228 On ihe unequal R^JleMbilll^ of Light. 



of efFeding this decompofition by employing plane furfaccs, nor curve furfaces, whicliliavir 

 not a very minute, and, as it were, evanefcent radius. It may, in fa£t, be conceived, that a 

 fmaii portion of a curved furface of a greater radius is a plane with regard to a particle of 

 light. The author indeed, explains this phenomenon in another manner ; but the faft, 

 independent of all explanation, is no* the lefs certain and acknowledged. 



Seition lo. Let H H H, fig. 3, now reprefent a very fmall bright poliQied cylinder (a metallic 

 fibre), and B R V K the cylinder on the fame axis, which is the fphere of aitivity of this 

 fmall body : each of them being reprefented by its circular fection. (Thefe two circles, 

 though very unequal, are confounded together in aftual obfervation.) 



A B reprefents a white ray, incident ^t the point B, on the cylinder, on its fphere of 

 a(Stivity. 



Suppofe the homogeneous rays to be unequally rcpulfive, and that the red be more ftrongly 

 repelled than the violet: Mr. Brougham admits this fuppofition (p. 267, or Philof. Jour- 

 nal, i, 592). 



On this hypothefis, the violet ray muft penetrate farther into the fphere of repulfion. 



The courfe defcribed by an homogeneous ray, within the fphere of repulfive aiElivity, muft 

 be formed of two equal and fimilar branches : and its axis rnufl. pafs through the centre of the 

 (phere or fedion. This follows from the principle before laid down (§6). 



And it is an immediate confequence of this remark, that this homogeneous ray will iflue 

 from the fphere of activity, under an angle of refledlion equal to the angle of incidence. 



So that all the homogeneous rays forming, in B, the fame angle of incidence, will be 

 reflefted under equal angles. 



But fince fome penetrate farther into the fphere of aftivity than others, they will come out 

 divergent; for this is the only condition by which the equality of the angles of refledtion can 

 be preferved. 



Fig. 3 is intended to fliew this e&£l. The red ray penetrating Icfs into the fphere or 

 cylinder of adtivity 3 R V K, defcribes the curve B O R, the axis of wliich paffes through 

 the centre C ; and it emerges through R G, making the angle of refleiSlion E R G = 

 A B D, the angle of incidence. The violet ray, penetrating deeper, defcribes the curve 

 B Q V, of which the axis liicewife pafTes through C. This ray emerges through V L, and 

 the three angles FVL, ERG, ABD, are equal. 



But the obferver feeing the arc B R V like a point, and knowing that the angles which 

 he meafures are the fum of the angle of incidence and of refleflion for the rays of each 

 kindy will be induced to think, that, under the fame angle of incidence, the angles of reflec- 

 tion vary: for he will find that the right line AB, forms unequal angles with the right lines 

 RG, VL; and, in a word, he will have all the fame appearances which prefented them- 

 felves to Mr. Brougham (§5). 



It is of importance to remark here, that though this philofopher affirms that the rays on 

 the confine of blue and green are refiefted under art angle of reflcftion equal t(^ the angle 

 of incidenee, he did not, nor could not, afcertain this equality by any dired experiment : it 



was 



