8 SCIENCE PROGRESS 



various types. Many investigations have been made to deter- 

 mine this, and they all agree in their main results. Briefly 

 stated, these are that the most distant stars are those of very early 

 type, and that the average distance gradually decreases with 

 advancing type, until the solar type is reached. The stars of 

 this type are on the average the nearest to us. Those of still 

 later types are more distant, and the red or M stars are the most 

 distant of all, after the B stars. The great average distance of 

 the helium or early-type stars is accounted for by the fact that 

 they are mostly in the distant galactic star clouds. We may 

 thus picture our sidereal system as consisting of a central 

 region containing stars of various types, but chiefly the later 

 ones, more or less uniformly distributed about their centre ; 

 outside this central region lies the belt of the Milky Way at 

 such an immense distance from our solar system that it has 

 not yet been measured with any certainty, constituted mainly 

 of early-type stars, and containing many nebulae and nebulous 

 clouds. An influence of the galactic belt on the inner region 

 is found in the tendency of its stars to concentrate towards the 

 galactic plane, more particularly in the case of the earlier types — 

 although we are not yet able to assign definitely the cause of 

 this, nor the way in which the system has arrived at its present 

 form. 



The further question arises as to whether our sidereal 

 universe is finite or infinite in extent. It is a well-known fact 

 that when the number of stars of magnitude in + i is compared 

 with the number of magnitude m, the former is found to be 

 between three and four times the greater. Now by the defini- 

 tion of the magnitude of a star, a star of magnitude m gives 

 2'5i2 times as much light as a star of ; magnitude m + i (log. 

 2*512 = 0'4, so that a ratio of 100:1 in apparent brightness 

 corresponds to a difference of 5 magnitudes). It follows from 

 these facts that all the stars of magnitude m + 1 give together 

 more light than all the stars of magnitude m. This result can 

 only hold for moderate values of m, because otherwise the 

 heavens would shine with a blaze of light greater than the 

 noonday sun, due to the innumerable faint stars. It follows 

 that at some stage the number of stars of a given magnitude 

 must reach a maximum and afterwards begin to decrease, and 

 this leads one to expect that the total number of the stars may 

 be finite in amount. That this is so is indicated by the counts 



