5, G] Survey of the Problem 9 



The distribution of globular clusters in the sky is somewhat surprising, it 

 being found* that they are practically confined to one hemisphere of the 

 sky. Not only this but there is a very marked condensation about one 

 point in the sky, 42 out of the 82 considered by Melotte lying within 30 of 

 a point in the galactic plane of latitude 325. 



Slipher has recently measured the radial velocities of ten star clusters, 

 and finds velocities ranging from 410 to -f 225 kms. a second, the mean of 

 the values, taken without regard to sign being 150 kms. a second. It is 

 clear that we have here to deal with velocities of the same order of magnitude 

 as the velocities of the spiral nebulae. 



Finally Shapleyf has attempted to estimate the distances of various 

 globular star-clusters, by assuming the absolute magnitudes of the cepheid 

 variables contained in them to be equal to those of similar cepheid variables 

 at known distances. He finds that probably, with one or two exceptions, 

 no globular cluster is nearer than about 30,000 light-years, corresponding to 

 a parallax of 0'00012 /X J. Thus the globular clusters, like the spiral nebulae, 

 appear to be independent of, and outside, our own system of stars. 



The formation of moving star clusters also exhibits a certain, although 

 not very great, degree of uniformity . A number of stars is said to form a 

 moving cluster when their velocities are sensibly the same, both in magnitude 

 and direction, and also when there is definite evidence of some further real 

 connection between the members of the cluster. The latter condition is 

 important because, by a procedure which is familiar to every student of the 

 Kinetic Theory of Gases, any collection of chaotically moving stars can be 

 resolved into parallel showers. ' Observational astronomy reveals the existence 

 of clusters of stars moving with equal velocities and also having physical 

 characteristics in common which suggest that they have some bond of common 

 origin. The cluster formed by the Pleiades provides perhaps the most super- 

 ficially obvious instance of a star cluster of this kind. Here we have a 

 group of stars, all of similar spectral type, all of approximately equal bright- 

 ness, concentrated in one region of space and moving with a common velocity ||. 

 A more thoroughly investigated cluster is the Taurus cluster which consists 

 of the Hyades and other neighbouring stars IF. A noteworthy cluster of 

 special interest is the Ursa Major cluster, which contains among other stars, 

 the stars ft, 7, 8, e and f Ursae Majoris of the " Plough."** There is a very 



* Cf. Melotte, Mem. E.A.S. 70, p. 176, and A. K. Hinks, Monthly Notices E.A.S. 71, p. 693. 

 t Proc. Nat. Acad. ScL 3 (1917), p. 479. 



* By a similar method Hertzsprung had previously estimated the distance of the lesser Magel- 

 lauic cloud to be of the same order (parallax -0001). Cf. Ast. Nach. 4692. 



Jj On this subject in general, see Chap. IV of Eddington's Stellar Movements. 

 || W. W. Campbell, Stellar Motions, p. 181. 

 IT Boss, Astrophya. Journ. 26 (1908), p. 31. 

 ** Ludendorff, Ast. Nnch. 180 (1909), 265, and W. W. Campbell, Stellar Motions, p. 175. 



