47l) Intelligence and Miscellaneous Articles. 



To control the accuracy of the mode of observation which we owe 

 to MM. Fizeau and Foucault, we made a great number of experi- 

 ments on the light of sodium by means of the very precise method 

 made known by M. Mascart in his excellent paper " On the Modi- 

 fications undergone by Light in consequence of the Motion of the 

 Luminous Source " ( Ann. jSc. cle VEcole Normale, 1872, t. i. p. 202). 

 Employing this method, which consists in operating on monochro- 

 matic light observed in a spectroscope with the slit very much 

 widened, M. Mascart found 21°-73 for the rotation of the light of 

 sodium at 15 degrees temperature. We arrived at the figure 

 21°' 727 at the temperature of 20 degrees, which corresponds to 

 21°*711 at 15 degrees. This slight difference appears to us to 

 come within the limit of the errors which this method permits 

 when applied to light which is not strictly monochromatic. 



On the other hand, we have made a great number of measure- 

 ments for the solar lines D 1 and D 2 by Fizeau and Foucault's 

 method. For D„ by varying the conditions of observation, we 

 obtained values between 21°*750 and 21°-714 : the mean of those 

 in which we have most confidence is 21°*736 at 20 degrees. For 

 D 2 we have found 21 0, 684. These numbers, between which falls 

 the value obtained by M. Mascart's procedure, appear to us to 

 prove the correctness as well as the delicacy of the method of MM. 

 Fizeau and Foucault. 



In making a first series of experiments the spectroscope we used 

 was furnished with lenses of quartz and a prism of Iceland spar, 

 with the fluorescent ocular for the observation of the ultra-violet 

 rajs*, and with the ordinary eyepiece for a certain number of 

 measurements taken in the luminous portion of the spectrum. In 

 a second series of experiments we made use of ordinary lenses of 

 optical glass, employing two flint-glass prisms for the lines com- 

 prised between A and Gr, and one prism only for the portion ex- 

 tending from h to M. 



* The spectroscope with fluorescent eyepiece, described by one of us, 

 has well answered its purpose for the very refrangible rays from N to R. 

 The solar light has to be reflected from a German-silver mirror (not a 

 silver one, because silver absorbs the highly refrangible rays, as Prof. 

 Stokes has already shown). To avoid the rotation which the collimator 

 lens of the spectroscope impresses on the polarized rays when it consists 

 of only one biconvex lens of quartz, a compensated lens is employed, 

 formed of two plano-convex quartz lenses, perfectly equal, one with right- 

 handed, the other with left-handed rotation, applied the one against the 

 other by their plane faces. The Iceland-spar prism has its edges cut 

 parallel to the axis ; it therefore gives two spectra, of which the most 

 deviating (the ordinary ray) is observed in preference. The substances 

 which appear to suit best for the fluorescent layer are an aqueous solution 

 of esculine for the portion of the spectrum from h to N ; and a plate of 

 uranium-glass for the rays of shorter wave-lengths : with this latter sub- 

 stance the lines R, S, and even T can be distinguished in the solar spec- 

 trum ; and in the spectra of the metals the most refrangible lines can be 

 8een — for instance, the twenty-fifth line of cadmium (X = 221-7, Mascart). 



