232 SCIENCE PROGRESS 



mass is almost equal to that of the whole atom, the mass of 

 the electrons being comparatively of little account. One 

 element is distinguished from another merely by the amount 

 of the charge on the nucleus — and, as a consequence, by the 

 number of electrons and the mass of the atom, since the mass 

 is supposed to be purely electro-magnetic. Hydrogen would 

 have unit positive charge and one electron ; helium, two of 

 each, and so on, an increase of one unit marking an advance 

 from one element to the next in the periodic table. Now 

 according to classical mechanics, electrons revolving round a 

 centre should radiate energy, and this is quite in accordance 

 with what atoms are believed to do. But the great difficulty 

 to be overcome is that, as a result, the orbits should gradually 

 become smaller and smaller until at last the electrons fall into 

 the nucleus and the atom ceases to exist. Moreover, as the 

 orbits shrink, the electrons should revolve faster and faster, 

 so that if we suppose — as seems natural — that the frequency 

 of the waves emitted is equal to that of revolution of the 

 electrons, we cannot understand the sharpness of the spectrum 

 lines, which tells of a very high degree of constancy in the 

 frequency. So long as we hold the laws of Newtonian mechanics 

 to be inviolable, there seems to be no possibility of escape from 

 this difficulty. We have, to some extent, therefore, to abandon 

 those laws. The departure is made by two distinct supposi- 

 tions. In the first place, we suppose that the electrons are 

 compelled to travel in certain defined orbits, separated from 

 one another by finite distances, and that in no circumstances 

 can they possibly revolve in the spaces between. In the second 

 place, we assume that, while an electron is revolving in any 

 one of those orbits, it radiates no energy, but that in other 

 respects the motion is in accordance with classical laws. We 

 can give no explanation of these requirements. The spectrum, 

 if we have understood it aright, issues an imperious demand 

 that they shall be granted, and condescends to no further 

 parley. Starting, then, from these premises, we find that 

 radiation takes place when an electron changes from one of 

 the possible orbits to another. The energy of the electron in 

 the stable state is a function of the radius of its orbit. Conse- 

 quently, when the orbit is changed, say to one of smaller 

 radius, there is a balance of energy left over. It is supposed 

 that this energy is radiated as a wave whose frequency is 

 proportional to its measure. Here again, if we attempt to 

 picture the details of the process by which energy of revolution 

 becomes transformed into wave energy, we are quite at a loss. 

 Energy of revolution vanishes, energy of wave motion appears, 

 and we simply equate them, making an arbitrary assumption 

 about the frequency. 



