20 On a simple Form of Saccharimeter* 



emerges with the planes of polarization of its two halves slightly 

 inclined to each other. It is of course always necessary to use 

 homogeneous light to avoid dispersion. 



Mr. Glazebrook has very kindly given me the following 

 numbers, which are taken at random from a large number of 

 sets of readings he has obtained for the electromagnetic rota- 

 tion of certain solutions of NaCl in water ; the difference of 

 thickness of the two plates being *1 mm., and the inclination 

 of the planes of polarization being therefore about 2° for 

 the sodium-light used. The circle to which the analyzer 

 was attached reads to 3'; but the vernier divisions can easily 

 be further subdivided by eye. 



Circle-readings. 



I. Current direct. 



23 45 





23 46 





23 45 





23 45 



Current reversed. 



21 



36 





21 



34 





21 



39 



II. Current direct. 



23 



15 





23 



16 





23 



18 



Current reversed. 



22 



19 





22 



19 





22 



20 



[II. Current direct. 



23 30 





23 



30 





23 



28 



Current reversed. 



21 45 





21 



47 





21 



48 



In order to vary the inclination of the two planes of polari- 

 zation to each other, one of the halves of the quartz plate 

 might be arranged like a Babinet's compensator, so that the 

 difference of the two might be varied at will. The chief ob- 

 jection to the method seems to be that the quartz plate has to 

 be adjusted very exactly perpendicular to the axis of the 

 pencil. 



A still simpler arrangement, which has yet only been tried 

 in a somewhat rough form, consists in a cell containing some 



