352 Mr. E. T* Glazebrook on 



probably varies as the pure resistance varies, and further 

 remembering that an opposing electromotive force which has 

 no existence apart from a combined resistance acts m any- 

 electrical test exactly as a resistance, it must be always very 

 difficult experimentally to separate them. All of course that we 

 can measure electrically is the difference of potential between 

 the carbons and the current passing between them ; and this 

 is what we have been measuring all through these two inves- 

 tigations. 



It may be here noted that in all probability the conduction 

 from particle to particle in a microphone is of the nature of a 

 small electric arc, or, rather, perhaps a convective discharge, 

 seeing that the resistance in a microphone varies with the 

 current used to measure it ; indeed it is probable, when the 

 pieces of carbon or other material employed, are so pressed 

 together that close intimacy of contact of the particles makes 

 the resistance tolerably independent of the current, that then 

 the pieces of carbon will not act as a microphone at all. 



We have to thank Messrs. W. Atkinson and Lt. B. Atkinson, 

 two of our students, for much assistance rendered us in these 

 experiments. 



LT. On Polarizing Prisms. By R. T. Glazebrook, M.A., 

 P.R.S., Fellow and Lecturer of Trinity College, Demonstrator 

 in the Cavendish Laboratory, Cambridge *. 



IN a paper on Nicol's prism (Phil. Mag. vol. x. 1880) 

 I have considered some of the defects of Nicol's prism 

 as a means of producing plane-polarized light. In the present 

 paper I propose to describe a form of polarizing prism free 

 from most of these. For many purposes, one of the great 

 objections to Nicol's prism is the lateral displacement pro- 

 duced by it in the image of an object viewed through it. If 

 we place a Nicol before the object-glass of a telescope, on 

 turning the Nicol round its axis the image moves across the 

 field. This has been remedied somewhat by cutting prisms 

 with their ends at right angles to their length, and making 

 the angle between the normal to the face on which the incident 

 light falls and the plane of Canada balsam such that the 

 •ordinary ray is totally reflected there while the extraordinary 

 ray passes through. But this is not entirely successful; for let 

 A B C D (fig. 1) be a section of such a prism by a plane parallel 

 to the edge B C and at right angles to the Canada balsam. 



* Communicated by the Physical Society of London; read 14th April, 



1883. 



