THE PLANETS. 425 



also atmospheres which, in consequence of the peculiarity of 

 their condensation, appear to us variable; in Saturn, indeed, 

 sometimes as interrupted bands. 



Although in the important classification of the planets into 

 two groups of interior and exterior planets, the general cha- 

 racters of absolute magnitude, density, flattening at the 

 poles, velocity of rotation, absence of moons, present them- 

 selves as dependent upon the distances, *. e., from their semi 

 orbital axes, this dependence cannot be affirmed of each 

 one of these groups. Up to the present time we are igno- 

 rant, as I have already remarked, of any internal necessity, 

 any mechanical law of nature, which (like the beautiful law 

 which connects the square of the periods of revolution with 

 the cube of the major axes) represents the above-named 

 elements of the order of succession of the individual planet- 

 ary bodies of each group, in their dependence upon the 

 distances. Although the planet which is nearest to the sun 

 (Mercury) is the densest, even six or eight times denser than 

 some of the exterior planets, Jupiter, Saturn, Uranus, and 

 Neptune, the order of succession, in the case of Venus, the 

 Earth, and Mars, or Jupiter, Saturn, and Uranus, is very 

 irregular. The absolute magnitudes do generally, as Kepler 

 has already observed (Harmonica Mundi, vol. iv. p. 194; 

 Cosmos, vol. i. pp. 77-82), increase with the distances; but 

 this does not hold good when the planets are considered indi- 

 vidually. Mars is smaller than the Earth, Uranus smaller 

 than Saturn, Saturn smaller than Jupiter, and succeeds 

 immediately to a host of planets, which, on account of their 

 smallness, are almost immeasurable. It is true the period of 

 rotation generally increases with the distance from the sun; 

 but it is, in the case of Mars, slower than in that of the Earth, 

 slower in Saturn than in Jupiter. 



The external world of forms, I again repeat it, can only be 

 represented in the enumeration of relations of space, as some* 



