COMET OF 1889-1896-1903 279 



below, where the unit of distance is the equatorial semi-diameter 

 of Jupiter. 



These values being obtained from the masses and distances as 

 given by Newcomb. 



The following table shows at a glance the character of the 

 approach to each of the four satellites : the first column con- 

 tains the number of the satellite, the second the smallest pos- 

 sible distance between the orbit of the satellite and that of the 

 comet in terms of the radius of Jupiter, and the third column 

 the radius of the sphere of attraction of the satellite. 



Thus there was no approach near enough to cause the 

 slightest change in the relative orbit of the nucleus about 

 Jupiter. In no case did the nucleus pass nearer to the satellite 

 than six times the radius of the " sphere of attraction " of that 

 satellite ; and at this distance the perturbations of the satellite 

 would be inappreciable. 



36. Again, as to the disruption of the comet ; the above 

 shows pretty clearly that it could not have been caused by the 

 action of any one of the four outer satellites. For the radius 

 of the comet, although a very uncertain quantity could hardly 

 have been larger than the radius of Jupiter, and in no possible 

 case did any of the cometary matter pass within the "sphere" 

 of a satellite. 



In all that goes before, the fifth satellite has been left out of 

 consideration. The mean distance of this body is about 2.6 

 radii of Jupiter, while the distance of the ascending node of the 

 comet's orbit was only 2.36 radii. Thus if the satellite was in 



