628 REPORT — 1900. 



as valid unlesa it explicitly restricted the argument to permanent systems ; yet 

 the conditions of permanency are unknown except in the simpler cases. The only 

 mode of discussion that is yet possible is the method of dynamical statistics of mole- 

 cules introduced by Maxwell. Now statistics is a method of arrangement rather 

 than of demonstration. Every statistical argument requires to be verified by com- 

 parison with the I'acts, because it is of the essence of this method to take things as 

 fortuitously distributed except in so far as we know the contrary ; and we simply 

 may not know essential facts to the contrary. For example, if the interaction of 

 the fether or other cause produces no influence to the contrary, the presumption 

 would be that the kinetic energy acquired by a molecule is, on the average, equally 

 distributed among its various independent modes of motion, whether vibrational 

 or translational. Assuming this type of distribution to be once established in a 

 gaseous system, the dynamics of Boltzmann and Maxwell show that it must be 

 permanent. But its assumption in the first instance is a result rather of the 

 absence than of the presence of knowledge of the circumstances, and can be 

 accepted only so far as it agrees with the facts ; our knowledge of the facts of 

 specific heat shows that it must be restricted to modes of motion that are homo- 

 logous. In the words of Maxwell, when he first discovered in 18G0, to his great 

 surprise, that in a system of colliding rigid atoms the energy would always be 

 equally divided between translatory and rotatory motions, it is only necessary to 

 assume, in order to evade this unwelcome conclusion, that • something essential to 

 the complete statement of the physical theory of molecular encounters must have 

 hitherto escaped us.' 



Our survey thus tends to the result, thnt as regards the simple and uniform 

 phenomena which involve activity of finite regions of the universal fother, 

 theoretical physics can lay claim to constructive functions, and can build up a 

 definite scheme ; but in the domain of matter the most that it can do is to accept 

 the existence of such permanent molecular systems as present themselves to our 

 notice, and tit together an outline plan of the more general and universal features 

 in their activity. Our well-founded belief in the rationality of natural processes 

 asserts the possibility of this, while admitting that the intimate details of atomic 

 constitution are beyond our scrutiny and provide plenty of room for processes that 

 transcend finite dynamical correlation. 



The following Papers were read : — 

 1. Note on M. Cremieu's Experiment. By Prof. G. F. FitzGerald, F.R.S. 



M. Cremieu has shown that, if his experimental methods can bear criticism as 

 •well as they seem to do, there is no induced electromotive force on a coil of wire 

 Burroundins a rotating disc when the strength of an electric charge on the disc is 

 chano-ing. He has deduced from this the conclusion that there is no magnetic 

 induction through the disc due to the moving charge such as Rowland's experi- 

 ments showed. This note is to point out that too little is known of the theory of 

 the ethereal effects of a charge of electricity forced to move by mechanical actions 

 for us to be quite sure that both M. Cremieu's and Rowland's observations may 

 not be true— Je., that it is possible that a charge of electricity, while it is being 

 accelerated by moving matter, may produce such an action on the surrounding 

 ether as to neutralise the electric force that would otherwise be produced by the 

 changing magnetic induction due to the moving charge. 



• 2. On the Creeping of Liquids and the Surface Tension of Mixtures. 

 hJDv. F. T. Tkoutox, F.E.S. 



3. On a Method of Investigating Correspondences between S2^eclra. 



By Hugh Ramage. 



The method is graphical ; spectral lines are plotted as abscissre, and the atomic 

 weights of the elements, or functions of the atomic weights, as ordinates. Con- 



