282 BRUSH— SOME DIFFRACTION PHENOMENA. [April 19, 



edges where they touched the rods, because it is only in these Hnes 

 of contact, or very near them, that the effects to be described are 

 produced. The glass rods were then removed and the bundle of 

 blades was mounted and used in the same manner as the two-blade 

 system O already described. 



With this device, which precludes reflection, all the eft'ects de- 

 scribed in connection with the mirror H may be reproduced, differ- 

 ing only, and differing but little, in brilliancy. As only about half 

 of all the edges (2 mm. across the edge of the bundle) are eft"ective 

 at any one time in producing visible fringes, it seems remarkable 

 that the latter are so brilliant. But we must bear in mind that, say, 

 twelve superposed fringe patterns will concentrate nearly all the 

 light into the bright bands, leaving the dark bands nearly black; 

 so that the contrasts should be nearly as strong as those produced 

 by the far greater . number of superposed patterns given by the 

 mirror H. 



The device R shows also something more of interest. Owing to 

 the limited number of patterns formed, failure in registry may be 

 seen at some points as division of a normal black band into two 

 narrower dark lines which merge when the tangent screw is slightly 

 turned, or the focal plane slightly moved ; and this phenomenon may 

 be shifted to different parts of the composite pattern by continuing 

 either or both of these adjustments. Thus relative shifting of vari- 

 ous fringe patterns, each more or less reinforced, is made obvious. 



