364 Mr. G. Dodd on 



multiplo-reflected rays could not fill up the dark spaces 

 which would result in the prismatic spectrum. 



In most Optical works, the relative proportions between 

 the different coloured spaces of the spectrum are given for 

 two or more substances : but neither the refracting angle 

 of the prism, nor the point of the side at which the ray- 

 enters, are stated as modifying the proportions between the 

 colours : the spectrum may be longer or shorter according 

 to the angle of the prism and the obliquity of the ray's 

 passage, but the relative proportions are not altered : now, 

 the points at which the internal multiplied reflexions take 

 place, depend almost entirely on these three conditions : 



1. The refracting angle of the prism. 



2. The parallelism or obliquity of the central ray's path, 

 with respect to the base. 



3. The point of immersion, whether midway between the 

 apex and the base, nearer to the base than to the apex, or 

 nearer to the apex than to the base. 



Let us suppose the prism to be of flint glass : the spectrum 

 is then of a certain nature, nearly, if not quite, equal in its 

 proportions, however the prism be held : that is, any 

 mode of using it which would lengthen or shorten one 

 coloured portion, would do so equally to the others : but, 

 supposing there to be three colours and three indices of 

 refraction, the rays, after multiplied reflexion within the 

 prism, would emerge at every imaginable position with 

 respect to the primary rays : the violet emersions, for in- 

 stance, might be all on their proper side of the green in 

 one experiment, while in another, they would be largely 

 mingled with the red on the opposite side, and this with the 

 same prism : the difference being due to a variation of the 

 point at which the light enters, or a variation of the angle 

 at which the primary ray passes through. 



Mr. Cooper has shewn in a diagram that the effect which 

 he attributes to internal reflexion may be produced, or at 

 least, that all the emersions of any one colour may be 

 parallel; but, I beWeve, he will find that that can only take 

 place when these three conditions are observed : 



1. The prism to be equilateral. 



2. The point of immersion to be midway between the 

 apex and base. 



