﻿92 , Pr. W. F. G. S.wahft.on*7<« 



experiment which ever had or ever would take place being 

 completely determined. It would be hopeless, and in fact 

 needless, tor us to inquire how the system originated, or why 

 it was that the present system was started rather than some 

 other, equally consistent with the electromagnetic scheme, 

 but showing types of motion which can never naturally evolve 

 out of the present universe. Since we cannot ascertain the 

 complete field of the universe at some instant, we are obliged 

 to fall back on more crude observation, and formulate certain 

 subsidiary laws which are broad enough to restrict the types 

 of field and of electronic motions which are to be considered 

 as possible, to those which it is possible for the electromagnetic 

 scheme to evolve out of the actually existing universe, that is 

 to say, to those fields and motions which we actually find in 

 practice*. These laws are of course to be of such a nature as 

 never to clash with the electromagnetic scheme, which, as we 

 have seen, is sufficient in itself to tell us the subsequent history 

 of a system when once that system is completely specified for 

 some instant. These subsidiary laws are in fact hardly 

 entitled to be called fundamental laws at all, being merely 

 substitutes for what we lack in being without complete know- 

 ledge of the field of the universe at some instant. The great 

 practical difficulty consists in formulating the true nature of 

 these subsidiary laws, which usually resolve themselves into 

 sets of more or less approximate truths, sufficiently accurate 

 for the particular purpose in hand. To illustrate this point 

 suppose we were to attempt to say what would happen if two 

 similarly charged spheres were placed near to each other and 

 left to themselves. Without further data it is impossible to 

 give any answer to the question . The electromagnetic scheme 

 would be satisfied if we were to say that the spheres remained 

 at rest, and gave as the field the sum of the fields due to each 

 sphere when alone. To proceed any farther in onr problem 

 it is necessary to partially make up for our deficiency of 

 knowledge of the electronic motions in the sphere by the 

 application of some tangential surface condition. Still we 

 shall be able to find an infinite number of solutions to fit the 

 case, and it becomes necessary to introduce the idea of a 

 Newtonian mass |, together with the accompanying notions 

 of forces and ordinary dynamical laws, and even then our 

 solution will be indeterminate without specification of the 



* On this view, gravitation, for instance, occupies the position of one 

 of the "subsidiary laws,'.' being 1 a land of empirical expression of. a 

 property of matter really inherent in the nature of the electronic fields^ 

 and nmtions originally started. 



t The introduction of the idea of Newtonian mass is only necessitated 

 by the deficiency of our knowledge of the electronic distributions and 

 motions in the molecules of the sphere. 



