LANTERN POLARISING APPARATUS 341 



reflected rays to be used as well as the transmitted ; and 



thus one image may be thrown upon the screen, and the 



complementary image at the same time upon the ceiling, if 



there happen to be one. Thin micro-glass is sold at per 



ounce in \\ x J- size, and twenty-four such plates, selected 



for flatness, make the best analyser, and will answer very well 



for a home-made instrument, both images being surprisingly 



good if the glass is flat. The 



plates may either be mounted 



in a round tube, by cutting a 



cork which fits the tube 



obliquely at the proper angle, u FIQ 188 



and cutting the glasses into 



ovals round a shape made to the cork section, after which 



the cork is hollowed through longitudinally into a tube ; 



or rectangular glasses may be fitted into a square tube, with 



a round nozzle at the end. 



A double-image prism should also be provided, unless two 

 are combined in a Huygens apparatus, when one of these 

 will of course suffice. 



195. The Nicol Prism Polariscope. The elbow in- 

 strument has two inconveniences : the polariser is fixed ; and 

 the lantern has to be de- 

 flected, which is awkward, 

 more especially if diagrams 

 are required between the ex- 

 periments. The most perfect Flo 189 

 instrument is unquestionably 



that in which the polariser, as well as analyser, consists of a 

 Nicol prism. It was formerly the custom to have two prisms 

 of nearly equal size for polariser and analyser, and such an 

 arrangement is still employed at the Eoyal Institution. As I 

 long ago pointed out, however, hi all ordinary experiments we 

 have to focus the rays by some sort of objective ; and it is mani- 

 fest from fig. 189 that just as many rays get through a smaD 



