352 Prof. D. B. Brace on Achromatic Polarization 



equation of condition (2 a) relates only to order and wave- 

 length, hence no attention need be given to the determination 



Fig.l. 



of d l and d 2 in observing whether achromatism exists. A 

 motion of the micrometer-screw of each compensator before 

 the slit caused the bands to move across the spectrum from 

 the violet to the red end, more coming in at the violet than 

 disappeared at the red end, thus causing the bands to crowd 

 together, while when the screw was moved so as to diminish 

 the thickness, more bands disappeared at the violet end than 

 came into the field at the red end, thus causing a spreading 

 out of the bands, which confirms the conclusions above 

 arrived at. When a prism was used at R the distance 

 between the bands increased from the red toward the violet 

 owing to the greater dispersion of the glass for this part of the 

 spectrum than the differential double refraction of the crystals. 

 It was found that for the values of d used the order of the 

 plate could be determined to a certainty according to the 

 method indicated under equation (15). This obviated the 

 difficulty hitherto met in determining the order of the plate 

 by measuring its thickness and index, which may introduce 

 serious errors, as indicated above in the discrepancies between 

 the results of Mouton and Dufet. It furthermore did not 

 require the use of wedges perfectly plane, which it is not 

 always possible to obtain, as, for example, in the case of mica 

 and other easily cleavable crystals, accurate data for which 

 are lacking. 



Plane parallel plates and wedges of mica, selenite, right 

 and left-handed quartz, Iceland spar, and aragonite were 



