488 



Professor Silcanus P. lliompson 



[May 17, 



rotates the light — say a tube with sugar solution in it. At once 

 the balance is upset, and I must, in order to get equality of illumi- 

 nation, turn my analyser through an angle equal to that of the 

 optical torsion. 



Of the same class are the polarimeters with special prisms made 

 in two parts slightly inclined to one another. 1 he earliest of these 

 was devised by the late Professor Jellett, of Dublin, and has been 

 followed by imitations of the same plan by Cornu, by Lippich, and 

 by Schmidt and Haensch. The beautiful " shadow polarimeter," by 

 the latter firm, which I here exhibit, has the divided prism, and a 

 quartz compensator. 



I have suggested two simpler methods of accomplishiug the same 

 end. In the first place, I have proposed to use twin-prisms. These 

 are made on a plan suggested to me by finding that Mr. Ahrens's 

 method of cutting calc-spar for prisms was admirably adapted for 

 making such prisms, either with wide or narrow angles between the 

 respective j^lancs of polarisation in the two parts of the visible field. 

 Two such twin-i)risms, one with 90°, the other with 2J°, between the 

 prisms, are here on the table. In the second place, I have essayed a 

 polarimeter, an example of which is before you, in which an arrange- 

 ment of twin-mirrors (each set at the polarising angle, but slightly 

 inclined to one another) is made to yield a half-shadow effect. 



Before I leave the subject of quartz I must refer to the famous 

 mathematical theory of Fresnel, who endeavoured to explain its action 



Fig. 11. 



Model illustrating recomposition of rectilinear motion irom two opposite circular 



motions. 



upon light by supposing that the jDlane-polariscd wave on entering it 

 is split into two waves, consisting of 02)positely circularly-polarised 

 light, which traverse the crystal with different speeds. On emerging 

 they recombine to form plane-polarised light, the jDlane of which, how- 

 ever, depends on the retardation of phase between the two components. 

 I here introduce a mechanical model to illustrate one of the points in 

 this theory — namely, the recombination of two circular motions to 

 form a straight-line motion. These two disks (Fig. 11), which turn 

 in opposite senses, but at equal rates, rei^resent two circularly-polarised 

 beams of light. The linkages, which connect two pins on these disks, 

 compound their motions at the central point, P, which executes, as 



