564 Mr. Arthur Stanley Eddington [March 26, 



Jupiter's year, or time it takes to go round the sun, is equal to 

 nearly twelve of ours. The satellite's " month," or time it takes to 

 go round Jupiter, is somewhat over two of our years. It follows 

 that the satellite goes rather more than five times round Jupiter 

 whilst Jupiter goes once round the sun. We may say that for 

 J VIII there are five months in the year, borrowing terminology 

 of the Earth-Moon system. Now this is a record. All the other 

 satellites in the solar system have a much greater number of 

 " months " to the " year," usually several hundreds, sometimes 

 even thousands. Our own moon previously held the record with 

 between 12 and 13 months to the year, but the new satellite far 

 outdoes that. Now this is a particularly interesting piece of record 

 breaking, because the number of satellite's months in the planet's 

 year expresses, generally speaking, the proportionate share which the 

 planet and sun have in controlling the motion. When, as is usually 

 the case, the number is several hundreds, that means that the share 

 of the planet is far greater than that of the sun, and the orbital 

 motion is easy to compute ; in the case of our Moon, which, as I have 

 said previously, held the record, the disturbance by the sun is con- 

 siderable, and tbe prediction of the motion is a very complicated 

 problem ; but, when, as in the case of J YIII, the number gets 

 nearly down to five, ordinary methods of calculation fail altogether, 

 and, when Mr. Cowell attacked the problem, he had to devise an 

 entirely new procedure for his purpose. 



These four newer satellites of Jupiter thus form a remarkably 

 varied group, each individual presenting its own particular problems. 

 It is hard to say which could least be spared, whether the fifth 

 satellite circling so closely and so rapidly ; or YI and YII, the 

 twins, with their curiously interlocked orbits ; or the eighth, which 

 moves at a distance from the planet so great that at first it could 

 hardly be credited. 



There is one circumstance, however, which renders J YIII the 

 most immediately interesting of all — it goes round Jupiter back- 

 wards. All the other seven go round in one direction; Jupiter 

 himself rotates in the same direction ; but the eighth satellite moves 

 round the opposite wa}'. This is a most significant fact ; if we are 

 right in our interpretation, it is a clue which reveals a curious 

 chapter in the past history of the solar system. The speculation 

 which it opens out is not a new one, but this fact gives a strong 

 confirmation to what must previously have been regarded as a plaus- 

 ible but daring hypothesis. In 1899 Phoebe, the ninth satellite of 

 Saturn, was discovered, and after considerably puKzling its dis- 

 coverer. Professor W. H. Pickering, by its unaccountable motion, 

 was at last found by him to have a retrograde motion : that is, to go 

 round Saturn in the opposite direction to his other satellites, just as 

 we now find J YIII behaves. Phoebe is a long way the outermost of 

 Saturn's satellites, just as J YIII is the outermost of Jupiter's, 



