1906.] o?i Recent Progress in Magneto-Optics. 293 



the circular polarisation is. There is no trace of rectilinear or of 

 elliptic polarisation.* (Fig. 8.) 



When I first looked for this circular polarisation, T did not have 

 the field of view divided into two parts, but the position of the line 

 was determined by means of a spider's thread. On the reversal of 

 the magnetising current, the luminous line moved. I do not wish to 

 disguise the fact that no observation has ever afforded me so much 

 pleasure as this one. 



It has already been remarked that we can also study the absorp- 

 tion lines which become visible when white light is transmitted 

 through the vapour. We then study the inverse effect. I shall use 

 it to show you at least something directly depending upon the effect, 

 because the effect itself is too young to appear before so large an 

 audience. The inverse effect for light parallel to the lines of force 

 plays a part in an experiment due to Righi.f Consider a horizontal 

 ray parallel to the axis of an electromagnet with pierced poles, and let 

 crossed Nicols be placed before and behind the instrument, as in 

 Faraday's experiment. A sodium flame in the field emitting two kinds 

 of circularly polarised rays, absorbs these same radiations, but does 

 not stop the radiations polarised in the opposite direction. These 

 remaining circularly polarised rays cannot be extinguished by a Nicol. 



The brilliant yellow spot which appears on the screen as soon as 

 the current is put on, is due to such rays. The explanation of this 

 experiment is not complete however, at least not for denser vapours. 

 The Faraday rotation of the plane of polarisation then plays a part, as 

 we shall see further on. 



The magnetisation of the spectral lines allows us to determine 

 whether positive or negative electrons are vibrating in a flame. From 

 the phenomena in the direction of the lines of force, it follows that 

 in a luminous gas the negative electrons give rise to all vibrations. 

 It does not follow, however, that the luminous molecules have a 

 negative charge. On the contrary, the researches of Lenard and 

 Stark show that at least part of the luminous spectra is emitted 

 by positively charged atoms. 



When a line is split up into a triplet, we can, by measuring the 

 amount of the effect, find out how much matter is loaded with the 

 revolving electron, or in other words, we can determine the ratio of 

 the charge e to the mass m of the electron. In this manner, I have 



made the first determination of this notable number -, and found it 



m 



of the order of magnitude of 10' electro-magnetic units per gramme. | 



The most accurate measurements of the present time for different 



* Cf. Larmor. Aether and Matter, p. 345, 1900. 



t Righi. C.R. 127, p. 216, 1898. C.R. 128, p. 45, 1899. Nuovo Cim. (9), 

 8, p. 102, 1898. 



X Zeeman. Verslagen Kon. Akademie, Amsterdam, November, 1896. § 23, 



