34 Prof. I\ E. Chase on our 



17. Jupiter and San thus appear to.be companion consti- 

 tuents) of a binary star; and the point of primitive rupture 

 should be sought at the secular-perihelion centre of gravity. 

 Bodies falling towards that point, on approaching Sun, are 

 subject to a force of about 1048 towards Sun, and 1049 towards 

 the slowly moving common centre of gravity. There are 

 therefore two nodal points, with the least resistance to motion 



nearly midway ( ^Keyi ) between them. If Sun is gaseous, as 



Hunt, Faye, and others have supposed, there should hence 

 arise linear oscillations of 2 x 2r synchronous with the circular 

 oscillations of %irr. The corona may perhaps be due to such 

 radial oscillations. 



18. The 15th accordance gives for the mass of Sun-*- Jupiter, 

 1049-875-2 = 1047-875; the 17th, 1049-871-2 = 1047-874, 

 Bessel's estimate being 1047-879 ±'235. 



19. The discrepancy between the two astronomical estimates 

 for the velocity of light seems to have arisen from ignorance 

 of the intranodal oscillation. Delambre, from his discussion 

 of more than 1000 eclipses of Jupiter's first satellite *, esti- 

 mated the time of light-passage from Sun to Earth at 493*198 

 seconds ; Struve, from the phenomena of aberration, at 497*827 

 seconds. If the time of traversing 212*86 solar radii is 493*198, 

 the time for 214*86 r should be 497*831, which differs from 

 Struve's value by less than yoW °^ ^ P er cem ^ 



20. If Earth was at the nebular nucleal surface when the 

 Jovi-Saturnian ring was nebularly atmospheric, the vis viva of 

 interior nucleal rotation varied as r, and the velocity of result- 

 ing planetary revolution as r*. We thus obtain for the theo- 

 retical time of present solar rotation, 



n/214-86 : 1 : : 365*2 days : 24*912 days. 



The lowest estimate from observation is Spicer's, 24*624 ; the 

 highest, that of Schwabe, 25*507. 



21. The laws of central forces require that provision should 

 be made for radial oscillations, tending towards the time-limit 

 of isoradial circular oscillations ( \/32 and v 8), for tangential 

 velocities, varying inversely as the times and directly as the 

 fourth root of central isoradial tendencies, for centres of oscil- 

 lation in lines of force, and for oscillations between systematic 

 and locally dominant centres. We have already seen (15, 17, 

 18) how closely the relative masses of Sun and Jupiter provide 

 for the last requirement ; if they provide also for the others, the 

 centre of oscillation for Sun's possible atmosphere should be at 



* Stockwell, Proc. Airier. Assoc, vol. xx. pp. 76, 77. 



