552 SCIENCE PROGRESS 



nebula at a distance of 20 million light-years, and with an 

 apparent diameter of 10', would actually have a diameter of 

 60,000 light-years, which is of the order of magnitude of our 

 own system. 



Distances of Stellar Clusters. — As in the case of spiral nebulae, 

 our knowledge of the distances of star clusters has to be obtained 

 indirectly ; far more confidence can, however, be placed in 

 the indirect estimates than in any direct measurements, where 

 such distant objects are concerned. H. Shapley (Proc. Nat. 

 Acad. Set. 3, 479, 191 7) has made some interesting estimates. 

 He shows that the extremes of range of variation of the Cepheid 

 type of variable stars in various clusters are about the same, 

 and further that in the same cluster the magnitudes of all 

 Cepheid type variables closely agree. It is believed that the 

 intrinsic magnitudes of all Cepheid stars are about the same 

 wherever the stars occur, and a means is thus provided of 

 estimating the distance of the clusters in which they occur. 

 Shapley thus finds parallaxes for several clusters which average 

 about o"*oooi, corresponding to a distance of about 30,000 

 light-years. He further finds that the difference between 

 the mean magnitude of the 25 brightest stars in a cluster 

 and that of the Cepheid variables in it is about i m .35 for all 

 clusters examined ; this result can be used to estimate the 

 mean magnitude of the Cepheids in clusters in which none have 

 been observed and hence to obtain an estimate of the distances 

 of such clusters by the preceding. 



The Threshold of Vision. — Three determinations of the 

 minimum light perceptible to the human eye have been made 

 recently, by H. E. Ives, Astroph. Journ. 44, 124, 19 10 ; 

 H.N. Russell, ibid. 45, 60, 191 7 ; and Prentice Reeves, ibid. 

 46, 167, 191 7, respectively. The first two of these determina- 

 tions were based upon the magnitude of the faintest visible 

 stellar object and on the pupillary area of the eye when com- 

 pletely adapted to darkness. The most direct and accurate 

 determination of the latter quantity is by the method used 

 recently by W. H. Steavenson, in which flashlight photographs 

 are taken of the eye when accommodated to darkness. He 

 thus obtained a value for the diameter of the pupil oi about 

 8" 5 mm. — considerably larger than previous estimates. This 

 result was not available to Ives, who used too small a value 

 for the diameter ; he further assumed that the faintest visible 



