CONTEMPORARY ADVANCES IN PHYSICS 397 



inner surface it will shine. The crystal should be very small in 

 comparison with the bulb, and the incident beam of waves extremely 

 narrow. We suppose that this beam comes in at a window, and 

 follows a diameter of the bulb, and the unscattered part after flowing 

 through the crystal goes out through another window. If the rays 

 are plane-polarized X-rays, the electric vector will remain constantly 

 parallel to some direction at right angles to the beam,. If they are 

 unpolarized X-rays, the electric vector will run or swing rapidly 

 around in the plane at right angles to the beam. Thus far it is 

 customary to use in crystal analysis X-rays which either are un- 

 polarized, or have the slight degree of polarization usually imprinted 

 on such rays at their excitation in a discharge-tube or a Coolidge tube.^ 

 We do not positively know whether electron- waves are polarized, but 

 we cannot alter their degree of polarization whatever it may be, and 

 it seems probable that they are not. 



This "imaginary bulb" is very nearly realized in practice, although 

 instead of a fluorescent screen it is customary to use a photographic 

 film on which the imprints of the diffraction beams are permanent 

 spots. The film is usually flat instead of spherical, so that the rings 

 presently to be mentioned are distorted from circles into ellipses. 

 Often an ionization-chamber (for X-rays) or a Faraday chamber (for 

 electrons) is swung around in arcs over a spherical surface centred at 

 the crystal, and the current which it reports is plotted as a function of 

 its position ; the curves then display peaks wherever the chamber is so 

 placed as to capture a beam. In what follows I shall use the terms 

 "diffraction beam," "diffraction peak" and "diffraction spot" almost 

 as synonyms. 



On the inner coating of the bulb, then, we shall see the diffraction- 

 pattern of the crystal. It will be an assemblage of luminous spots 

 arranged in a symmetrical array recalling the symmetry of the crystal 

 lattice. In making this statement I am anticipating what is presently 

 to be proved, or rather to be deduced from the fact that the atom- 

 groups of the crystal are marshalled on a lattice. Indeed, if the 

 incident waves are monochromatic or nearly so, we shall not see even 

 spots unless by happy accident or by a careful choice of wave-length. 

 Most waves are not diffracted by a three-dimensional crystal lattice. 

 If we could reduce our crystal to a single plane of atom-groups forming 

 a two-dimensional network, there would be a pattern of spots for 

 any wave-length whatever. If we could isolate a single row of atom- 

 groups, there would be rings of light on the coating of the bulb, 



- Because the electrons which produce the rays in falHng onto the target of the 

 tube are all moving along parallel lines when the impinge upon it. 



