26 SCIENCE PROGRESS 



embrace all the known results of such phenomena as chemical 

 combination and valency, molecular cohesion, radiation, dia- 

 magnetism and paramagnetism, ionisation, Zeeman effect, 

 series laws of spectral lines. The earlier view of Thomson as to 

 the atom consisting of a " sphere of positive electricity," with 

 coplanar rings of electrons embedded in it, and in motion 

 around the centre, has gradually given way to the " nucleus ' 

 theory of Rutherford, in which the positive charge which 

 serves to bind the electrons in their orbits is no longer diffused 

 through the whole volume " occupied " by the atom, but is 

 concentrated within a very small portion of its volume at the 

 centre, having in fact a size which must be smaller than that 

 of the electron itself if we are to account for the whole mass 

 of the atom on purely electromagnetic grounds. This view, 

 in one form or another, holds the field to-day, and, despite its 

 manifest difficulties, it has had a remarkable success. For 

 instance, it explains very readily the scattering oi" a rays 

 which is experienced when these particles are passing through 

 gases or thin foils of metal, a fact with which it is difficult to 

 reconcile the Thomson theory of atom structure. The most 

 serious obstacle to its acceptance is the instability of the 

 electrons in their orbits around a minute centre of force at- 

 tracting as the inverse square law. For one thing, any sort of 

 stability in the orbits is impossible at all unless the orbits are 

 coplanar, and it is difficult to see how even these would sur- 

 vive some of the shocks to which every atom must be exposed 

 even in its ordinary thermal encounters with other atoms. 

 Added to this is another difficulty. All electrons in orbital 

 motion are being accelerated, and such acceleration is accom- 

 panied by radiation of the electron's energy leading gradually 

 to contraction of the orbit and ultimate coalescence with the 

 atomic nucleus — a state of affairs quite irreconcilable with 

 the constant properties of atoms as we meet them in physical 

 and chemical experiments. Yet orbital motion is a postulate 

 we cannot abandon without at the same time encountering 

 just as apparently insurmountable difficulties in explaining 

 the magnetic properties of matter, the Zeeman effect, etc. Of 

 course, all these -difficulties arise in the application of the 

 classical Newtonian, dynamical principles to atomic models as 

 pictured above. The Newtonian dynamics has had great 

 success among those phenomena where the velocities of moving 



