soMii coxir.Mi'nK iRv .inr.ixcis ix I'liy^ics rii .us 



fHMuU'iilly in its own f-ishion.'*. TluTi-fori' tin- (|uantiin) imisl hv widt-r 

 than the widest slit which displays rlear ditTraitiDn-phenoniena- and 

 this makes it at least a millimetre wide! Mut this is not the limit! 

 ("ut another slit in the screen, parallel to the first one, a distance d 

 away from it. Where the widening; ililTracted li^ht-beams from the 

 two slits interpenetrate one another, they will produce interferern-e- 

 pat terns of li^ht and shade, accurately the sanie as we should expect 

 if the wavefront is wider than the distance d. The quantum must 

 therefore be wider than the Krt*'>tt-'!^l distance between two slits, the 

 liv;hi-l)eams passing throuijh which are able to interfere with one 

 another. The slits may be put cpiite far apart, and the light-beams 

 brought together by systems of prisms and mirrors. This is the 

 principle of Michelson's famous method of determining the diameters 

 of stars. He obtained interference fringes when the two beams of 

 light were taken from portions of the wavefront twenty feet apart!^^ 



Therefore the quantum is twenty feet wide! This is the object 

 from which an atom one ten-millionth of a millimetre wiile can suck 

 up all its energy! this is what enters as a unit into collision with an 

 electron ten thousandfold smaller yet! 



The evidence is now before the reader: iiol tlu- entire exidenre for 

 either of the two conceptions of radiation, but, I think, a fair sampling 

 for both. If either view has been ine(|uitably treated, it is the un- 

 dulatory theory which has been underrated; for, as I have said already 

 but cannot say too often, the e\idence that light partakes of the nature 

 of a wave-motion is tremendously extensi\e and tremendously com- 

 |ielling; it seems the less powerful only because it is so thoroughly 

 familiar, and through much repetition has lost the force of no\elty. 

 .Still, it is not necessary to hold all the rele\ant facts continually in 

 mind. If one could reconcile a single typical fact of the one sort, such 

 as the interference between beams of light brought together from par- 

 allel courses far apart, with a single outstanding fact of the other sort, 

 sach as the instantaneous emergence of electrons with great energy 

 from atoms upon which a feeble beam of light has only just been 

 directed — if one could unify two such phenomena as these, all of the 

 others would probably fuse spontaneously into a harmonious system. 

 Hut in thinking about these things, there is one more all-important 



''■ ( )nc might, of course, inquire, why should a piece of the uavefronl of a quantum, 

 rut out of it by the edges of a slit, expaml after passing through the slit when the 

 quantum itself apparently rushes through spaie without expanding? 



" It might be argued that these quanta from stars have come an enormously 

 long way, and possibly have had a lx;tter chance to expand than the quanta passing 

 across a laboratory room from an X-ray tube or a mercury arc to a metal plate. 

 However, since the photoelectric cell is used to measure the brightness of a star, they 

 evidently prixluce the same sort of photoelectric effect as newborn quanta. 



