Wl EDWARD M. NELSON ON A NEW FORM OF POLARISER. 



placed close to it was capable of rotation, then the rotation 

 of that plate would give the same efiect as the rotation of a 

 Nicol. Now as it is difficult to cut 1/4 wave-length plates of 

 a large size, the following is a means of reducing the beam so 

 that it can pass through a disc of small diameter, and then be 

 parallelised to full size upon emergence. B are the 1/4 wave- 

 length plates. Please note particularly that for all ordinary- 

 microscopical work the rotation of the polariser is not essential, 

 therefore the two lenses and the 1/4 wave-plates may be omitted. 

 A useful size for a microscope would be : hole 1-5 in. diameter ; 

 d - 1-75 in. ; MMi = 1-82 in. ; c = 0-5 in. ; G = 3-33 in. (which 

 may with advantage be increased to 3-5 in.) ; length of box 

 6-5 in. ; width 1*75 in. ; depth 4-0 in. The box should have a 

 slide at G so that the plates and mirrors can be readily got at 

 for cleaning. The above are inside measurements. Polarisation 

 by means of transmitted light through a pile of plates has been 

 tried. The results were exceedingly poor, in fact quite worthless 

 when compared to those obtained by reflection, which, when 

 20 or 24 plates were used, were practically as good as by a NicoL 

 The percentage of polarised light given by any number of 

 plates can be determined by the following formula : x = 1*2135 -j- 

 2-20752/ + 3-5245?/2 — 1-9237?/^ + 0-2702?/* ; ?/ = ^^ ; P = cc^ ; 

 n is the number of plates, and P the percentage of polarised 

 light. Example, let w = 8, then y = 2, x=^ 8-66 and P = 75-0. 

 So we see that with 8 plates three-quarters of the incident light 

 will be polarised by reflection at the proper angle. 



Journ. Quekett Microscopical Club, Ser. 2, Vol. XIV., No. 84, April 1919. 



