on the Affectio7is and Properties of Light. 359 
necessary; but its great importance to the whole theory will, 
I hope, plead my excuse. 
Let us now suppose that a homogeneal beam passes through 
the spheres of flexion, it will follow that no divergence can 
take place from the bending power of the body ; so that we 
have only to estimate the effect produced by the reflexion, and 
to inquire whether the different reflexibilities of the rays can 
cause the images to vary their sizes according as they are form- 
ed by different rays. In fig. 3 . let AB be the body, C D the li- 
mit of its sphere of reflexion, and I P a beam of homogeneal 
rays, as red, incident at P and reflected to R, forming there 
the image Rr. It is evident that the greater reflexibility of the 
rays I P can only alter the position of the centre of Rr, making 
it nearer the perpendicular than the centre of an image formed 
by any other rays would be. But the greater length of Rr 
shews that a greater quantity of rays is reflected, or that the 
same quantity is spread over a greater space, and that in the 
following way. Let I F fi be a beam of violet-making rays 
entering A B C D, and reflected so as to form the image R v. 
The force exerted by A B decreasing according to some law 
(of which we are as yet ignorant) as the distance increases, is 
not sufficient to turn the rays back till they have come a cer- 
tain length within A B C D. But for the same reason it turns 
back all that it does reflect before they come nearer than a cer- 
tain distance ; between these two limits, therefore, the rays are 
turned back. But the limits are not the same to all the rays ; 
some begin to be turned at a greater distance from the body 
than others, and consequently are reflected to a greater dis- 
tance from the middle ray of the incident beam. Thus if I Ff i 
be changed to a red-making beam, it begins to be turned back 
