196 CORRELATION OF PHYSICAL FORCES. 



mass of matter come from nowhere, and formed from noth- 

 ing. The impossibility, humanly speaking, of creating or 

 annihilating matter, has long been admitted, though, perhaps, 

 its distinct reception in philosophy may be set down to the 

 overthrow of the doctrine of Phlogiston, and the reformation 

 of chemistry at the time of Lavoisier. The reasons for the 

 admission of a similar doctrine as to force appear to be equally 

 strong. With regard to matter, there are many cases in 

 which we never practically prove its cessation of existence, 

 yet we do not the less believe in it : who, for instance, can 

 trace, so as to re- weigh, the - particles of iron worn off 

 the tire of a carriage wheel ? who can re-combine the parti- 

 cles of wax dissipated and chemically changed in the burning 

 of a candle? By placing matter undergoing physical or 

 chemical changes under special limiting circumstances, we 

 may, indeed, acquire evidence of its continued existence, 

 weight for weight and so we may in some instances of force, 

 as in definite electrolysis : indeed the evidence we acquire of the 

 continued existence of matter is by the continued exertion of 

 the force it exercises, as, when we weigh it, our evidence is 

 the force of attraction ; so, again, our evidence of force is 

 the matter it acts upon. Thus, matter and force are corre- 

 lates, in the strictest sense of the word ; the conception of 

 the existence of the one involves the conception of the exis- 

 tence of the other : the quantity of matter again, and the de- 

 gree of force, involve conceptions of space and time. But 

 to follow out these abstract relations would lead me too far 

 into the alluring paths of metaphysical speculation. 



That the theoretical portions of this essay are open to ob- 

 jection I am fully conscious. I cannot, however, but think 

 that the fair way to test a theory is to compare it with other 

 theories, and to see whether upon the whole the balance of 

 probability is in its favour. Were a theory open to no ob- 

 jection it would cease to be a theory, and become a law ; 

 and were we not to theorise, or to take generalised views of 



