82 Mr. R. Potter on a particular Modification 



angle about 43 degrees. By using a homogeneous light pro- 

 duced by a coloured solution and two plane mirrors of spe- 

 culum metal, or rather one mirror polished as for a New- 

 tonian telescope and then cut into two across the middle, I 

 obtained sufficient brightness to trace readily the whole phe- 

 nomena. I immediately found that Professor PowelPs con- 

 clusion from using the prism with a small angle, was prema- 

 ture, and that the same portions of the pencils did not interfere 

 after refraction which would have interfered before, if the 

 prism had not been interposed; but that interference then 

 took place between rays which had passed at a greater di- 

 stance from the angle of the prism. Another phenomenon 

 which greatly attracted my attention was, that when the eye 

 and eye-glass were withdrawn further from the prism, the in- 

 terference occurred between other parts of the pencils which 

 had passed at still greater distances ; and that when the eye 

 and eye-glass were withdrawn still further, all appearances 

 of interference at length ceased. This last effect arose where 

 the prism became too narrow to allow a sufficient breadth of 

 the pencils to pass, or when the mirrors were not sufficiently 

 inclined to each other to give the required overlapping of the 

 pencils for interference to take place at such a distance from 

 the prism. The inflected or diffracted bands produced by 

 the edges of the mirrors give a certain criterion by which to 

 judge of this other interfering' light; and in the act of drawing 

 the eye and eye-glass from the prism, we see the phenomena 

 which we consider traverse gradually over those inflected or. 

 diffracted bands, and finally become as gradually lost in the 

 single light of the other mirror. This appearance takes 

 place whatever may be the angle at which the pencils emerge 

 from the prism. Another fact which demands equal attention 

 is, that the breadths of the fringes produced by the inter- 

 ference vary with the inclination of the light to the surfaces of 

 the prism : thus from the angle of minimum deviation towards 

 a perpendicular incidence on the first surface, the fringes be- 

 come narrower and narrower, and on the contrary side of the 

 angle of minimum deviation they grow larger. 



The above will be more easily comprehended on referring 

 to fig. 1, where a and b are the images of the luminous 

 point o, produced by the two mirrors fg and gh ; a' and h* 

 the secondary images after refraction by the prism dee. 

 Let a i and b k be the axes of the pencils drawn perpendicu- 

 lar to the line joining the points a and b. Let m n be a line 

 parallel to a i and bk 9 and exactly intermediate between them. 

 After refraction, these lines must be considered as in the di- 

 rections a' z 7 , b' /c 1 , ?n' n 1 . Now where the prism does not inter- 

 vene, the central band produced by the interference of the 



