Ill 



by Sir John Herschel, who has investigated the case in which the 

 two surfaces of the glass belong to a pair of concentric spheres, and 

 the hole in the screen is situated in the common centre of curvature. 



A set of coloured bands has since been observed by Dr. Whewell 

 in a common plane mirror. These bands are seen when a candle is 

 held near the eye, at the distance of several feet from the mirror, 

 and is viewed by reflexion. It is necessary that the first surface of 

 the glass should be a little tarnished. The theory of these bands 

 had not been worked out, and it had even been doubted by some 

 philosophers whether they were of the nature of the coloured rings 

 of thick plates. 



In this paper the author gave a general investigation, which in- 

 cludes as particular cases the theory of the rings formed on a screen 

 in Newton's experiment, and that of the bands which Dr. Whewell 

 had observed in a plane mirror, and which are not thrown on a 

 screen, but viewed directly by the eye. He also exhibited to the 

 meeting a variation of Newton's experiment, in which an extremely 

 beautiful system of rings is very easily produced without sunlight. 

 The face of a concave mirror of quicksilvered glass was prepared by 

 pouring on it a mixture consisting of one part of milk to three or 

 four of water, and then holding the mirror vertically in front of a 

 fire to dry. When the flame of a taper, or of an oil-lamp with a 

 small wick, is placed in front of a mirror thus prepared, in such a 

 position as to coincide with its inverted image, a beautiful system of 

 rings is seen encompassing the flame. These rings appear to have 

 a definite position in space, like a bodily object. The rings thus 

 formed, which are evidently of the nature of Newton's coloured 

 rings of thick plates, may be made to pass in a perfectly continuous 

 manner into the coloured bands observed by Dr. Whewell. 



The author has compared theory and experiment in various par- 

 ticulars, and has found the agreement perfect. It will be sufficient 

 to mention here one result of theory, which is of great generality 

 and of considerable elegance. It applies to the system of rings seen 

 by reflexion in a mirror, either plane or curved, when a luminous 

 point is placed anywhere near the axis, and the eye occupies any 

 other position likewise near the axis. The result is as follows : — 

 Join the eye with the luminous point, and likewise with its image, 

 whether it be real or virtual, and find the points in which the join- 

 ing lines, produced if necessary, cut the mirror. Describe a circle 

 having for diameter the line joining these two points. This circle 

 will be the middle line of the bright colourless fringe of the order 

 zero, and on each side of it the colours will be arranged in descend- 

 ing order. 



June 2, 1851. 



On a new Elliptic Analyser. By Professor Stokes. 



After mentioning some of the inconveniences and inaccuracies 

 attending the use of a Fresnel's rhomb in the analysis of elliptically- 

 polarized light, and alluding to some other methods which had been 



