ATMOSPHERES OF THE TLANETS RUSSELL 161 



it meets with some obstacle or resistance on its outward way. For 

 the moon this velocity is 2.4 kilometers per second; for the earth, 

 11.2; for Jupiter, CO. 



Now the molecules of any f^as are continually flying about in all 

 directions, with average speeds which depend upon their weights. 

 At 0° Centigrade the average speed for a hydrogen molecule is 1.84 

 km/sec.; for oxygen, 0.46; for carbon dioxide, 0.39. If an at- 

 mosphere of hydrogen could be put upon the moon, every molecule 

 that was moving but a little faster than the average would fly off at 

 once into space, unless it was thrown back by collision with another, 

 and tlie atmosphere would diffuse away in a very short time. With 

 an escape velocity three times the average speed, enough fast-moving 

 molecules would get away to reduce the atmosphere to half its orig- 

 inal amount in a few weeks (according to Jeans). The rate of loss 

 falls off very rapidly beyond this, so that, with an average velocity 

 one fifth that of escape, the atmosphere would remain for hundreds 

 of millions of years. 



The moon's surface reaches a temperature exceeding 100° C. dur- 

 ing every rotation, and it follows that neither air nor water vapor 

 could permanently remain above its surface. If at any time in its 

 past history, it has been really hot, like molten lava, it could have 

 retained no trace of atmosphere. For Mercury, the escape velocity 

 is half as great again as for the moon ; but the planet, being so near 

 the sun, is much hotter, and it can hold only the heaviest gases. 

 Mars, with an escape velocity of 5 km/sec, could not hold hydrogen 

 but should retain water vapor — as it appears to have done — and all 

 heavier gases. Venus and the earth, at their present temperatures, 

 should retain even hydrogen, and the major planets would do so 

 even if incandescent. 



This reasoning explains the cases of Mercury and the moon, and 

 leads to the important conclusion that all smaller bodies, such as the 

 asteroids and satellites, must be wholly devoid of atmosphere — ex- 

 cept perhaps bodies like Neptune's satellite, which is relatively mas- 

 sive, and must be very cold. We cannot be sure about Pluto, for we 

 know neither its size nor its mass; but it is probable that, at most, 

 it may have a thin atmosphere, like Mars. 



The same principle was invoked, shortly after its discovery, to ex- 

 plain the great difference in mean density between the major and the 

 terrestrial planets. The moon, Mercury, Mars, Venus, and the earth 

 all have densities between 3.3 and 5.5 times that of water. The rest 

 are almost certainly what we know the earth to be, spheroids of 

 rock, with cores of metallic iron of varying sizes. For the major 

 planets, the densities range from 1.6 for Neptune to 0.7 for Saturn. 

 Moulton suggested, about 1900, that they contained great quantities 



