.>rAssi-:s oi' \ isuAL iuxarv stars. 463 



40 years at a distance of about i" with measured parallaxes of 

 about o". I, or at a radial distance of 10. These stars, it is at once 

 apparent, are of about the Sun's mass. Are there any of these 

 stars nearer than 10 radials ? The table says, No! Are there 

 an^• of these 40-year stars (9° i)er annum) of enormous mass and 

 at great distance but about 5th magnitude? Again the table 

 replies, Xo ! Do we know of any very close pair o".25 or so, 

 and faint, say yth magnitude, moving over 9° a year? If so, it 

 might be a very massive star at a great distance. Thus a double 

 star 1,000 times the Sim's mass, at a radial of 1,000, and at a 

 distance of o".2^, would move over 2°. 8 a year. There is no 

 such jiair. 



When every allowance is made for the weakness of some of 

 the assumptions, it is still impossible to resist the main conclu- 

 sion to be drawn from the tables, namely, that the stars are not 

 models of the Sun — that in general their gravitative power is very 

 much smaller indeed, and in some cases, as is shown in Table II, 

 seems to be quite insignificant. This especially applies to the spec- 

 tral classes B, A, and F. Classes O and M are not represented in 

 either table, indicating, again, very small gravitative power, whilst 

 classes G (solar type) and K have in a few cases gravitative 

 power of the same order as the Sun. 



If we take the simple means of Col. 9 in Table 1 by the classes 

 A, F, G, and K, we find the gravitative ])Ower is as follows: — 



A 0.02 



F 0.27 



G 1 . 23 



K 330 



And this table contains all the best-known and nearest stars. 

 The enormous disproportion thus revealed between lutninosity 

 and energy should not be surprising — a rude analogy is fur- 

 nished by considering a locomotive and its electric headlight, in 

 which practically all the energy is in the nearly black body of 

 the locomotive, and all the luminosity in the small lamp-filament. 

 The cases of a fe\\- t}'pical pairs will be specifically dealt with. 



X Orion. — Magnitude 3.49 {components 2^.66 and 5.56). 



Spectrum Oe^. 



This is one of the few double stars of the Oe class, and 

 like them all, show's no change whatever. That this is a real 

 and not optical pair is proved by their common pro]^er motion 

 o".028, which in the century this pair has been under observa- 

 tion, would have increased the distance by 2". 8, or from about 

 4" .4 to y" .2\ there has been no change. The effective gravita- 

 tive power of this pair and of all other Oe stars must be quite 

 trifling. 



There are no double stars of spectral type Oa to Od. The spectrum 

 Oe has helium and hydrogen lines and two bright bands at 4633 and 4688- 

 metallic lines are absent. 



