145 



lillinu tie deep into the substance, it is only a relatively thin layer of 

 yMalline substance that is engaged in the phenomenon considered, 

 and in every case the number of "reflecting" net-planes is a finite 

 mi-. Only when the reflected wave-trains are in the same phase, i. . 

 they interfere with phase-differences of A or a multiple of A, 

 interference-maximum will occur. Now if bS be the plane per- 

 mlioular to the incident beam of radiation, and aS that perpendi- 

 ular to the "reflected" beam L\ L' 2> the difference in the path 

 travelled by a ray coming from V lt and by that coming from F 2 , 

 obviously be bP + Pa. But bP = Pa, is the projection of 

 le distance d between two consecutive net-planes V v and V t 

 ipon the direction of the incident an emergent beam, and therefore 

 jqual to dsinQ, when < is the glancing angle which the incident 

 pencil makes with the planes V l or V 2 . The whole phase-difference 

 therefore: 2d.sin<p and an interference-maximum will only 

 :cur, if this difference be equal to A, or to a multiple of it: 2A, 3A, 



etc. 

 It is obvious that from the equation: 



it follows that for constant d and for each definite value of A, the 

 angle < can only have definite' values (f)^ $,. $3, e tc., the sines of 

 which are rational multiples of each other. Conversely: if V l be given 

 a certain crystal, d is wholly determined by the internal specific 

 fstal-structure, and when homogeneous radiation of a known 

 /ave-length A be used, we have only to measure (P lt <p z , etc., to 

 id the distance d between two consecutive layers parallel to F t . 

 )n the other hand it must be clear that from all wave-lengths present 

 the incident radiation, only that which is equal to A will be 

 ^fleeted under the angles mentioned, when the plane is in a fixed 

 )sition. The reflection at such a fixed set of net-planes under a 

 :onstant glancing angle $ has therefore the effect of separating 

 mly a special wave-length A out of the total number of wave-lengths 

 present in the incident rays; it changes the incident radiation, being 

 a complicated wave-motion, into a "monochromatic" one of definite 

 wave-length A corresponding to a certain glancing angle Q, and 

 therefore such a reflection has a pronounced selective action. It 

 may be remarked here that F x need not be a limiting plane of the 

 crystal; the so-called reflection occurs within the crystal, and at 

 the parallel, equidistant net-planes present therein. 

 When the radiation falling on the crystal is monochromatic, the 



10 



