208 ASTRONOMY. 



The first line gives the ratio of the light of Jupiter to that of each of 

 the satellites, the last the albedo relative to Jupiter. To obtain the ab- 

 solute albedo for each satellite, the relative albedo must be multiplied 

 by the albedo of Jupiter (O.GIO). The view that the light of the satel- 

 lites is variable is not confirmed by Prof. Pickering's observations. 



A large number of jihotometric measures of the eight satellites of 

 Saturn were made in 1877-8, special attention being devoted to the 

 variations of light of lapetus, which were found to extend from 40 to 

 140 (taking the mean brightness as 100). The following formula was 

 found to represent the variations of light : 



L = 100 — 50 sin v + 10 cos 2v, 

 where L is the light of the satellite at any longitude v. 



On this formula Prof. Pickering remarks, "The absence of the term 

 c .cos V shows that if the variation in light is due to unequal brightness of 

 the two hemispheres of lapetus, one-half of each will always be turned 

 towards Saturn. In other words, it would present to an observer at the 

 north pole of Saturn the appearance of a half moon, the semicircle to 

 the right being about four times as bright as that to the left. 



" Some doubt is thrown on the term e cos 2t', since the comparisons 

 with Saturn do not confirm its presence. Since the value of e is positive, 

 it cannot be sux)posed to indicate that lapetus is elongated in the direc- 

 tion of Sat'^rn, although a slight elongation of this kind is probably im- 

 plied in the assumption that the satellite turns once on its axis during 

 each of its revolutions. An elongation suificient to i)roduce this effect 

 might be caused by tbe attraction of Saturn, but it v.ould be far too 

 small to be perceptible photometrically. To make e = -f 1, the equato- 

 rial diameter of the satellite turned towards Saturn should be exceeded 

 by that at right angles to it in the ratio of 9 to 11 5 so great a differ- 

 ence does not seem probable. The term e cos 2v could also be accounted 

 for by two bright or dark spots on the satellite. A dark spot on one 

 side covering less than a hemisphere would also give a variation in light 

 closelj" resembling that given by that in the formula. The most natural 

 explanation, however, is that the dark and light portions are irregularly 

 distributed on lapetus, like the land and water on our Earth, and that 

 one hemisphere is, on the whole, much darker than the other. The 

 smaller variations may be assumed to be such that the formula given 

 above represents them closely. " 



The following are the equivalent diameters of the satellites in miles 

 foundby Prof. Pickering: Mimas292±9; Encleadus 370±10; Tethys 

 570±1S5 Dione 542±17; Ehea 745±27; Titan 140G±52; Hyperion 

 193±5; lapetus, mean 4SGzt4, max. 574, min. 307. 



The satellites of "Uranus and Neptune have the following equivalent fj 

 diameters (in miles): Titania 5SGzt:lC; Oberon 544±15; satellite of fl 

 Neptune 22G0zliG0. The large equivalent diameter of the satellite of f| 

 Neptune is noteworthy. 



Determinations are also given of the equivalent diameters of some i 



