JEther and Electrons. 705 



"Without some rotational motion, there could be nothing in- 

 volving circulation around any closed circuit ; the region 

 filled by the medium being simply-connected. Thus there 

 can be no (infinitesimal) displacement which involves circu- 

 lations, and in regard to which the system is labile. 



If now we introduce a locked distribution of infinitesimal 

 strain in accordance with the specification quoted in § 14, 

 the elastic properties of the system remain unaffected, in 

 virtue of the result obtained in § 11, so that still no motion 

 involving circulation in any closed circuit can take place 

 without calling up elastic resistance. On these grounds the 

 aetherial circulation which is specified by Professor Larmor as 

 equivalent to the translation of his electron through the aether 

 does not appear to me to be admissible as a mode of motion 

 unopposed by the elasticity of the medium. 



16. It is readily admitted that the possibility of a " free " 

 strain- figure has yet to be demonstrated; but none the less 

 it may be fairly contended that if mobile strain-figures must 

 necessarily be of the free type, the development of theory on 

 the basis of this result is deserving of some attention. 



17. Continuing the comparison of free with locked strain- 

 figures, we may now consider how- they stand in relation to 

 the balance of positive and negative electricity in the known 

 universe. Though any large excess of positive or negative 

 electrification would tend to spread itself through space, and 

 not to become concentrated by gravitation like electrically 

 neutral matter, yet it is hard to resist the conclusion that the 

 ratio of total positive to total negative charges,, taken through- 

 out the universe, is extraordinarily near to unity. If our 

 electrons are so constituted that they can only appear in 

 pairs, a positive and an equal negative being always produced 

 simultaneously* in the ideal process of genesis, this equiva- 

 lence of total charges receives a ready explanation. On the 

 other hand, an electron of the free strain-figure type is 

 supposed to exist independently of any complementary com- 

 panion, and we are led to postulate that the formation of 

 positive electrons having a given total charge, and the for- 

 mation of negative electrons having a complementary total 

 charge, are a priori equally probable. Then a very exact 

 balancing of charges, where enormously great numbers of 

 electrons are concerned, becomes an overwhelming probability. 

 We have no right, however, on the score of symmetry, to 

 make the assumption just referred to ; since experiment 

 shows that the circumstances are very far from symmetrical. 

 It is nothing more than a conjecture that a fuller knowledge 



* Cf. Larmor, ' ^Ether and Matter,' p. 326. 

 Phil Mag. S. (5. Vol. 13, No. 78. June 1907. 3 B 



