354 Prof. D. B. Brace on Achromatic Polarization 



selenite of the thirty-third order cut with its edge parallel to 

 one of the principal axes, one face being in the plane of 

 principal cleavage. Adjustment was made so as to give 

 three bands in the field. This gave almost perfect coincidence 

 of the two sets of bands. On increasing the thickness the 

 approximate coincidence between the alternate bands was 

 obtained by five passages of the sodium line, thus showing 

 the original order to be the fifth. On increasing the thick- 

 ness of the other wedge so as to give five passages the approxi- 

 mate coincidence of the bands was again obtained, those of 

 the quartz being shorter than those of the selenite in the 

 blue. By continuing this process of increasing the order of 

 each up to the thirty-second (the limit of the selenite wedge) 

 this shortening of the quartz in the blue increased, the red 

 bands extending beyond those of the selenite. On intro- 

 ducing a plate of selenite before the wedge and adjusting for 

 coincidence, and then increasing the orders successively up to 

 the sixtieth, it was found that a better coincidence in the 

 yellow and toward the blue was obtained on increasing the 

 order of the selenite by one at the forty-ninth order, making 

 the order fifty for selenite and forty-nine for quartz ; this, 

 however, increased the disagreement in the red still more. 

 Achromatism was apparently not possible except in parts of 

 the spectrum, the ratio depending on the portion of the 

 spectrum, the reciprocal of this ratio being even possible for 

 the red, say. No anomalous distribution of the bands of 

 selenite could be detected in the region studied by Mouton 

 and Dufet, namely '491 /a, where the bands of the selenite 

 could be made to coincide with those of the quartz. 



Comparison of Mica and Left-handed Quartz. 



Considerable difficulty was experienced in obtaining a 

 wedge of mica. One of the twenty-eighth order was, how- 

 ever, finally ground by closely cementing together two or 

 three pieces upon a plate of glass and careful polishing. The 

 angle was 6° 1' ', the edge being parallel to one of the principal 

 axes and one face that of principal cleavage. High orders 

 could not be used on account of the great absorption of light. 

 The quartz wedge was the same as previously used. 



Starting with four bands visible in the field, which was as 

 low as comparisons could be made, and increasing the thick- 

 ness of each to one passage of the sodium line, the best coin- 

 cidence for the central band was obtained with orders of eight 

 quartz and nine mica, the orders being determined as above 

 by doubling the number of bands in one spectrum. The 



