93 COSMOS. 



viatmg biu little from circles. It is only in the ca&e of oin 

 rnoon, and perhaps in that of the first and innermost of the 

 satellites of Saturn (0*068), that we discover an eccentricity 

 greater than that of Jupiter ; according to the very exact ol> 

 rcrvations of Bessel, the eccentricity of the sixth of Saturn'? 

 satellites (0*029) exceeds that of the Earth. On the extremes! 

 limits of the planetary system, where, at a distance nineteen 

 times greater than that of our Earth, the centripetal force of 

 the Sun is greatly diminished, the satellites of Uranus (which 

 have certainly been but imperfectly investigated) exhibit tin 

 most striking contrasts from the facts observed with regard to 

 other secondary planets. Instead, as in all other satellites, of 

 having their orbits but slightly inclined toward the ecliptic 

 and (not excepting even Saturn's ring, which may be regard- 

 ed as a fusion of agglomerated satellites) moving from west tc 

 east, the satellites of Uranus are almost perpendicular to the 

 ecliptic, and move retrogressively from east to west, as Sir 

 John Herschel has proved by observations continued during 

 many years If the primary and secondary planets have been 

 formed by the condensation of rotating rings of solar and plan- 

 etary atmospheric vapor, there must have existed singular 

 causes of retardation or impediment in the vaporous rings re- 

 volving round Uranus, by which, under relations with whbh 

 we are unacquainted, the revolution of the second arid fourth 

 of its satellites was made to assume a direction opposite to that 

 of the rotation of the central planet. 



It seems highly probable that the period of rotation of all 

 secondary planets is equal to that of their revolution round 

 the main planet, and therefore that they always present to 

 the latter the same side. Inequalities, occasioned by slight 

 variations in the revolution, give rise to fluctuations of from 

 G to 8, or to an apparent libration in longitude as well as 

 in latitude. Thus, in the case of our moon, we sometimes 

 observe more than the half of its surface, the eastern and 

 northern edges being more visible at one time, and the west 

 ern or southern at another. By means of this libration* we 

 are enabled to see the annular mountain Malapert (which oc- 

 casionally conceals the Moon's south pole), the arctic land- 

 scape round the crater of Gioja, and the 'argf gray plane near 

 Endymion, which exceeds in superficial extent the Mare Va~ 

 porum. Three sevenths of the Moon's surface are entirely 



* Beer and Madler, op. cit., 183, s. 208, and $ 347, 8 33^; ami U 

 :hi'ir I'/tyt. Kennlniss der himniL Kurp-.r, s. 4 uncJ 69. Tub I (Physic 

 al History of the Heavenly Bodies). 



