Mr. R. T. Glazebrook on NicoVs Prism. 247 



would reach b during the intervals between the discharges 

 emitted by b. But as, according to all experimental evidence, 

 these intervals have even greater duration than the discharges 

 themselves, the whole light from a would gain access each 

 time to the dark surface, and the dark surface would, contrary 

 to experience, vanish altogether. 



This result, combined with that of the spectroscopic experi- 

 ment, shows again the improbability of the convective action in 

 the discharge, against which hypothesis I have already adduced 

 experimental evidence in the paper laid before the Academy 

 in the preceding year. 



Berlin, Oct. 20th, 1879. 



XXXIV. Notes on NicoVs Prism. By R. T. Glazebrook, 

 M.A.,Felloic of Trinity College, and Demonstrator of Physics 

 in the Cavendish Laboratory, Cambridge*. 



IN using a NieoPs prism to measure the rotation of the 

 plane of polarization of light, two adjustments are neces- 

 sary : (1) the incident light should be parallel to an edge of 

 the rhomb of spar which constitutes the prism ; (2) the axis 

 of rotation of the prism should also be parallel to this edge. 



I have lately found it requisite to discuss the effects that 

 errors in these adjustments would produce in the results of 

 some experiments I have been engaged in; and it appeared 

 that the discussion may be useful to others working with a 

 Nicol's prism. 



Let A B G D E F G H be the prism, A B C D being the face 



of incidence. Let the incident-wave normal be parallel to 

 A E, and suppose, for simplicity, that A B C D is a rhombus. 



* Communicated by the Author. 



