262 SCIENCE AND METHOD. 



disc. It is clear, then, that a mass starting without 

 velocity from the surface will arrive at the centre with 

 varying velocities, according as it has started from the 

 surface in the neighbourhood of the middle of the disc 

 or from the edge of the disc. In the latter case the 

 velocity will be considerably greater. 



Now up to the present we have assumed that the 

 individual velocities of the stars, the velocities we 

 observe, must be comparable to those that would be 

 attained by such masses. This involves a certain 

 difficulty. I have given above a value for the dimen- 

 sions of the Milky Way, and I deduced it from the 

 observed individual velocities, which are of the same 

 order of magnitude as that of the Earth upon its orbit ; 

 but what is the dimension I have thus measured ? Is 

 it the thickness or the radius of the disc? It is, no 

 doubt, something between the two, but in that case 

 what can be said of the thickness itself, or of the 

 radius of the disc ? Data for making the calculation 

 are wanting, and I content myself with foreshadowing 

 the possibility of basing at least an approximate 

 estimate upon a profound study of the individual 

 motions. 



Now, we find ourselves confronted by two hypo- 

 theses. Either the stars of the Milky Way are 

 animated with velocities which are in the main 

 parallel with the Galactic plane, but otherwise dis- 

 tributed uniformly in all directions parallel with 

 this plane. If so, observation of the individual 

 motions should reveal a preponderance of components 

 parallel with the Milky Way. This remains to be 

 ascertained, for I do not know that any systematic 

 study has been made from this point of view. On the 



