8 Rheinberg-: Common Basis ofthe Theories ofMicroscopic Vision. XIX, 1. 



Next consider the poiiit B, (fig. 4). From the diagram it is 

 evident that the point Yis two half wave lengths distance from T", 

 aud that undiüatious therefore reach M from ^ exactly a whole 

 wave length later than from Y". The diagram fiirther shows, that 

 dividing the wave front J^Yinto two parts at C\ each point between 

 J^ and C has a corresponding point between C and Y exactly -|- a 

 wave leugth behind it. As each of these poiuts uentralises the other, 

 there can be no effect prodnced at R at all. 



We have then at P' (fig. 6) a point of maximum ilhimination, at 



R a point devoid of liglit, and at Q 

 in between, the light is approximately 

 I of that at P'. It natiirally follows 

 in like manner that between these points 

 P' QP, there is a gradnal transition. 

 Figiire 7 represents the intensity of a 

 disc of light thiis formed. 



Now consider the point S (fig. 5). 

 Here we see that undiilations from Y 

 reach the point '^/^ wave lengths behind 

 those from Jl. Divide the wave front 

 XY into three parts XD^ DE, EY. 

 It is Seen that XD and DE nentra- 

 lise each other as regards their etfect 

 at 5 in precisely the same way as 

 indicated with regard to R in figure 4. 

 Y' But the remaining part E Y produces 

 some light at S. We fonnd that when 

 6- the Variation in distance from nil to -J 



wave length was spread over the whole 

 of the wave front under consideration (fig. 3) that the amplitude of 

 the light at Q was f^^- that at P'. Now that this Variation up 

 to -J- wave length is distributed over E F, or only about ^ of XY 

 the amplitude at S cannot exceed \ X | = '-/g. By squaring this 

 to obtain the relative intensity of the light, we get ^1^^ or something 

 less than ^/.,o of the light at P. 



The point T (fig. 6) is such that light reaches it from Y four 

 half wave lengths later than from X. X Y could therefore be divided 

 into 4 parts, each consecutive part of whicli would cancel each other, 

 and prevent any light reaching P. 



We have now found that we have a ring devoid of light at 



\T im 



Y'" 



Y" 



