ATMOSPHERES OF THE PLANETS — RUSSELL 163 



would most of the other light gases such as neon and nitrogen (which 

 at the temperature even of the sun's surface is dissociated into 

 atoms). Free oxygen, too, would escape, but a good deal might be 

 retained in combination with silicon and the metals. As the gaseous 

 mass cooled, by expansion and radiation, drops of molten metal and 

 lava would form within it, as Jeffreys suggests, and fall toward the 

 center, building up a molten core. After the first turbulence was 

 over, there would remain a molten planet surrounded by an atmos- 

 phere containing heavy inert gases, such as argon, perhaps some car- 

 bon dioxide, and as much of the nitrogen and neon as had failed to 

 escape, Menzel and I, a few years ago, noticed that neon, while ap- 

 parently fully as abundant in the stars and nebulae as argon, is but 

 J^oo as abundant in the earth's atmosphere; while nitrogen, which 

 is cosmically an abundant element, showing strong spectral lines, 

 forms but a very small portion of the earth's mass. It appears, 

 therefore, that a mass of the earth's magnitude must have lost almost, 

 though not quite, the whole of its primitive atmosphere. 



Still following Jeffreys, it appears that, as the molten earth 

 cooled, the 2,000-mile deep sea of lava solidified first at the bottom 

 (where the melting point was greatly raised by pressure) and so 

 gradually to the surface. During this process great quantities of 

 gases, mainly water vapor, must have been evolved from the solidi- 

 fying magma, and escaped to the surface, forming a new atmosphere 

 which now would not escape, since the surface was cooler. With 

 solidification would come rapid superficial cooling, and an ocean 

 would bathe the rocky crust, leaving an atmosphere of moderate 

 extent. Carbon dioxide — evolved from the magma, and perhaps 

 partly primitive — would be a major constituent, along with nitrogen, 

 argon, neon, and other minor left-overs. The presence of free 

 oxygen seems very unlikely, for practically all volcanic rocks and 

 gases are unsaturated with respect to this element — the former 

 containing much ferrous iron and the latter being often actually 

 combustible when they meet the air. 



The present rich supply of oxygen appears to be a byproduct of 

 terrestial life. (This suggestion is more than a century old.) The 

 earth, indeed, may be regarded as an intensively vegetated planet, 

 from whose atmosphere the greedy plants extract the remaining 

 residue of carbon dioxide so rapidly that if it were not returned to 

 the air by combustion, respiration, and decay, the whole supply 

 would be exhausted in a decade or so. Oxygen removed from the 

 atmosphere by these processes is speedily returned by plants; but 

 there is another process of slow depletion which is irreversible. 

 During rock-weathering, about half the ferrous iron of the rocks is 

 oxidized to the ferric state. Goldschmidt (from whose admirable 



36923 — 36 12 



