VOL. VII.] PHILOSOPHICAL TRANSACTIONS. 715 



prism's reflection. I have sometimes designed to try, how a fixed star, seen 

 through a long telescope, would appear by interposing a prism between the 

 telescope and my eye. But by the appearance of Venus, viewed with my naked 

 eye through a prism, I presage the event. 



2. Concerning the second experiment, I have occasionally observed, that by 

 covering both ends of the prism with paper, at several distances from the mid- 

 dle, the breadth of the solar image will be increased or diminished as muc:h as 

 is the aperture of the prism, without any variation of the length t or, if the 

 aperture be augmented on all sides, the image on all sides will be so much 

 augmented, and no more. 



3. Of the third experiment I have occasion to speak in my answer to another 

 person ; where you will find the effects of two prisms in all cross positions of 

 one to another described. But if one prism alone be turned about, the coloured 

 image will only be translated from place to place, describing a circle, or some 

 other conic section, on the wall on which it is projected, without suffering any 

 alteration in its shape, unless such as may arise from the obliquity of the wall, 

 or casual change of the prism's obliquity to the sun's rays. 



4. The effect of the fourth experiment I have already insinuated, telling you 

 (in page 679) that light, passing through parts of the prism of divers thick- 

 nesses, did still exhibit the same phaenomena. 



Note, that the long axes of the two prisms, in the experiment described in 

 the said page 679, were parallel one to another. And for the rest of their posi- 

 tion, you will best apprehend it by this scheme : where, let EG design the 

 window (fig. 6, pi. 15) ; F the hole in it, through which the light arrives at the 

 prisms; ABC the first prism, which refracts the light towards PT, painting 

 there the colour in an oblong form ; and a[3y the second prism, which refracts 

 back again the rays to Q, where the long image PT is contracted into a round 

 one. 



The plane ay to BC, and j3y to AC, I suppose parallel, that the rays may 

 be equally refracted contrary ways in both prisms. And the prisms must be 

 placed very near to one another ; for if their distance be so great, that colours 

 begin to appear in the light before its incidence on the second prism, those 

 colours will not be destroyed by the contrary refractions of that prism. 



These things being observed, the round image Q will appear of the same 

 size which it does when both the prisms are taken away, that the light may 

 pass directly towards Q from the hole without any refraction at all. And its 

 diameter will equal the breadth of the long image PT, if those images be 

 equally distant from the prisms. 



If an accurate consideration of these refractions be designed^ it is convenient 



4x2 



