142 Mr. A. K Tutton [May 2, 



fracting edge is parallel to an axis of the ellipsoid. Employing it in 

 the usual manner to throw a spectrum on the screen, we observe 

 that instead of a single spectrum, as when we use a prism of glass, 

 we have two spectra produced, differing considerably in their dis- 

 persion. Moreover, on placing a Nicol prism in the path of the light, 

 we see that these spectra consist of polarised light, for one extin- 

 guishes when the Nicol is arranged with its vibration direction at 0°, 

 and the other when the Nicol is at 90°. One of the spectra, in fact, 

 is formed by light vibrating parallel to the refracting edge, and 

 therefore to that axis of the ellipsoid which runs parallel with it, 

 and the other by light vibrating at right angles to that direction. 

 If, in cutting our prism, we make this perpendicular direction to 

 coincide with a second axis of the ellipsoid, the two spectra, when 

 set to their minimum deviation, will, provided we know also the 

 angle of the prism, at once afford us the data for computing the 

 refractive indices corresponding to the two axes of the ellipsoid in 

 question. A second prism, similarly cut so as to have vibration 

 directions parallel to one of these axes and to the third axis, will 

 afford us the third refractive index as well as a repetition of one of 

 the first two. 



You observe that the two spectra are considerably separated on 

 the screen. If the two axial directions are those of maximum and 

 minimum velocity, the amount of separation is a measure of the 

 double refraction. Now it has been observed that the amount of the 

 double refraction also varies progressively according to the atomic 

 weight, and this fact may be illustrated realistically by means of a 

 carefully made lantern slide, reproducing the actual positions, as 

 seen in the spectrometer, of the two spectra afforded by a triplet of 

 salts. The triplet chosen consists of potassium zinc sulphate, rubi- 

 dium zinc sulphate and caesium zinc sulphate. The spectra are 

 represeuted by the images of the spectrometer slit for two wave- 

 lengths of light, those corresponding to the red and greenish-blue 

 hydrogen lines. The spectra were produced by three prisms of the 

 respective salts, of precisely the same angle, near 60°. You see that 

 the two upper spectra, representing the potassium salt, are the furthest 

 apart, and that the separation diminishes for the rubidium salt, and 

 it becomes so much less in the case of the csBsium salt that the two 

 lower spectra formed by this salt partially overlap. 



This relative behaviour as regards double refraction is quite 

 general throughout all the four series of salts. 



The point can be further illustrated by the curves which you now 

 see projected on the screen. The two outer curves represent the 

 maximum and minimum refraction, and the inner one the inter- 

 mediate refraction along the third rectangular axis of the ellipsoid. 

 You observe that the two outer curves converge towards each other 

 as the atomic weight of the alkali metal increases. 



This leads in four particular cases to perhaps the most interesting 



