VOL. XXXV.] PHILOSOPHICAL TRANSACTIONS. 295 



the colours separated most at the greatest^ and least at the least distance. This 

 mistaking a reflection for a refraction, has been the occasion of several more 

 errors and difficulties, to be met with in Siguier Rizzetti's book. 



Exper. 3. — A candle k, fig. 6, reflected from the surface ab of the prism 

 ABC, appeared very faintly to the eye at e, as a weak image at k ; because the 

 rays incident at i, pass most of them through the prism, and go on to k, 

 separating from each other according to their different degrees of refrangibility; 

 while a few of them are reflected to the eye in the direction ie. 



But if the prism be in the position acb, fig. 7, most of the rays of the candle 

 K, incident at i, on the plane ab, after having passed perpendicularly through 

 the plane bc, are reflected, and passing perpendicularly through ac, go into 

 the eye at e, which sees a very strong image of the candle at k, while very ievi 

 rays go down to r to produce colours. 



This shows that the rays of light pass with more facility through glass (a 

 dense) than through the air (a rare) medium; contrary to Rizzetti's assertion, 



Exper. 4. — To make this more evident, and compare together the facilities 

 with which light passes through the two mediums, the Doctor took a cube of 

 glass, of 3 inches the side, AabBdDC, fig. 8, whose section is abcd, and. view- 

 ing it from E, to see by reflection the candle k, he saw two images of it ; one 

 at k very faint, and reflected from the upper surface ab, and the other at x. very 

 strong, and reflected from the lower surface CD. Now it is evident, that the 

 brightness of the images >c, is to that of the image k ; as the facility with 

 which the rays, in these circumstances, pass through the glass, or through 

 the air : and those are easily compared, because both the images are seen at 

 once. 



Exper. 5. — The line pi, fig. Q, being perpendicular to the reflecting plane 

 AB, of the triangle acb, the Doctor brought the candle k by degrees so near to 

 p, as to diminish very much the angle of incidence kip; which made the image, 

 or appearance of the candle at k, become fainter by degrees, and at last as faint 

 as in fig. 6. 



Exper. 6. — Having made the experiment as at fig. 7, the Doctor pressed an- 

 other prism DFG, fig. 10, close to the prism abc, and when he pressed them 

 together but gently, some of the rays from the candle r, passed through the 

 lower prism, and falling on a paper at r, made a reddish spot ; but when he 

 pressed them very hard, the spot became much wider, white in the middle, and 

 only tinged with red about the edges : at the same time the eye saw a black 

 spot in the image of the candle at k ; and a stander-by looking obliquely at the 

 place I, where the glasses touched, saw, as it were, a little hole through the 

 prismSj as large as the spot k. But if the prisms be pressed together but gently, 



