Theory of Molecular Action according to Newton's Law. 125 



In what way the alternative of attraction or repulsion is de- 

 fined, I do not profess accurately to specify. I prefer, for 

 the present, to consider matters of detail as open for future in- 

 vestigation. That I may be allowed to do so it is necessary 

 that I should premise the grounds on which I consider them 

 as not yet satisfactorily established. Whether the molecules 

 of matter attract or repel each other is perfectly indifferent ; 

 I believe either hypothesis will do very well. Neither does 

 it signify whether the particles of matter attract or repel those 

 of the other fluid (called aether), provided it be allowed that 

 the latter can come in contact with and rest against the former. 

 But whether the particles of aether attract or repel each other 

 is a question of more importance, and one which, when de- 

 cided, will probably settle the other two. The prima facie 

 probability is that they act by repulsion. It is argued in 

 favour of this supposition, that were it not so, the slightest 

 displacement which should bring two particles near each other 

 would of necessity cause them to run together. That this 

 argument is fallacious will appear presently, when we shall 

 show that they would not instantaneously tend either to unite 

 or to separate. Another argument is that when they had 

 once come in contact they could never again be separated. 

 This argument applies with equal force against any hypothesis 

 of attractive particles. At the same time I do not think the 

 arguments in favour of the hypothesis of attraction to be by 

 any means conclusive. 



The popular grounds on which I rested this hypothesis 

 (Trans. Camb. Phil. Soc. vol. vi. p. 178) can, of course, only 

 be held as an illustration. That they are quite insufficient to 

 build anything upon, is obvious enough ; but it is most com- 

 pletely shown by Mr. Earnshaw in his memoir on the Nature 

 of Molecular Forces, to which I am about to direct attention 

 presently. (See Art. 8.) Nor is the argument deduced from 

 an approximate estimation of the value of the function which 

 expresses the time of vibration of a particle, at all conclusive. 

 It will be found in my memoir (Trans. Camb. Phil. Soc. vol. vi. 

 p. 183 and 24-1). It rests on the assumptions, first, that the 

 principal effect is due to the particles in the immediate neigh- 

 bourhood of that whose motion we are ihvestigating ; secondly, 

 that the effect of the action of any particle is independent of 

 its position relative to the direction of transmission. The 

 former assumption is doubtless admissible to a certain extent ; 

 the latter, I believe, not at all. The attractive nature of the 

 particles is still further supported by an argument which I do 

 not now regard as satisfactory. It is this: — We have good 

 reason to suppose that the vibrations of the air are normal, in 



