1!»4 F. K. Wright — Transmission of Light through 



observations, differences of a degree or more being not uncom- 

 mon. The rotatory effects of the glass mounts and lenses are 

 included in these results, and indicate that in actual micro- 

 scopic work with tilted slides or with interference figures, the 

 observed azimuths of the planes of polarization of any given 

 transmitted wave may differ several degrees from the actual 

 azimuth of the refracted ware within the crystal. This, how- 

 ever, is the azimuth sought for in such observations. The 

 measurements prove that for such tilted thin plates there is no 

 position of total extinction, but rather a region of minimum 

 illumination. Methods based on such phenomena can there- 

 fore furnish only approximate results. 



On the horizontal line 10 of Table II, the values obtained by 

 graphical construction of the azimuths of the planes' D (fig. 12) 

 for different positions of the plate are listed. These will be 

 considered in a later section. 



Nephelite. 



In nephelite the birefringence is weak and the two refracted 

 waves, resulting from a single plane polarized incident wave, 

 follow approximate^' the same direction of propagation. A 

 satisfactory determination of the uniradial azimuths on thin 

 plates of nephelite is, in consequence, not possihle, and has not 

 been attempted. The positions of maximum extinction of the 

 plate tilted at different angles were, however, determined and 

 are listed in Table III. For these measurements a polished 

 plate of nephelite, so cut that the optic axis included an angle of 

 27° 18' with the normal to the plate, was mounted (« Na of object 

 glass 1*511, of cover slip, 1*520) in Canada balsam (n Na = 1*537) 

 and readings made in strong sodium light. On the horizontal 

 lines, 1 to ±, of Table III, the calculated values of r, 8, e and &' 

 are listed ; on line 5 the positions of extinction are recorded, 

 as determined by direct observations on rotating the plate 

 until the maximum darkness was attained. Each angle given 

 is the average of 20 readings, 10 for +i, +co, and 10 for 

 + i, —&). The individual settings agreed only fairly well, 

 and even for the average values a high order of accuracy can- 

 not be claimed. For these measurements the ordinary micro- 

 scope lens system was used with the exception of the condensor, 

 which had been removed. On line No. 6, Table III, the results 

 of settings made with the aid of the bi-quartz wedge plate and 

 strong sodium light are recorded, each angle listed being the 

 average of 10 readings, 5 for the position + i, + &>, and 5 for 

 the position + i, — co. Since in the case of tilted plates the 

 emergent light is not strictly plane polarized and there is con- 

 sequently no position of total extinction, the positions of mini- 

 mum illumination as determined by simple rotation under 



