1889.' 



on Optical Torque. 



481 



First, I have made the end faces principal planes of section (Fig. 5) ; 

 secondly I have made the axis of vision cross the crystallograjDhic 

 axis at right angles, so getting a flatter field, a shorter length, a 

 wider angle, and less loss of light by reflection. Mr. Ahrens, the 

 prism-cutter, on whose able assistance I have relied during the last 

 six or seven years in cutting these prisms, has aided me with his 

 ingenuity in devising a method of cutting up the spar so as to give 

 these advantages with a minimum waste of material. He has farther 

 devised a method of putting a polarising prism together in three 



Fig. .5. 



Fig. 6. 



S. P. Thompson's modification of the Nicol prism. Ahvens's triple prism. 



instead of two pieces — illustrated in the diagram (Fig. 6) — which 

 gives a still wider angle. The prism which I shall use as analyser 

 in the next ex|)eriments is one of these forms. 



Unfortunately at present there is a spar-famine, pieces of Iceland 

 spar of a size and purity suitable for the making of large polarisers 

 such as that I employ being not now procurable at any price. To 

 avoid the excessive cost of large Nicols I have lately got Mr. Ahrens 

 fco construct for me a large reflection-polariser, on the plan of 

 Delezenne, but modified by Mr. Ahrens in detail. In this prism the 

 light is first turned to the proper polarising angle by a large total- 

 •eflection prism of glass, and then reflected back, parallel to its 

 )riginal path, by impinging upon a mirror of black glass covered by 

 I single sheet of the thinnest patent plate glass to increase the 

 ntensity of the light. This form of polariser, depicted in Fig. 7, is 

 [uite equal for projection purposes to a Nicol prism of equal aperture, 

 Lud is much less costly. This one has 2| inches clear aj^erture. 



Having so far reviewed the apparatus for polarising and analysing 

 '. will return to the apparatus set with its prisms crossed, so that the 

 nalyser completely extinguishes the polarised light emitted from the 

 •olariser. 



If in the space between polariser and analyser anything be intro- 

 uced which can either resolve obliquely the polarised vibrations or 

 vist them bodily round, then there will not be complete extinction ; 

 le amount of light passing the analyser depending in the one case 

 Q the obliquity of the resolution, in the other upon the degree to 

 hich the vibrations are twisted or rotated upon themselves. 



