674 



NATURE 



[February i8, 19 15 



STAR CLUSTERS. 



OF all the telescopic objects in the sky none 

 are more beautiful or more fascinating' than 

 the condensed, globular star clusters. Their be- 

 wildering" complexity renders them unsuitable for 

 direct study at the telescope, but photography has 

 now brought them within the range of systematic 

 investigation. The technical problem which they 

 present is by no means easy, and demands high 

 resolving power for success. The fine examples 

 reproduced herewith, M3 and Mi 3, have been 

 very kindly sent from the Mount \^'ilson Solar 

 Observatory, and illustrate admirably the work 

 of the famous 5 ft. mirror constructed 

 by Mr. G. W. Ritchey. 



Considerable attention was given to 

 the star clusters by Sir John Herschel, 

 whose attempts to depict them by hand 

 met naturally with small success. Cer- 

 tain curious irregularities which he be- 

 lieved to exist in the distribution of the 

 stars may be attributed to a purely sub- 

 jective origin, or they may be accounted 

 for by the absorptive influence of ex- 

 ternal dark nebulous masses. No great 

 importance is now^ attached to them, 

 and in the main the stars may be con- 

 sidered as distributed with radial sym- 

 metry. But one curious feature noticed 

 by Sir John Herschel has been confirmed 

 by later study. The stars in a cluster 

 tend to divide into two classes of magni- 

 tude, a brighter and a fainter, separated 

 by a distinct interval. Can this be a 

 visible division of stars presumably at 

 the same distance and of nearly equal 

 age into the two classes of giant and 

 dwarf stars inferred by Hertzsprung and 

 H. N. Russell? 



About twenty years ago Prof. S. I. 

 Bailey, at that time at Arequipa, devoted 

 considerable study to photographs of the 

 chief globular clusters. His work pro- 

 ceeded on two lines. On one hand 

 he made systematic counts of the stars 

 recorded, thus laying the foundation for 

 statistical investigations of their ar- 

 rangement in space. And on the other 

 he investigated the magnitudes of the 

 stars, and was thus led to the remark- 

 able discovery that several clusters 

 contain a high proportion of variable stars, 

 a ratio of i in 7 in the extreme case of M3. 

 His detailed results for the clusters to Cen- 

 tauri and M3 have been published In two 

 beautiful memoirs. The type of variation is of a 

 distinct character, though a few isolated examples 

 have been found elsewhere In the sky, with a 

 period of about twelve hours and a rapid rise to 

 maximum. In the case of M3 the variation Is sin- 

 gularly true to one type, the range between maxi- 

 mum and minimum being two photographic mag- 

 nitudes. Some clusters, notably Mi 3, are almost 

 entirely devoid of such variables ; where they do 

 NO. 2364, VOL. 94] 



occur they are apparently confined to the stars ot 

 the brighter order of magnitude. 



The question of the distribution of stars in 

 clusters was discussed by Prof. E. C. Pickering. 

 Using counts on the clusters u Centauri, 47 

 Tucanae and M13 (Herculisj, he formed the im- 

 portant conclusions: (i) that the law of distribu- 

 tion is essentially the same for different clusters, 

 (2) that the bright stars and the faint stars of a 

 cluster obey the same law. He represented 

 graphically the curve of apparent (projected) 

 density for different distances from the centre, and 

 attempted without success to reproduce It by 

 assuming laws of the form i—r- and (i— r)" foi 



Fig. I. — M3 Canes Venatici. Exposure 4h. 



the density in space. The latter form was als 

 tested by Mr. W. E. Plummer with much tl 

 same result on an extensive series of measures of 

 the stars in Mi 3. 



The next important contribution to the subject 

 is due to H. v. Zeipel, who measured the positions 

 of the stars in M3 (Can. Ven.). By adapting the 

 solution of a certain integral equation studied by 

 Abel he showed how the law of distribution in 

 space may be deduced numerically from the ob- 

 served distribution as it is seen in projection. 

 Later he compared the law of density in space 

 arrived at in this way with that which obtains ii^ 



