole, and the maximum belt of the auroras will for this reason surround the magnetic and the geographical poles, as has already been pointed out (compare page 122). The negatively charged solar dust will thus be concentrated in two rings above the maximum belts of the polar lights. Where the dust collides with molecules of the air, it will produce a phosphorescent glow, as if these molecules were hit by the electrically charged particles of radium. This phosphorescent glow rises in the shape of a luminous arch to a height of about 400 km. (250 miles)—according to Paulsen—and the apex of this arch will in every part seem to lie in the direction where the maximum belt is nearest to the station of the observer. That will fairly coincide with the direction of the magnetic needle.

The solar corona of a sun-spot maximum year is of a very different appearance (Fig. 31). The streamers radiate straight from the sun in almost all directions; and if there be some privileged directions, it will be those above the sun-spot belts. The velocity of the solar dust is evidently so great that the streamers are no longer visibly deflected by the magnetic lines of force of the sun. Nor is this charged dust influenced to any noticeable degree by the magnetic lines of force of the earth. It will in the main fall straight down in that part of the atmosphere in which the radiation is most intense: As these "hard" rays of the sun[8] seem to issue from the faculæ of the sun which are most frequent in maximum sun-spot years, some polar lights will also be seen in districts which are far removed from the maximum belt of the auroras, especially when the number of sun-spots is large. The opposite relation holds for the "soft" streams of solar dust which fall near the maximum belt of the polar lights. These streams occur most frequently with low sun-spot frequency, as we know from observations of the solar corona. Possibly they are carried along by the stream of harder rays in maximum years. The polar lights corresponding to these rays therefore attain their maximum with few sun-spots. Hard and soft dust streams occur, of course, simultaneously; but the former predominate in maximum sun-spot years, the latter in minimum years.

That the periodicity of the polar lights in regions without the maximum belt follows very closely the periodicity of the sun-spots was shown by Fritz as early as 1863. The length of the period varies between 7 and 16 years, the average being 11.1 years. The years of maxima and minima for sun-spots and for northern auroras are the following:

MAXIMUM YEARS

MINIMUM YEARS

There are, in addition, as De Mairan proved in his classical memoir of the year 1746, longer periods common to both the number of sun-spots and the number of auroras. According to Hansky, the length of this period is 72 years; according to Schuster, 33 years. Very pronounced maxima occurred at the beginning and the end of the eighteenth century, the last in the year 1788; afterwards auroras became very rare in the years 1800-1830, just as in the middle of the eighteenth century. In 1850, and particularly in 1871, there were strong maxima; they have been absent since then.

The estimates of the heights of the polar lights vary very considerably. The height seems to be the greater, on the whole, the nearer the point of observation is to the equator, which would well agree with the slight deflection of the kathode rays towards the surface of the earth in regions which are farther removed from the pole. Gyllenskiöld found on Spitzbergen a mean height of 55 km.; Bravais, in northern Norway, 100 to 200 km.; De Mairan, in central Europe, 900 km.; Galle, again, 300 km. In Greenland, Paulsen observed northern lights at very low levels. In Iceland he fixed the apex of the northern arch which may be considered as a point where the charged particles from the sun are discharged into the air at about 400 km. Not much reliance can be placed upon the earlier determinations; but the heights given conform approximately to the order of magnitude which we may deduce from the height at which the solar dust will be stopped by the terrestrial atmosphere.

The polar lights possess, further, a pronounced yearly periodicity which is easily explicable by the aid of the solar dust theory. We have seen that sun-spots are rarely observed near the solar equator, and the same applies to solar faculæ. They rapidly increase in frequency with higher latitudes of the sun, and their maximum occurs at latitudes of about fifteen degrees. The equatorial plane of the sun is inclined by about seven degrees towards the plane of the earth’s orbit. The earth is in the equatorial plane of the sun on December 6th and June 4th, and most distant from it three months later. We may, therefore, expect that the smallest number of solar-dust particles will fall on the earth when the earth is in the equator of the sun—that is, in December and June—and the greatest number in March and September. These relations are somewhat disturbed by the twilight, which interferes with the observation of auroras in the bright summer nights of the arctic region, while the dark nights of the winter favor the observation of these phenomena. The distribution of the polar lights over the different seasons of the year will become clear from the subjoined table compiled by Ekholm and myself:

In zones where the difference between the lengths of day and night of the different seasons is not very great, as in the United States, and in districts in which the southern light is observed (about latitude 40° S.), the chief minimum falls in winter: on the northern hemisphere, in December; on the southern hemisphere, in June or July. A less pronounced minimum occurring in the summer. Twice in the course of the year the earth passes through the plane of the solar equator. During these periods a minimum of solar dust trickles down upon the earth, and that period is characterized by a larger number of polar lights which is distinguished by a higher elevation of the sun above the horizon. We may expect this; for most solar dust will fall upon that portion of the earth over which the sun is highest at noon. The two maxima of March or April and of September or October, when the earth is at its greatest distance from the plane of the solar equator, are strongly marked in all the series, except in those for the polar districts Iceland and Greenland. There the auroral frequency is solely dependent upon the intensity of the twilight, so that we find a single maximum in December and the corresponding minimum in June. More recent statistics (1891-1903) indicate, however, a minimum in December. For the same reason the summer minimum in countries of high latitudes, like Sweden and Norway, is very much accentuated.

Similar reasons render it difficult for most localities to indicate the daily periodicity of the polar lights. Most of the solar dust falls about noon; and most polar lights should therefore be counted a few hours after noon, just as the highest temperature of the day is reached a little after noon. On account of the intense sunlight, however, this maximum can only be established in the wintry night of the polar regions, and even there only when a correction has been made for the disturbing effect of the twilight. In this way Gyllenskiöld found a northern-light maximum at 2.40 P.M. for Cape Thordsen, on Spitzbergen, the corresponding minimum being at 7.40 A.M. In other localities we can only ascertain that the polar lights are more intense and more frequent before than after midnight. In central Europe the maximum occurs at about 9 P.M.; in Sweden and Norway (in latitude 60° N.), half an hour or an hour later.

A few other periods, approximately of the length of a month, have been suggested with regard to polar lights. A period lasting 25.93 days predominates in the southern lights, where the maximum exceeds the average by 44 per cent. For the northern lights in Norway the corresponding excess percentage is 23; for Sweden, only 11.[9]

The same period of nearly twenty-six days had already been pointed out for a long series of other especially magnetic phenomena which, as we shall see, are very closely connected with auroras, and it had also been found in the frequency of thunder-storms and in the variations of the barometer. This periodicity has often been thought to be connected with the axial rotation of the sun. The Austrian scientist Hornstein has even gone so far as to propose that the length of this period should be carefully determined, "because it would give a more accurate value for the rotation of the sun than the direct determinations." We know now that the length of the solar revolution is different for different solar altitudes, a circumstance with which observations of sun-spot movements at different latitudes had already made Carrington and Spörer familiar, but which was not safely established before Dunér’s spectroscopical determination of the movement of the solar photosphere. Dunér found the following sidereal revolutions for different latitudes of the sun to which the subjoined synodical revolution would correspond. (By sidereal revolution of a point on the sun we understand the time which elapses between the two moments when a certain star passes, on two consecutive occasions, through the meridian plane of the point—that is to say, through a plane laid through the poles of the sun and the point in question. The synodical revolution is determined by the passage of the earth through this meridian. On account of the proper motion of the earth the synodical period is longer than the sidereal period.)

That the periods of rotation of the solar photosphere, and, in a similar way, the periods of the spots, the faculæ, and the prominences, should become so considerably longer with increasing latitudes is one of the most mysterious problems of the physics of the sun. Something similar applies to the clouds of Jupiter, but the difference in that case is much smaller—only about one per cent. The clouds of our atmosphere behave quite differently, a fact which is explained by our atmospheric circulation.[10]

In our case, of course, the position of the sun with regard to the earth—that is to say, the synodical period—can alone be of importance. We recognize that the period of 25.93 days does not at all agree with any period of the solar photosphere. The solar equatorial zone differs least, and it would be appropriate to reckon with this period, since the earth never moves very far from the plane of the solar equator, and returns to that plane, at any rate, twice in the course of a year.

But there is another peculiarity. The higher a point is situated in the atmosphere of the sun, the shorter is its period. Thus the synodical period of the faculæ near the equator is on an average 26.06, the period of the spots 26.82, of the photosphere 27.3 days. Faculæ situated at higher levels revolve still more rapidly, and we are thus driven to the conclusion that the period to which we have alluded agrees with the period of the faculæ situated at higher levels in the equatorial zone of the sun, and is probably dependent upon them. That would conform to our ideas concerning the physics of the sun. For the faculæ are produced in the ascending currents of gas and at rather lower levels than the spherules which are expelled by the radiation pressure. This radiation pressure is strongest just in the neighborhood of the faculæ.

For the same reason the repulsion of the solar dust becomes particularly powerful when the faculæ are strongly developed—that is to say, just in the time of pronounced eruptive activity of the sun which is characterized by many sun-spots.

We must thus imagine that the radiation of the sun will be stronger in times of strong eruptive activity than during the days of low sun-spot frequencies. Direct observations of the intensity of the solar radiation which have been made by Saveljeff in Kieff confirm this assumption. It must be pointed out, however, that another phenomenon investigated by Köppen seems to contradict this conclusion. Köppen ascertained that in our tropics the temperature was by 0.32° Cent. (nearly 0.6° F.) lower during sun-spot maxima than the average, and that five years later, a year before the sun-spot minimum, it reached its maximum value of 0.41° Cent. (0.7° F.) above the average. A similar peculiarity can be traced to other zones, but on account of the less steady climates it is much less marked there than in the tropics. A French physicist, Nordmann, has fully confirmed the observations of Köppen. On the other hand, Very, an American astronomer, has found that the temperature in very dry (desert) districts of the tropics (near Port Darwin, 12° 28´ S., and near Alice Springs, 23° 38´ S., both in Australia) is higher at sun-spot maxima than at minima; but Very was in this research guided merely by the records of maximum and minimum thermometers. From Very’s investigation it would appear that the solar radiation is really more intense with larger sun-spot numbers.[11] This, it must be remarked, is only noticeable in exceedingly dry districts in which there is no cloud formation worth mentioning. In other districts the cloud formation which accompanies sun-spot maxima interferes with the simplicity of the phenomena. The cooling effect of the clouds seems in these cases by far to surpass the direct heating effect of the solar rays, and in this manner Köppen’s conclusion would become explicable. If we could observe the temperatures of the atmospheric strata above the clouds, their variation would no doubt be in the same degree as that in the desert.

Finally, we have to note another period in the phenomena of the polar lights—the so-called tropical month, whose length is 27.3 days. The nature of this period is little understood. It is possibly connected with the electric charge of the moon. The peculiarity of this period is that it acts in an opposite way in the northern and southern hemispheres. When the moon is above the horizon, it seems to prevent the formation of polar lights; but for this case the difficulties of observation caused by the moonlight must, of course, be taken into consideration.

Celsius and Hiorter observed in 1741 that the polar lights exercise an influence on the magnetic needle. From this circumstance we have drawn the conclusion that the polar lights are in some way due to electric discharges which act upon the magnetic needle. These magnetic effects, the disturbances of the otherwise steady position of the magnetic needle, are not influenced by the light of the sun and moon, and can therefore be studied to greater advantage than auroras. We have already pointed out that it is only the aurora of the radial, streamer type which exerts this magnetic influence (compare Figs. 42 and 43).

These magnetic variations show exactly the same periods as the northern lights and the sun-spots. As regards, first, the long period of 11.1 years, our observations prove that the so-called magnetic disturbances of the position of the magnetic needle faithfully reflect the variations in the sun-spots. This connection was discovered in 1852 by Sabine in England, by Wolf in Switzerland, and by Gautier in France. Even the more irregular diurnal variations in the magnetic elements are subjected to a solar period. The magnetic needle points in our districts with its north end towards the north—not exactly, though, being deflected towards the west. This western deviation or declination is greatest soon after noon, about one o’clock, and this diurnal change is greater in summer than in winter, and the fluctuation of the position of the magnetic needle greater in daytime than at night-time. It is, therefore, manifest that we have to deal with some solar effect. This becomes still more distinct when we study the change with reference to the daily variation in the number of sun-spots. In the subjoined table the variation in the declination has been compiled for Prague for the years 1856 to 1889. Only years with maxima and minima of sun-spots and of magnetic variations have been noticed in this table:

DAILY VARIATIONS IN DECLINATION

We see that the maxima and minima years of the two phenomena very nearly coincide. The accord is so evident that we may calculate the diurnal variation as proportional to the increase in the number of sun-spots. This is shown by the two last lines of the table.

The yearly variation is again exactly the same as that of polar lights, as the following table indicates, in which the disturbances of magnetic declination, horizontal intensity, and vertical intensity are compiled for Toronto, Canada; for comparison the means of these three magnitudes are added for Greenwich. The unit of this table is the average annual variation:

TORONTO

GREENWICH

The daily variation of the disturbances has been analyzed by Van Bemmelen for the period 1882-1893 and for the observatory of Batavia, on Java. The maximum occurs there about 1 P.M., and is about 1.86 times as great as the average value for the day. The minimum of 0.48 occurs at 11 P.M. Between 8 P.M. and 3 A.M. the disturbances are almost as rare as about 11 o’clock at night.

The variation is greatest with that declination which has its maximum of 3.26 at 12 M., and its minimum of 0.14 at 11 P.M.

The period of almost 26 days first investigated by Hornstein has also been refound in the magnetic variations and disturbances by Broun, Liznár, and C. A. Müller. It must be added, however, that Schuster does not consider these data as in any way conclusive.

The moon has also a slight influence upon the magnetic needle, as Kreil proved as early as 1841. The effect is in a different sign in the northern and southern hemispheres, and may be likened to a tidal phenomenon.

The ultra-violet rays of the sun are strongly absorbed by the atmosphere, and they cause an ionization of the molecules of the air. This ionization is, on the whole, more marked at higher altitudes. The ascending air currents carry with them water vapor which is condensed on the ions, particularly on the negative ions. In this way most clouds become negatively charged; this interesting fact—i.e., that they are more frequently charged with negative than with positive electricity—was first proved by Franklin in his kite experiments. When the rain-drops have fallen, the air above remains positively charged; this has been observed during balloon ascensions. The clouds which are formed at high levels are most strongly charged; for this reason thunder-storms over land occur mostly in the summer-time. The thunder-storms also show the 26-day period, as Bezold has proved for southern Germany, and Ekholm and myself have shown for Sweden.

A vast amount of material concerning these questions and magnetic phenomena in particular has been collected by the various meteorological observatories and is awaiting analysis.

Although some observers like Sidgreaves question the correlation of sun-spots and polar lights or magnetic disturbances, because strong spots have been seen on the disk of the sun without any magnetic disturbances having been noticed, yet the view predominates that the magnetic disturbances are caused by sun-spots when the sun-spots cross the central meridian of the sun which is opposite the earth. Thus Maunder observed a magnetic storm and a northern light succeeding the passage of a large sun-spot through the central solar meridian on the 8th to the 10th of September, 1898. The magnetic effect attained its maximum about twenty-one hours after the passage through the meridian.

Similarly Riccò established in ten instances, in which exact determination was possible, a time interval of 45.5 hours on an average between the meridian passage of a spot and the maximum magnetic effect. Riccò also submitted to an analysis the data which Ellis had collected and which Maunder had investigated. He found for these instances, on an average, almost exactly the same numbers, the time interval being 42.5 hours. That would correspond to a mean velocity of the solar dust of from 910 to 980 km. per second. On the other hand, we have no difficulty at all in calculating the time which a spherule of a diameter of 0.00016 mm. (those particles travel fastest) and of the specific gravity of water would need in order to reach the earth, under the influence of solar gravitation and of a mechanical radiation pressure 2.5 times as large from the outside of the sun. The time found, 56.1 hours, corresponds to a mean velocity of 740 km. per second. In order that the solar dust may move with the velocity calculated by Riccò, its specific gravity should be less than 1—viz., 0.66 and 0.57. This value looks by no means improbable, when we assume that the spherules consist of hydrocarbons saturated with hydrogen, helium, and other noble gases. We should also arrive at larger velocities for the solar dust, as has already been pointed out with regard to the tails of comets, when we presume that the particles consist of felted marguerites of carbon or silicates, or of iron—materials which we regard as the main constituents of meteorites.

It should, perhaps, be mentioned that the most intense spectrum line of the polar lights has been found to be characteristic of the noble gas krypton. As this gas is found only in very small quantities in our atmosphere, it does not appear improbable that it has been carried to us together with the solar dust, and that its spectrum becomes perceptible during the discharge phenomena. The other spectrum lines of the polar lights belong to the spectra of nitrogen, argon, and of the other noble gases. The volumes of noble gases which are brought into our terrestrial atmosphere in this manner would in any case be exceedingly small.

The electrical phenomena of our terrestrial atmosphere indirectly possess considerable importance for organic life and, consequently, for human beings. By the electrical discharges part of the nitrogen is made to combine with the oxygen and hydrogen (liberated by the electric decomposition of water vapor) of our air, and it thus forms the ammonia compounds, as well as the nitrates and nitrites, which are so essential to vegetable growth. The ammonia compounds which play a most important part in the temperate zones appear especially to be formed by the so-called silent discharges which we connect with auroras. The oxygen compounds of nitrogen, on the other hand, seem to be chiefly the products of the violent thunder-storms of the tropics. The rains carry these compounds down into the soil, where they fertilize the plants.

The supply of nitrogen thus fixed amounts in the course of a year to about 1.25 gramme per square metre in Europe, and to almost fourfold that figure in the tropics. If we accept three grammes as the average number for the whole firm land of the earth, that would mean 3 tons per square kilometre, and about 400 million tons per year for the whole firm land of 136 million square kilometres. A very small portion of this fixed nitrogen, possibly one-twentieth, falls on cultivated soil; a larger portion will help to stimulate plant growth in the forests and on the steppes. We may mention, for comparison, that the nitrogen contained in the saltpetre which the mines of Chili yield to us has risen from 50,000 tons in 1880 to 120,000 tons in 1890, to 210,000 tons in 1900, and to 260,000 tons in 1905. The nitrogen produced in the shape of ammonium salts (sulphate) by the gas-works of Europe amounts to about one-quarter of the last-mentioned figure. To this figure we have, of course, to add the production of coal-gas ammonia in the United States and elsewhere. Yet even allowing for this item, we shall find that the artificial supply of combined nitrogen on the earth does not represent more than about one-thousandth of the natural supply.

As the nitrogen contents of the air may be estimated at 3980 billion tons, we recognize that only one part in three millions of the nitrogen of the atmosphere is every year fixed by electric discharges, presuming that the nitrogen supply to the sea is as great per square kilometre as to the land. The nitrogen thus fixed benefits the plants of the sea and of the land, and passes back into the atmosphere or into the water during the life of the plants or during their decay. Water absorbs some nitrogen, and equilibrium between the nitrogen contents of the atmosphere and of the sea is thus maintained. Hence we need not fear any noteworthy depletion of the nitrogen contents of the air. This conclusion is in accord with the fact that no notable accumulation of fixed nitrogen appears to have taken place in the solid and liquid constituents of the earth.

The reader may remember (compare page 57) that during the annual cycle of vegetation not less than one-fiftieth of the atmospheric contents in the carbon dioxide is absorbed. Since oxygen is formed from this carbon dioxide, and since the air contains about seven hundred times as many parts per volume of oxygen as of carbon dioxide, the exchange of atmospheric oxygen is about one part in thirty-five thousand. In other words, the oxygen of the air participates about one hundred times more energetically in the processes of vegetation than the nitrogen, and this is in accordance with the general high chemical activity of oxygen.

Before we close this chapter we will briefly refer to the peculiar phenomenon known as the Zodiacal Light, which can be seen in the tropics almost any clear night for a few hours after or before sunset or sunrise. In our latitudes the light is rarely visible, and is best seen about the periods of the vernal and autumnal equinoxes. The phenomenon is generally described as a luminous cone whose basis lies on the horizon, and whose middle line coincides with the zodiac. Hence the name. According to Wright and Liais, its spectrum is continuous. It is stated that the Zodiacal Light is as strong in the tropics as that of the Milky Way.