600 Stoney — Cause of Double Lines in Spectra. 



series, to 3223 revolutions during each apsidal circuit in the most refrangible of 

 the observed pairs. Exact information on these points will be obtainable if ade- 

 quate observations can be made. 



S{g). The actual form of the elliptic orbit of each of the partials can be ascer- 

 tained from observations on the brightness of the lines. See below, p. 603. 



This is a considerable body of information about the motion in the molecules 

 which causes series S, all of which is within our reach if adequate observations 

 can be made. 



Series D. 



When we turn to the series of diffuse lines, we find that it resembles series S in 

 most respects, with, however, three notable points of difference : — the lines are 

 brighter in series D than in series S ; they are diffuse instead of sharp ; and some 

 satellites (or supposed satellites) are present. 



That the lines are brighter betokens that the partials of the primary motion of 

 the D electron are motions in larger ellipses than those of the 8 electron ; and how 

 much larger may be ascertained so soon as measures of their relative brightnesses 

 shall have been made. As to satellites, three supposed satellites are recorded, one 

 apparently midway between the lines of that which Professor Rydberg regards as 

 the fourth pair of the series. If, when adequate observations are made, it is found 

 to be really midway between them, it will indicate that the corresponding partial 

 Kes in a plane which is inclined to the plane of the apsidal motion at an angle which 

 can be determined so soon as the relative brightnesses of the three lines (the double 

 line and its satellite) shall have been measured. See Corollary,* p. 577. 



No similarly placed satellites are recorded of the other terms of series D, If 

 upon an adequate scrutiny it is found that there are none such, it will indicate that 

 all the partials of the primary motion, except one, lie in the plane of the apsidal 

 motion. But as there is one which does not lie in that plane, the primary motion 

 of electron D must be a curve of double curvature. 



Professor Rydberg thinks the observations warrant the conclusion that Ak is 

 the same In all the pairs of this series, and even that it is the same in all the three 

 series. If it has the same value in all the terms of series Z), it will indicate that 

 the primary motion of electron D is in an orbit of double curvature which, without 

 changing its form, shifts round in a definite plane (which is the plane of all but one 

 of its partials) with an apsidal motion of which the frequency is Ak / 2. 



* What it is here convenient to regard as apsidal motion in a plane inclined to the plane of the partial, 

 is identical with that which in the Corollary (p. 577) is perhaps more accurately described as precessional 

 motion unaccompanied by apsidal motion. 



