290 THE GRAMMAR OF SCIENCE 



Possibly it is idle to inquire so long, at any rate, as the 

 conceptual ether remains as little defined as at present. 

 Our notions of the ether are so essentially bound up with 

 the conception of its continuity, while our notions of gross 

 " matter " are, on the other hand, so closely associated 

 with the idea of the discontinuity of matter, that we 

 are inclined to treat as fundamental for ether-elements 

 the method in which they act in each other's presence, 

 and for gross " matter " corpuscles the method in which 

 they act when isolated. On this account the law of 

 inertia, as we postulate it for gross " matter " corpuscles, 

 may be considered as a feature of mechanism very prob- 

 ably flowing from the structure of the prime-atom itself 



^ 5. — TJie Third Law of Motion. Mutual Acceleration is 

 determined by Relative Position 



Let us now proceed a stage further and postulate the 

 next simplest field ; let us suppose two corpuscles taken 

 and their motions determined relatively (p. 208) to a 

 frame through a third corpuscle, which, however, like that 

 on p. 287, we will consider to be at such a distance as to 

 be quite isolated from their influence. What must we 

 conceive as happening ? In the first place, because two 

 corpuscles are in the same field must we consider them as 

 having a certain definite position relative to each other ? 

 Certainly not. We find ourselves compelled to consider 

 them as capable of taking up a great variety of positions 

 with regard to each other. Does, then, the fact that they 

 are in the same field, or in a certain relative position in 

 that field, determine with what velocities we are to 

 consider them as moving ? Again we must answer : No 

 — at any rate for particles. In order to construct motions 

 which will effectively describe our sequences of sense- 

 impression we are forced to suppose that particles may 

 move through the same relative position with every 

 variety of velocity. What, then, must we consider as 

 determined when we know the relative position of two 

 corpuscles ? It is their accelerations, the rates at which 



