RECENT ADVANCES IN SCIENCE 283 



that if the age of the Earth is admitted to be large, then it 

 must be assumed that some catastrophic event must have 

 occurred in the sidereal system at no very distant epoch. 



Stellar Dynamics. — The well-known phenomenon of star- 

 streaming indicates that the stars, instead of moving in all 

 directions at random, have a predominant tendency to move 

 in two opposite directions. Prof. Turner suggested that these 

 preferential directions are directed along a line passing through 

 the centre of the stellar system, and that the stars are, in 

 reality, describing elongated orbits about the centre, the two 

 streams being the aggregate of the stars moving respectively 

 towards and away from the centre. The objection to this 

 explanation was that it apparently necessitated a great con- 

 centration of stars near the centre — greater than appeared 

 admissible. Prof. Eddington (M.N. R.A.S. vol. 75, p. 366, 

 191 5, March) has discussed the dynamics of a globular stellar 

 system which, for the sake of simplicity, was supposed to 

 possess spherical symmetry. Starting from Schwarzschild's 

 ellipsoidal law of velocities, which is in good agreement with 

 observation, he has succeeded in deducing a density law which 

 does not give too great a concentration at the centre, and also 

 gives a density which diminishes continuously from the centre 

 outwards. Although the system discussed is an ideal one, this 

 article is a valuable contribution towards our knowledge of 

 stellar dynamics, and it is to be hoped that the author will 

 succeed in extending his results to a system more nearly ap- 

 proaching our sidereal system in its construction. 



The Nature of Cepheid Variation. — The term Cepheid is 

 given to a class of short-period variables in which the rise 

 from minimum to maximum brightness occupies less than half 

 the period of variation. The brightness varies continuously, 

 and so the variability cannot be due to eclipses. Their spec- 

 troscopic study has led to the belief that they are, nevertheless, 

 binary systems, whose period of orbital revolution is equal to 

 the period of their light variation. The maximum brightness 

 is found to occur near the point of maximum velocity of 

 approach. No satisfactory theory of the dependence of the 

 brightness on orbital motion has been advanced. R. H. 

 Curtiss (Astroph. Journ. 20, p. 186, and also Pub. Astron. Obs. 

 Univ. Mich. vol. i. 191 5, p. 104) has attempted an explanation 

 by supposing the existence of a resisting medium around the 



