70 MORRIS LOEB 



ities and then he assumes that e> the electrostatic charge, 

 is constant; therefore m, the mass, varies. Now, the value 

 of e is derived from Faraday's law, which would never have 

 been announced if Faraday had not dealt with the equivalent 

 weights as fixed mathematical quantities. In fact, just so 

 far as Thomson substantializes electricity by giving it atomic 

 structure, with invariable mass, the chemical atom becomes 

 wavy and matter evanesces into the ghost-like form which 

 energy has assumed in the chemical mind. If our scientific 

 terms are, as it were, to receive the reciprocals of their present 

 significance progress may ultimately result, but we should 

 enter into topsy-turvy dom with our eyes open. 



The electron theory possesses the merit of furnishing a 

 working hypothesis upon which to coordinate the various 

 electrical phenomena of vacuum tube and radio-active origin : 

 chief among which is the increased conductivity of gases. 

 Either direct current measurement or the more sensitive 

 electrometer, determinative of the decrease of electrostatic 

 potential, indicates that gases begin to conduct electricity 

 when affected by ultra-violet light, by cathode and X-rays, 

 by radium, thorium, etc. Ingenious experiments have proved 

 that portions of the gas are positively, others negatively, 

 charged; that they behave as if ionized; the numbers, masses 

 and charges of the hypothetical ions have been measured and 

 found to agree with the assumption that the negative ions 

 have the magnitude of the electrons, the positive ions that of 

 the regular molecules, i.e., the negative ions are always very 

 small and mobile, with the same value for all gases; the posi- 

 tive ions are, at least, 1000 times as large, and vary for dif- 

 ferent gases. If the gas moves away from the locality of ion- 

 izing influence, its conductivity disappears gradually at a rate 

 to suggest reunion of the ions. Plausible, if not quite conclu- 

 sive, reasoning connects the ionization hypothesis with the 



