258 . SCIENCE AND METHOD. 



nearest stars. The density is accordingly lo^* times 

 as small ; now the velocities are upon the same scale, 

 and therefore the radius must be lO^ as great, or i,ooo 

 times the distance of the nearest stars, which would 

 give about a thousand million stars in the Milky Way. 

 But you will tell me that these hypotheses are very 

 far removed from reality. Firstly, the Milky Way is 

 not spherical (we shall soon return to this point) ; and 

 secondly, the kinetic theory of gases is not compatible 

 with the hypothesis of a homogeneous sphere. But if 

 we made an exact calculation in conformity with this 

 theory, though we should no doubt obtain a different 

 result, it would still be of the same order of magni- 

 tude : now in such a problem the data are so uncertain 

 that the order of magnitude is the only end we can 



aim at. 



And here a first observation suggests itself Lord 

 Kelvin's result, which I have just obtained again by 

 an approximate calculation, is in marked accordance 

 with the estimates that observers have succeeded in 

 making with their telescopes, so that we must conclude 

 that we are on the point of piercing the Milky Way. 

 But this enables us to solve another question. There 

 are the stars we see because they shine, but might 

 there not be dark stars travelling in the interstellar 

 spaces, whose existence might long remain unknown? 

 But in that case, what Lord Kelvin's method gives us 

 would be the total number of stars, including the dark 

 stars, and as his figure compares with that given by 

 the telescope, there is not any dark matter, or at least 

 not as much dark as there is brilliant matter. 



Before going further we must consider the problem 

 under another aspect. Is the Milky Way, thus con- 



