APPENDIX TO MEMOIR OF PELTIER. 189 



mastered tliem; but lie had experimented little, and all that remains to us of 

 his in this branch of physics is the improvement which he applied to the cyano- 

 polarimeter of Arago. 



Every one knows how much the azure color of the sky varies with the quan- 

 tity and state of the vapors diffused in the atmosphere; every one knows also, 

 since the discovery of M. Arago and the researches of MM. Quetelet and Dele- 

 zeune, that the air polarizes light and that the intensity of this polarization is 

 not the same at all points of the sky, nor the same for the same point at all 

 hours. There was nothing, for a longtime, wherewith to measure the variations 

 of the azure of the sky but the cj^anometer of Saussure ; for the cyanometer of 

 Arago, as designed by him in 1817, had never been realized. As to the polari- 

 zation of the atmosphere, there existed for its study only the polariscope of 

 Savart and that of Arago. But the cyanometer of Saussure is a very imperfect 

 instrument which can yield none but very imcertaiu results; as regards the 

 polariscopes of Savart and Arago, they are both, it is true, extremely sensitive, 

 but as they are destitute of the means of measurement, they could not serve for 

 exact observations. 



In the sitting of 25th of October, 1841, Arago communicated to the Academy 

 of Sciences a polarimetcr of his own invention. This instrument was the 

 polariscope proposed by the same savant in 1811, but to which a particular 

 apparatus had been adapted. The polariscope of Arago becomes a polarimeter 

 by the sole addition of one or more plates of glass with parallel faces, placed 

 in front of the old instrument. These plates are movable. A graduated circle 

 indicates the inclination under Avhich the light has traversed them, before pene- 

 trating into the polariscope, properly so called. The proportion of polarized 

 light contained in the pencil observed is deduced from the angle at which it is 

 necessary to adjust the plates of glass in order to perceive no longer any trace 

 of color athwart the whole apparatus. 



In the sitting just mentioned, Arago had presented to the academy the instru- 

 ment as constructed and aiTanged by himself; at a succeeding session, Novem- 

 ber 15, he submitted to the inspection of the academy this same polarimeter 

 constructed upon his model, but executed by M. Soleil ; this instrument is known 

 as the cyano-polarimeter of Arago. Capable of serving at once as a cyanometer 

 and polarimetej, it was, beyond doubt, greatly superior to the instruments pre- 

 viously in use for studying the variations of the blue color of the sky and the 

 differences in the quantity of light polarized by the atmosphere ; yet was it not 

 without defects : first, as concerns cyanometry, it wanted several important 

 means of measm'ement; then, as regards polarimetry, it could in reality render 

 service in only two rectangular planes : in the plane, namely, of the meridian, 

 and in that of the equator of the aerial sphere, of which the sun is one of the 

 poles, and the anti-sun the other pole ; outside of these two planes, it could be 

 of no utility. Peltier applied himself to correct these defects, and completely 

 succeeded in doing so. 



Optical principles of cyanometry. — If we take a crystal having a single axis 

 of double refraction, such as Iceland-spar, the beryl, &c., and cut from its mass 

 a slip of which the two faces shall be exactly perpendicular to that axis, and 

 if we then cause a ray of polarized white light to fall peipendicularly on this 

 slip, so that it shall traverse the crystal exactly in the direction of its axis, the 

 ray will undergo modification. If we now analyze it on its emergence with an 

 achromatic double-refracting prism, taking care to place the principal section of 

 this prism in the plane itself of the polarization of the raj^, the ordinary image 

 contains the entire ray; that is to say, the complimentary tints are black and 

 white, and there is no coloration. Quartz, however, forms an exception to this 

 rule. When, in effect, a ray of polarized white light is made to pass through a 

 lamina of quartz (rock crj'stal) cut perpendicularly to the axis, and this ray, as 

 in the previous case, exactly follows the direction of the axis, if we in like man- 



