ON STELLAR DISTRIBUTION AND MOVEMENTS. 20!) 



On the other hand, a resultant force of all the matter of our system 

 seems adequate to produce the stellar motions, though one would rather 

 expect to find more definite traces of a central force and of distinctly 

 radial motions. Perhaps the mistake is to attribute a centre of gravity 

 to the stellar system, which of course does not act as though it were 

 concentrated at a point. A useful, though very rough, idea of the 

 magnitude of the force which the collective attraction of the stars 

 exerts may be gained from the comparison of light and gravitation ; 

 since both vary as the inverse square law, we may assert that roughly 

 speaking the attraction towards any part of the sky is measured by the 

 total light of that part of the sky. The total light of the sky seems 

 to be about that of 2,400 first magnitude stars, and I suppose the total 

 gravitation would bear the same ratio to that of an average first magni- 

 tude star. Owing to the central position of the sun the pull is pretty 

 uniform in all directions, and there can be little resultant force, but 

 on an eccentrically placed star it would be very appreciable, considering 

 that it would act in nearly the same direction for an almost unlimited 

 time. 



A very interesting suggestion as to stellar velocities has been put 

 forward by Dr. Halm. 21 According to his view the Orion stars have 

 small velocities not because they are young but because they are 

 heavy. He proposes a law of equipartition of energy, so that in a 

 mixture of stars of different types the average value of m v 2 would be 

 the same for all types. This is the law which prevails in a mixture of 

 gases containing heavy and light molecules. On the observational 

 side our knowledge of the masses of stars is rather small; but such 

 determinations as have been made appear to support Halm's con- 

 tention. The Orion stars especially are considerably heavier than 

 the others. Also, if we consider the stars as being all formed about 

 the same epoch, it would be natural to expect that the evolutionary 

 development has proceeded fastest in the least massive stars, so that 

 the order of mass may well be also the order of evolution. But if the 

 argument of the preceding paragraphs is sound, the theoretical causes 

 for the law of equipartition cannot be admitted. It is true that equipar- 

 tition of energy ultimately results, not only when the individuals con- 

 sidered interact by collisions, as in the case of a gas, but also when any 

 kind of interchange of momentum takes place between the individuals 

 subject to the conservation of energy. The gravitational perturbations 

 of the neighbouring stars might thus in time bring about equipartition 

 of energy. But from what has already been said it appears that these 

 attractions have so far had an insignificant effect on stellar velocities. 

 The time required to produce equipartition would seem to be far too 

 great. 



Meanwhile there is yet a third possibility that might be considered. 

 Suppose that Orion stars move slowly, not because they are young, 

 or because they are heavy, but because they are distant. Velocities of 

 stars being mainly attributable to a central force of the universe might 

 be expected to increase from the outside towards the centre. I admit 



* Halm, loc. cit. 



s 2 



