162 ANNUAL. REPORT SMITHSONIAN INSTITUTION, 193 5 



of light substances, which the smaller terrestrial planets had not been 

 able to keep from diffusing away into space. This has been fully 

 confirmed by later studies. 



From the ellipticity of a planet and the changes in its satellites' 

 orbits caused by the attraction of its equatorial bulge, information 

 may be obtained regarding the degree to which the density increases 

 toward its center. Applying this to Jupiter and Saturn, Jeffreys 

 concludes that they contain cores of rock and metal, like the inner 

 planets, surrounded by vast shells of ice — frozen oceans thousands of 

 miles deep — and above this, again, atmospheres of great extent. 

 Throughout most of the atmospheres, the pressure must be so great 

 that the gas is reduced to a density as great as it would have if lique- 

 fied, or even solidified, by cooling. Indeed, Wildt believes that the 

 enormous pressure would actually solidify even the " permanent " 

 gases. 



Now this outer layer is of low density — less than 0.78 for Jupiter 

 and 0.41 for Saturn — according to Wildt's calculations. This ex- 

 cludes all but a few possible constituents. Frozen oxygen has a 

 density of 1.45, nitrogen 1.02, ammonia 0.82. Only hydrocarbons 

 (methane 0.42, ethane 0.55), helium (0.19), and hydrogen (0.08) 

 come within the limits even for Jupiter. We can therefore conclude, 

 from considerations of density alone, that the outer parts of Jupiter 

 probably, and of Saturn certainly, contain great quantities of free 

 hydrogen or helium. Uranus and Neptune are similar to Jupiter. 



It is generally believed that the planets have been produced, in 

 some way or other, from matter ejected or removed from the sun. 

 No really satisfactory theory of the process of formation has yet 

 been devised; but no other hypothesis has yet done better, and the 

 isolation of the sun and planets in space makes a common origin 

 highly probable. 



Now we know the composition of the sun — at least of its outer 

 layers — much better than we do that of the planets. Quantitative 

 spectroscopic analysis, though still beset with difficulties, has ad- 

 vanced far enough to show that most of the sun's outer layers is 

 comj^osed of hydrogen; next come helium, oxygen, and carbon, fol- 

 lowed by nitrogen, then silicon and the metals. A mass of matter 

 removed from the sun and allowed to cool without serious loss would 

 therefore closely resemble the major planets. If small enough to 

 lose all its atmosphere, it would be like the moon or the asteroids — 

 though there are difficulties in seeing how such small masses could 

 have escaped diffusing away altogether before the more refractory 

 constituents solidified. 



The history of a body of intermediate mass is more interesting. 

 Hydrogen and helium would be lost while it was still very hot. So 



