in the Hufner Type of Spectrophotometer. 485 



in the plane of incidence is (see Schuster's ' Theory of 

 Optics/ first edition, p. 49) 



sin^-tf) 

 2Cl sin 2 (£i + 0)' 



The intensity of the transmitted light if that light be polarized 

 in the plane of incidence will then be 



- [_ sin-(6' 1 + tf)j 



As in a prism at the angle of minimum deviation the angle 

 of incidence on entering the prism is equnl to that of re- 

 fraction on leaving it, the intensity of the transmitted light 

 if that light be polarized in the plane of incidence, is, after 

 passing right through the prism, 



f sin'(gi-fl) 



For the case under consideration this works out to ^-x*651* 



The polarization due to the prism in the plane at right 

 angles to this will bo found, when evaluated, to be negligible* 



When, then, a spectrum produced by such a prism is 

 observed through a polarizing prism, and the latter is 

 rotated, the beam varies in intensity from 1 to '651 (for 

 the particular wave-length for which the yu, of the prism 

 = 1-658). 



It will be seen how extremely serious is this polarizing 

 effect of the dispersion prism. Want of leisure has pre- 

 vented me from reading up fully the bibliography of the 

 subject, but although this error has been previously mentioned 

 and seems fairly well known, there certainly seems to have 

 been insufficient attention brought to bear on the subject, as 

 there are still polarization photometers made which suffer 

 from the full amount of this error. 



The Hiifner instrument even in its original form is to a 

 very great extent protected from this failing by the presence 

 of the so-called Hufner rhomb in front of the slit. It will be 

 seen at once that this produces a partial polarization of the 

 upper beam of light in a plane perpendicular to that of the 

 partial polarization produced by the dispersion prism. By 

 applying Fresnel's equations as in the case of the dispersion 

 prism above, we can find the amount of this partial polarization. 

 The rhombs in the first instrument made were of Jena boro- 

 silicate crown of md = . 1*517, the angle of the front and back 



