160 ANNUAL REPORT SMITHSONIAN INSTITUTION, 19 3 5 



At this high pressure the lines flowed together, and produced dif- 

 fuse bands ; but the agreement of these with the planetary bands was 

 so complete as to be decisive. 



A further, and wholly conclusive, test could be added. The funda- 

 mental frequencies of vibration of the methane molecule were al- 

 ready known, from observations in the infrared. For the higher 

 harmonics of these vibrations the frequencies are not exact multi- 

 ples of the lowest, but nevertheless bear a simple numerical relation 

 to them (as is well known in the case of other gases). Applying 

 this test, the strongest bands (including Huggins' band in the orange, 

 and the one coincident with the blue hydrogen line) were found to be 

 harmonics, from the third to the eighth, of one of the fundamental 

 frequencies, while another slower vibration was represented by all 

 its harmonics from the eighth to the sixteenth. The remaining 

 bands were accounted for by combinations of these harmonics with 

 other known frequencies, all of types consistent with the well-estab- 

 lished rules which govern band spectra. Forty-one bands in all 

 have been identified. Many of these appear only in Uranus and 

 Neptune, and have not yet been produced in the laboratory, but the 

 harmonic relations just mentioned make their identification certain. 

 The higher gaseous hydrocarbons, ethane, ethylene, and acetylene, 

 all have bands in places clear of disturbance by the methane ; and all 

 were looked for in vain. All the planetary bands of any importance 

 are accounted for by methane alone — it is a clean sweep. 



From the published data, it appears that the amount of methane 

 above the visible surface of Jupiter is of the order of one mile-at- 

 mosphere. There must be much more on Uranus, and especially on 

 Neptune — 25-mile atmospheres, according to Slipher and Adel. 



There is still plenty of work to do upon these bands, but mainly 

 for the theoretical investigator. Adel calculates that the band at 

 A 5430, when fully resolved, should consist of 18 different overlap- 

 ping systems of many lines each. Fortunately, the astrophysicist 

 need not wait to draw his conclusions till this has been completely 

 analyzed. 



The results of observation can be summarized in a sentence. Large 

 planets have atmospheres containing hydrogen compounds; middle- 

 sized planets, atmospheres containing oxygen compounds; and small 

 planets no atmospheres at all. The reason, in the last case, was 

 found by Johnstone Stoney, in 1897. It is simply that small bodies 

 have not sufficient gravitative power to keep their atmospheres from 

 diffusing away into the vacuum of interplanetary space. At the sur- 

 face of any planet, there is a certain velocity of escape, depending 

 only on its mass and radius. A body projected from its surface, 

 in whatever direction, with this or any higher velocity, will fly off 

 in a parabolic or hyperbolic orbit and never return — unless, indeed, 



