CONDITION OF THE LARGER PLANETS. 193 



called his ' square shouldered ' aspect. These observa- 

 tions are far too well-authenticated, and were made by 

 observers far too skilful, to be open to doubt or cavil. 

 They cannot possibly be explained except by assuming 

 that the outlines of Jupiter and Saturn are variable to 

 such an extent that the variations appreciably affect 

 the figure of the planets. Such variations, involving 

 differences of level two or three thousand miles, are 

 utterly incredible, and in point of fact impossible, in 

 the case of planets like our earth. The heat generated 

 by such changes would of itself suffice to melt and in 

 large degree to vaporise the crust for many thousands 

 of square miles around the scene of upheaval or depres- 

 sion, so that we should thus have, but in another way, 

 the heat which my theory indicates. On the other 

 hand, such changes of outline in a planet whose appa- 

 rent outline is not formed by its real surface, but by 

 cloud layers thousands of miles above the real surface, 

 are very easily explained. Nay, they are to be ex- 

 pected (though only as rare phenomena). We know 

 that cloud-belts sometimes form, or are dissipated, 

 rapidly on the face of the disc. Equally, therefore, 

 they must sometimes form or become dissipated rapidly 

 at parts of the planet so placed as to form the apparent 

 outline. There would then be a rapid change of out- 

 line, such as must have occurred in the case of the 

 apparent reappearance of Jupiter's second satellite. 

 Slowerchanges inthe cloud-belts would correspond tothe 

 changes of shape observed in Saturn's case, and would 

 explain Schroter's observation that at times the outline 

 in. o 



