We reach the same conclusion when investigating the ability of the planets to reflect the sunlight falling upon them. The bodies which possess an atmosphere hold also suspended therein clouds of water or ice, and also of dust, whirled up from below. These floating particles reflect light far more efficiently than the solid or fluid surface of a planet. The Moon can now reflect 7.3 per cent. of the sunlight and Mercury 6.9 per cent. (H. N. Russell, Proceedings Nat. Acad., 1916). These numbers lie so close that they may be considered practically the same within errors of observation.
It is therefore probable that Mercury is as devoid of an atmosphere as the Moon. The opposite extreme is represented by Venus, which reflects not less than 59 per cent. of the sunlight received, according to H. N. Russell. Terrestrial clouds were found by Abbot to return 65 per cent. We believe from astronomical observations that the entire surface of Venus is hidden behind a thick opaque cloud-covering. The slight difference between O.65 and O.59 may be due to errors of observation, but also to a small absorption of light in those parts of Venus’s atmosphere which are outside of the clouds. Saturn and Jupiter are very similar to Venus in this regard with 63 and 56 per cent. respectively. The gases above the clouds on these planets extinguish to a considerable extent the sunlight reflected from the clouds, as apparent from their spectra. (Compare [Fig. 13].) Hence the value 0.63 given by Russell for Saturn is probably too high. Concerning Jupiter it has been observed that its red light becomes deeper when the sunspots are few, but whiter when the spots are numerous. The sunspots have been found to favour the formation of high clouds, such as cirrus, and this would seem to apply also to Jupiter; when spots are plenty, the clouds are high, and consequently the absorbing layers above, which cause the red colour, are thinner, so that Jupiter will then shine with a whiter—less red—lustre than when the sunspots are rare.
The two outmost planets, Uranus and Neptune, return, according to Russell, 63 and 73 per cent. respectively of the sunlight received. These figures are probably too high. They do not agree well with Slipher’s records of their spectra ([Fig. 12]).
There now remains Mars. This planet approaches the Moon inasmuch as it reflects only 15.4 per cent. of the sunlight arriving to the orb. Everything points to the conclusion that the atmosphere of Mars is very thin. Lowell estimates, on somewhat meagre grounds, however, that on each square metre of the planet rests only 22 per cent. of the mass of air supported by each square metre of the Earth’s surface.
It would naturally be very interesting to ascertain the amount of sunlight our Earth throws back into space. This we cannot measure, as we cannot place our instruments outside of the Earth’s cloud-mists nor can we read them there. Not less than 52 per cent. of the Earth is covered with clouds, whose whiteness (Latin: Albedo) is 65. Thus the clouds alone return 0.52 × 0.65 = 0.338 parts of the sunlight. Of this portion a fraction amounting to about 4 per cent. is extinguished in the air above. The remainder is 0.325. Atmosphere and suspended dust reduce the sunlight over the cloud-free part, i. e. 48 per cent. of the Earth, by 60 per cent, half of which returns to space, while the other half reaches the ground in the form of light from the sky, and of this fraction again about 4 per cent. is reflected into space; these two items added give 0.15. Finally, the 40 per cent. sunlight directly received by the Earth’s surface is reflected to the extent of 6 per cent. by the oceans and by the generally moist ground; deserts and bare rocks reflect about twice as much, but their total area is comparatively small; of this 6 per cent. reflected light, 70 per cent. reaches outside space; thus we obtain 0.48 × 0.40 × 0.06 × 0.70 = 0.008. In all, therefore, the amount of reflected sunlight is 0.338 + 0.15 + 0.008 = 49.6 per cent. If the air were free from clouds, the reflexion-number or Albedo would be 33 per cent., or considerably higher than that of Mars. When now half or a little more (52%) of the Earth’s surface is overcast with clouds and this portion therefore has the whiteness of Venus, the figure 49.6 (Russell calculates the figure 45) for the entire Earth naturally falls closer—almost 3.6 times—to 59, the figure of Venus, than to 15.4, the figure for Mars. We may also compare the value 33 per cent., which applies to the cloud-free portion of the Earth, with the value 15.4 per cent. for Mars, which is almost without clouds, and with the value 7.3 per cent. for the Moon, which has neither clouds nor dust, because it lacks an atmosphere. We can then conclude that the atmosphere of our Earth holds almost three times as much dust suspended over each square metre as does Mars, and this in spite of the smaller gravitational force on Mars, which is about 37.5 per cent. of that on Earth. Taking proper account of the low temperature on Mars we may easily compute, by means of a formula given by Stokes, that a particle of dust should sink 2.3 times slower on Mars than on Earth. When, nevertheless and in spite of frequent but thin mists, so few particles of dust float in the atmosphere of Mars, the conception inevitably comes to our mind that the air on that planet must be extremely rarefied so that the wind-puffs have little power to raise the dust from the ground. Lowell estimated the barometric pressure at the surface of Mars to be about 64 mm. (2.52 inches), and Proctor gives about twice this figure. There appears to be ground for considering already the former value too high; both are very uncertain. If we accept that of Lowell, we find that each square metre of the surface of Mars supports a column of air, whose mass is only about one-fifth of the mass resting on each square metre of ocean surface on Earth.
Fig. 14. The planet Venus, with sunlit atmosphere (to the left), as observed by Langley at the Venus passage December 6, 1882.
The dense clouds which float above Venus have long ago led to the assumption that the atmosphere of that planet must be far deeper than that of the Earth. Its strong refractory power has also contributed to this belief. When Venus is close to the sun-disc the dark body stands forth surrounded by a ring of light (see [Fig. 14]). It is, however, recognized that this phenomenon requires no greater air density than that on Earth for its appearance. In this connection we should remember that the inside limit of the vapour shell which we in this manner observe, is the cloud-wall, not the ground. And these clouds, we have every reason to believe, float on account of the heat prevailing at a great height in the atmosphere, so high in fact that they form an impenetrable wall already where the cirrus clouds appear in our sky. If these suppositions are correct, the light-ring mentioned is caused by a quarter only of the air-masses on Venus, and its total air-covering must be far deeper than that of the Earth. The latter occupies probably in this respect as well as in reference to position in space a middle ground between Mars with its extremely thin and Venus with its comparatively dense atmosphere. If so, we might expect the atmosphere on Mercury to be denser yet, while we already have seen that it is almost wholly lacking on that planet. The explanation is that Mercury has lost its spin around its own axis and therefore always presents the same side to the Sun—just as the Moon and probably all other satellites turn one side only toward their respective central bodies—hence the opposite side becomes so cold that all gases are there condensed to fluids or solids except the two most volatile ones, hydrogen and helium, which on the other hand leave the planet on its hot side. If Venus, therefore, as held by several astronomers from Schiaparelli to Lowell, always turned one side only toward the Sun, this planet also would be without any perceptible air-covering. According to investigations by Bjelopolsky of Venus’s spectrum, which investigations, however, are in complete disagreement with corresponding measurements by Slipher, that planet has a period of rotation on its own axis of about 29 hours. This figure is very uncertain and a new determination is therefore highly desirable.
In order to understand the atmospheres of the planets, it is of great interest to ascertain the composition of the air that surrounds the Earth. Our knowledge in these matters has grown considerably of late. We shall in the main follow the presentation by Dr. Wegener of Marburg.
We know at present with considerable accuracy which gases enter into the air. Besides the previously well-known nitrogen and oxygen which contribute the bulk, 78.1 and 20.9 per cent. respectively, of the total volume at the earth’s surface, we find water vapour in proportions changing with localities and times, and it is for this reason left out when fixing the various percentages; further, carbon dioxide 0.03 volume per cent. and the rare gases discovered by Rayleigh and Ramsay, argon, 0.932 per cent., neon 0.0012 per cent., helium 0.0004 per cent., krypton 0.000005 per cent., and xenon 0.0000006 per cent. Each one of these constituents diminishes in quantity with height in accordance with the so-called barometer-formula and the rate is the more rapid the heavier the gas. Krypton and xenon, therefore, which are two and one half and four times heavier than oxygen, occur mainly in the lower strata. The percentage of helium on the other hand, a gas eight times lighter than oxygen, should increase rapidly with height. If the air consisted of a mixture of oxygen and helium at 0° C. (32° F.) the former would decrease to one half at an elevation of 5 km. (3.1 miles) but the latter not before we had ascended 40 km. (25 miles) (eight times higher than for oxygen, as the weights are in the ratio of 1 to 8). At that altitude, oxygen would have decreased in the proportion 1:28 = 1:256. When, as actually is the case, there is 50,000 times more oxygen than helium at the surface this ratio should decrease in the proportion 128:1 at a height of 40 km. (25 miles). Ninety kilometers (56 miles) above the surface helium should overbalance oxygen and thereafter rapidly gain in preponderance. This holds true provided no agitation takes place in the form of vertical currents of air.