VOL. LXXXVII.] PHILOSOPHICAL TRANSACTIONS. 207 



were distended, becoming very bright and beautiful; when the inclination was great, 

 and when it was still increased, another set of colours emerged from the side next 

 the speculum, and was concave to that side. Here I stopped the motion, and the 

 image on both sides of the focus had 3 sets of fringes, and 4 fringes in each set; 

 but when viewed through a prism, as before described, the numbers greatly in- 

 creased, both the fringes and the dark intervals decreasing regularly. The ap- 

 pearance to the naked eye is represented in fig. 10, where adc being the image, 

 A and c are the sets of fringes at the edges, and b the 3d set, there being none at 

 e and d the sides, since the light which illuminates these quarters comes not from 

 the edges of the speculum in so great inclinations. I now viewed the surface of 

 the speculum, and saw it, in the place answering to b in the image, covered with 

 fringes exactly corresponding with those at b ; and on changing the figure of that 

 part of the speculum's edge between them and the sun, the fringes likewise had 

 their figure altered in the very same way. On moving the speculum farther round, 

 b came nearer to a in the image, according as the fringes on the speculum receded 

 from that side which formed them; and before they vanished alike from the speculum 

 and image, they mixed with the colours at a in the image, and formed in their 

 motion a variety of new and beautiful compound colours: among these I particu- 

 larly remarked a brown chocolate colour, and various other shades and tinges of 

 brown and purple. Just before the fringes at b appeared, the space between a and 

 c was filled with colours by reflexion, totally different in appearance from the 

 fringes; but I could not examine them so minutely as I wished in this broad image, 

 I therefore made the following experiment. 



Observ. 11. At the hole in the window-shutter I held the speculum, and moved 

 it to such an inclination that the colours by reflexion might be formed in the image; 

 they were much brighter and far more distended than the fringes, and were in every 

 respect like the images by reflexion in the common way, only that the colours 

 were a little better and more regular. They were also seen on the speculum as the 

 3d set of fringes had before been in observ. 10; but by letting the rays fall on the 

 half next the chart, and inclining that half very much, I could produce them, 

 though less distinctly, by a single reflexion. I now held a plain metal speculum 

 so, that the rays might be reflected to form a white image on a chart: On inclining 

 the speculum much, I saw the image turn red at the edge; it then became a little 

 distended; and lastly, fringes emerged from it well coloured, and in regular order*, 

 with their dark intervals. This may easily be tried by candle-light with a piece of 

 looking-glass, and those who without much trouble would satisfy themselves of 

 the truth of the whole experiment contained in this and the last observation, may 

 easily do it in this way with a concave speculum ; but the beauty of the appearance 

 is hereby quite impaired. After this detail it is almost superfluous to add, that the 

 fringes at b, fig. 6, are formed by deflexion from the edge of the speculum next 

 the sun, and then falling on it are reflected to the chart; that the images by re- 

 flexion are either formed by the light being decompounded at its first reflexion, 



