I. 



THE CONSTITUTION OF NATURE. 



WE cannot think of space as finite, for wherever in 

 imagination we erect a boundary we are compelled to think 

 of space as existing beyond that boundary. Thus by the 

 incessant dissolution of limits we arrive at a more or less 

 adequate idea of the infinity of space. But though com 

 pelled to think of space as unbounded, there is no mental 

 necessity to compel us to think of it either as filled or as 

 empty ; whether it is filled or empty must be decided by 

 experiment and observation. That it is not entirely void, 

 the starry heavens declare ; but the question still remains, 

 Are the stars themselves hung in vacua? Are the vast 

 regions which surround them, and across which their light 

 is propagated, absolutely empty ? A century ago the 

 answer to this question would be, &quot; No, for particles of 

 light are incessantly shot through space.&quot; The reply of 

 modern science is also negative, but on a somewhat differ 

 ent ground. It has the best possible reasons for rejecting 

 the idea of luminiferous particles ; but, in support of the 

 conclusion that the celestial spaces are occupied by matter, 

 it is able to offer proofs almost as cogent as those which 

 can be adduced for the existence of an atmosphere round 

 the earth. Men s minds, indeed, rose to a conception of 

 the celestial and universal atmosphere through the study 

 of the terrestrial and local one. From the phenomena of 

 sound as displayed in the air, they ascended to the phe- 



