296 PHILOSOPHICAL TKA.NSACTIONS. [ ANNO ] 728. 



then all the other phasnomena disappear, except the first little spot at r, as in 

 %. 11. 



When the candle is seen by reflection from the lower surface of a prism, as 

 in the 7th, gth and 10th figures, the rays pass quite through that surface, and 

 are turned up again, by its attraction, in curve lines, so as to re-enter the prism 

 and then, going out again through the surface ac, go up to the eye at e. In 

 this case, the most refrangible rays, being the most easily inflected, make the 

 least curves, whose vertices are nearer the glass than those of the greater curves 

 made by the least refrangible rays. This is proved by experiment 6, where the 

 under prism only attracts down, from the reflection of the upper prism, the 

 red making rays, as in fig. 11, where the plate of air between the prisms is of 

 some small thickness. But when the prisms, whose surfaces are a little convex 

 are pressed hard together, the lower prism is near enough to attract rays of a 

 great degree of refrangibility ; and therefore the spot then becomes white in 

 the middle ; and only red about the edges, which are produced by such parts of 

 the lower prism as are not so near the upper. 



There are two circumstances in the ()th experiment, which disapprove 

 Rizzetti's assertion (page 125) viz. That there is a sensible reflection even where 

 glasses touch ; for when the prisms couch at i, fig. 10, the black spot appear- 

 ing in the image of the candle k, shows that there is at i a deficiency of those 

 rays, which, coming from the middle of the candle, used to be reflected up to 

 the eye at e, and therefore that ab, the reflecting surface of the upper prism, 

 ceases to reflect in a little space round about i, where the upper surface df of 

 the under prism touches it ; the rays, which before were reflected, now going 

 down to make the spot at r. The other circuinstance is this ; that whereas a 

 paper at k is invisible to an eye at e, by the interposition of the prism dfg ; 

 when another prism acb is laid over it, and pressed hard, there appears to be a 

 hole, of about -i- of an inch (more or less in diameter as the prismatical surfaces 

 are more or less fiat) through which the paper at k becomes visible ; this being 

 the place of contact where the reflection downwards, of the surface df, ceases. 

 This happens, because those rays, which, coming from the candle k, were 

 bent in curves under the surface ab of the upper prism, about several points 

 near i, are by the nearness of the surface df of the lower prism, brought down 

 to R, instead of being turned up again to the eye at e; while those rays which, 

 coming from the paper at k, through the surface gf of the lower prism, and 

 passing through its upper surface fd, were bent in curves about several points 

 near i, are prevented from turning down again to r, and are brought up to the 

 eye at e, which consequently must see a round part of the paper at k, just as 

 large as the place of contact, which appears like a hole; or as if the two prisms. 



