CH. XVIIL] ATOMIC RADIATION 165 



It is easy to grant that, whenever there are actual 

 collisions of sufficient suddenness, some radiation must 

 be emitted ; but we cannot help asking, why does 

 not the quiet orbital revolution of electrons round 

 atoms, in a substance not in a high state of 

 thermal disturbance and not possessing specially 

 massive atoms, why does not this also give rise to 

 a perceptible amount of true radiation and loss of 

 energy ? One answer that has been given is as 

 follows : 



The radiators are not isolated or independent, and 

 surface radiation is maintained by layers at greater 

 depth in the substance. Moreover the radiators are 

 so close together that they are in all sorts of phases 

 within the first quarter wave length, a length which 

 embraces a multitude of them; wherefore a multitude 

 is a worse radiator than one, because they interfere 

 and produce but little external or distant effect ; like 

 the two prongs of a fork, or two neighbouring organ 

 pipes, or the front and back of a vibrating wire. See 

 Larmor, Ether and Matter, page 232 ; the proposition 

 by which he considers the question settled is that the 

 vector sum of accelerations equals zero. 



Of course it must not be forgotten that radiation 

 of a low temperature order is as a matter of fact 

 always going on from all substances ; that energy is 

 conserved, and constancy of temperature persists, 

 merely because loss is equal to gain, because absorption 

 compensates radiation, not because radiation ceases ; 

 and that to make an estimate of the amount of 

 radiation, so occurring, it would be necessary to 

 suppose the body in an enclosure at absolute zero : 

 when undoubtedly its kinetic energy would rapidly 

 leak away, and be dissipated. But this refers to 

 questions connected with ordinary radiation, whereas 



