528 Prof. R. W. Wood on Separation of Close 



quartz should be used, as the crystals from Madagascar show 

 irregularities when examined by polarized light. 



A large and very clear crystal was selected and examined 

 between crossed nicols with sodium light. There appeared 

 to be no internal irregularities of a nature such as were 

 exhibited by a beautiful block of quartz loaned to me bv 

 Mr. Twymau, which was possibly cut from a Madagascar 

 crystal. In this block the D x and D 2 bands, instead of 

 appearing uniformly parallel, were deformed at one point 

 in sharp zig-zags ; no trace of anything could be seen bv 

 unpolarized light. It is not always possible to judge a 

 crystal before it is cut, but if the faces are reasonably plane- 

 and clear, it is usually possible to tell whether variations in 

 the intensity of the sodium light result from internal troubles, 

 or from small differences in thickness. From the selected 

 crystal a block measuring 85 mm. x 60 mm. x 32 mm. was- 

 cut parallel to the axis and polished *. 



As the degree to which the plate needs to be plane-parallel 

 can easily be attained by the use of the spherometer, and as 

 the side faces do not have to be accurately parallel to the 

 axis, the preparation of the plate presents no great difficulty.. 



It was examined with sodium light between crossed nicols 

 before the polishing stage was reached, and found to give 

 absolutely uniform illumination, which is what is required- 

 Glass plates and benzine were of course used to render the 

 block transparent. M. Bertin increased the uniformity of 

 thickness considerably even after this test, and then polished 

 the six surfaces of the block. This block when used for 

 separating the D lines (32 mm. thickness) has an area of 

 50 square cms. Used in the other position, 60 mm. thick- 

 ness, it will separate lines 3 Angstrom units apart. Used 

 " end-on " its property of natural rotation can be utilized in 

 certain experiments. 



The double refraction of Iceland spar is very much greater 

 than that of quartz, a plate less than 3 mm. thick being re- 

 quired for the D lines. It would, however, have to be plane- 

 parallel to a degree 10 times as great as is the case with 

 quartz. A thick block of spar, made plane-parallel by the 

 methods in use in the construction of modern interferometers,, 

 could very likely be used for suppressing the strong central 

 components of multiple lines, when photographing or ex- 

 amining the fainter satellites. Experiments in this direction 

 are now in progress. 



* The work was done by M; Bertin, 13 quai St. Michel, Paris, whom 

 I have instructed as to the methods of testing by polarized sodium light. 

 The price of the crystal, including cutting- and polishing, was 140 francs- 



