THE SCIENTIFIC METHOD 139 



by assigning a proper variation either to the numerator 

 or the denominator. It therefore becomes a mere mat- 

 ter of expedience which of the two quantities, mass or 

 electrical change, shall be supposed constant. In addi- 

 tion, the quantity, e/m, is itself a constant for all 

 velocities which can be attained by bodies which are 

 appreciably large. So the whole question of variation 

 is more or less academic, in that it does not become 

 important unless we are discussing hypothetical atomic 

 systems. 



To say that e is a constant is an assumption based 

 solely on an analogy to the experimental laws of elec- 

 trolysis; but in electrolysis, when we obtain equal 

 electrical charges we also find equivalent masses of 

 matter. In the discharge of electricity through gases 

 and in radio-activity the matter deposited is too small 

 to be measured. This is a fundamental difference and 

 vitiates an analogy between the two. For example, we 

 measure the amount of current in a vacuum tube by an 

 electrical device, and at the same time we measure the 

 deflection of the current by an electric and magnetic 

 field; in other words, all quantities and forces are 

 electrical, and we say that equal currents in this case 

 require equivalent quantities of matter. But it has 

 not been shown to be impossible or even improbable 

 that electrons, associated with equal quantities of mat- 

 ter but having different velocities, might show different 

 electrical charges; or that electrons producing equal 



