ATMOSPHERES OF THE PLANETS — RUSSELL 157 



while lines produced in the Earth's atmosphere are of course un- 

 affected. Were this shift great enough the planetary and telluric 

 lines would appear double, and the former, even though faint, could 

 readily be detected. The greatest available shift is not enough to 

 resolve the lines completely ; but measures of the blended lines suffice 

 to show whether any important planetary contribution is present. 

 A still more delicate test is afforded by microphotometer measures 

 of the contours of the lines, which would reveal even a slight asym- 

 metry. These observations are very exacting, requiring high dis- 

 persion and a great deal of light, so that the best evidence is that 

 from the great coude spectrograph of the 100-inch telescope at 

 Mount Wilson. St. John and Nicholson found, in 1922, that there 

 was no perceptible trace of planetary lines in Venus, and Adams and 

 Dunham, in 1934, have come to the same conclusion in the case of 

 Mars. An amount of oxygen, on either planet, equal to a thousandth 

 part of that above an equal area on earth, could certainly have been 

 detected. For water vapor, the tests have so far been less delicate, 

 and are not fully decisive — though the quantity present on either 

 planet must be small. More delicate tests, with stronger lines, may 

 soon be made on new red-sensitive plates. 



There can be no reasonable doubt, on quite different evidence, that 

 some small amount of water vapor is actually present in Mars' 

 atmosphere. Radiometric observations of the planet's heat show 

 definitely that the surface rises to temperatures above 0° Centigrade 

 at noon every day in the Martian tropics, and at the pole at mid- 

 summer, though falling far below freezing at night. The polar caps 

 must therefore really be composed of snow, and evaporate into water 

 vapor, even if the pressure is so low that the ice turns directly into 

 vapor without melting. The only plausible alternative suggestion — 

 carbon dioxide — would volatilize at much lower temperatures than 

 the actual polar caps do. But, judging from the amount of solar 

 heat available to evaporate them, the polar caps must be very thin, 

 probably only a few inches thick. The vapor resulting from the 

 gradual sublimation would never attain any considerable density, 

 and might easily fail of detection by the tests which have so far been 

 practicable. 



No such independent evidence is available for Venus, but Adams 

 and Dunham, in 1932, discovered, in the infrared region of her spec- 

 trum, three beautifully defined bands with heads at A7820, A7883, and 

 A.8689, and evidently of atmospheric origin. They had not then been 

 observed elsewhere; but an immediate suggestion regarding their 

 origin was obtained from the theory of band-spectra — by that time 

 well developed. The spacing of the individual lines in a band arises 

 from the rotation of the molecule and depends upon its moment of 

 inertia. For the new planetary band, it showed that the otherwise 



