38 4< Mr. Potter on FresneVs Experiment of Interferences 



usually employed to throw the sun's light through the win- 

 dow-shutter into the darkened room, I have employed an 

 equatorial method of mounting, which keeps all the appa- 

 ratus in the adjustment it is first placed in, and at the 

 same time enables us to follow readily the sun's daily mo- 

 tion, keeping at the same time the room quite dark. 



To obviate all objections which attach to the luminous 

 point being formed by a common lens, which has different 

 foci for differently coloured rays, I have employed a spheri- 

 cal mirror, as better even than any achromatic lens; and 

 again, as it was suggested to me that an objection might 

 be raised if the rays crossed in a real focus, I have used a 

 convex mirror, so that they diverge from a virtual focus 

 without having crossed : also to enable me to use the lumi- 

 nous point the smaller, I have generally used the two mirrors 

 slightly inclined, of polished speculum metal, which reflects 

 many times the quantity of light which glass reflectors do in 

 the position which my apparatus requires. With these pre- 

 cautions, I find, when the sun is perfectly unclouded, and 

 near the meridian, high above the horizon, that the central 

 band is black. When there are clouds before the sun's disc, 

 however slight, the central band is more difficult to fix upon, 

 and generally either white or doubtful. The discrepancies 

 which have been to me a puzzle for so many years, I am now 

 able to solve, and I announce the following new principle of 

 interferences : — 



When light in a state of interference is made to interfere 

 again, the result is of an opposite character to nxihat it nsoould 

 have been if the light had been in the first instance in the 

 ordinary state. 



This proposition, which is in itself reasonable, might have 

 been anticipated, and solves all the anomalies. The light 

 falling on the small hole in the thin plate of metal, formerly 

 mentioned, as used to form the luminous point, is thrown 

 into a state of interference by diffraction at the edge of the 

 hole, and hence the central band is seen white. Again, the 

 sun's light passing through thin clouds, or through the va- 

 pours of the atmosphere when near the horizon, is thrown 

 more or less into a state of interference by diffraction at the 

 edges of the particles of the vapours, and gives results which 

 are either doubtful or with a white centre : for where part 

 of the light is in its original state and part in a state of inter- 

 ference, the bright and dark centred bands will be super- 

 posed, the central band of the one set over the central band 

 of the other, and thus produce indistinct phcenomena in which 

 the intensity of the one or the other species may prevail. 



