414 



KNOWLEDGE. 



November, 1911. 



ruin' enhance their mutual heautw and in another 

 the topaz and the sapphire nutglow each other. 

 Imaj^ine oneself the inhabitant of a planet lit by a 

 ruin' sun and an emerald sun. Consider the brilliant 

 pictures and the shadows, every shadow dt)Lible, one 

 coloured red and the other green. These wonders 

 seem too much for full description. So I will not tr\' 

 to describe the effects as seen li\ the iiih;ibitaiits o( 

 planets uitlnii a coloured 

 star cluster: I leave 

 to the readers 

 " Knowledge " to 



a 



'ine the scener\' 



It 



of 



im- 



for 



themselves. 



Collisions in 

 Clusteks. 



The great Humboldt 

 asked this most sug- 

 gestive question : " How 

 can these systems be 

 maintained — how cati 

 the suns, crowciing at 

 the interior of star 

 clusters, fulfil their re\'o- 

 lutions freely and with- 

 out clashing." Do they 

 re\'ol\'e without clashing? 

 The answer is: Almost 

 certainlv the\' occasion- 

 ally clash. 



N o v a e s o ni e t i m e s 

 occur in star clusters 

 as do main variables. 

 We Ikinc alread\' proxed 

 that the impact of suns 

 absolutely must produce 

 new stars, and Nature 

 seems to offer no other 

 suggestion of then- 

 origin. E\'ery such im- 

 pact must strew the 

 space about the point 

 of collision w ith meteoric 

 and gaseous matter, 

 which must exercise 



A tvpic-il star clustKr with a condensed centre ; a more 



difliised one, tliat (jf w Centauri, was shown on page J4J ol 



the September issue. 



a retarding e 



ffect 



upon 



it, and produce other 

 stars over the cluster. 



the suns passing through 



imi)acts strewing variable 



l'2ver\ such collision lessens the nuitixf energy 



of the (luster, and helps to weld it together, 



until hiialh', after eons have passed, a \'ast nebulous 



sun m.'i\' result. 



Till-; Genesis oi- Clusters. 



How do star clusters originate ? Possibly they 

 are the nuclei of vast third l>odies th.it have been 

 striK:k from gigantic suns, or perhaps the result 

 of whirling coalescence, where great angular motion 

 prevented condensation to a centre. We ha\'e 

 already seen that selective molecular escape must 

 rob the bodies at unstable temperatures of their 



light gases, whilst a vast expanding rotating mass of 

 elements of greater atomic weight will be left behind, 

 that is a rotating meteoric swarm. Such a rotating 

 mass will not (|uickl\' tend to concentrate to a centre, 

 the angular velocit\' will keep the particles in orbits. 

 These orbits must be continually crossing one 

 another, and although the general rotation of the 

 mass will be m om- direction, as the swarm will be 



rougliK' spherical, the 

 indivitlual particks w ill 

 be in orl)its in all 

 azimuths. Hence, in- 

 stead of concentrating 

 to a [ilane as they wnuld 

 [probably do were they 

 subject to ;i great 

 central controlling force, 

 the\' will tend to aggre- 

 gate into larger and 

 larger fragments, the 

 whole retaining some- 

 what of a globular form. 

 Small swarms of this 

 character would prob- 

 abh' give rise to some 

 of the comets, but it 

 would not be the only 

 method of their forma- 

 tion, as impact shows 

 us, man\' modes by 

 which dust swarms may 

 originate. \'er\' gigantic 

 swarms would, as the 

 ages rolled on by con- 

 tinued coalescence gradu- 

 allv form star clusters 

 in which the constituent 

 suns would be somewhat 

 similar in mass. 



Impacts within star 

 clusters would not be so 

 liable to produce double 

 stars as under ordinary 

 circumstances, because in 

 proportion to the mass 

 there must be a very con- 

 siderable orbittil proper 

 motion. Hence, although we should expect a fair 

 number of very transitory novae and a great number 

 of xariables in special parts of star clusters, or in 

 special clusters, we should not expect a great 

 number of dt)nble stars. The whole subject of the 

 internal motions of star clusters is one worthy of 

 ver\' careful stud\' on the [:art of the astronomer. We 

 shall require to know much more than we do about 

 these motions, before there can be any certainty as 

 to the origin of clusters. 



The suggestion here niadc as to their genesis has 

 considerable elements of probabilit}', but is certainly 

 not an absolutely demonstrated deduction, as is the 

 partial impact origin of new stars bv the formation 

 of the tliird bod\'. Nor does it compare in 



Cluster M.i 



e'anes 'Venatici. 



