TN QUARTZ (3F LIGHT IN THE VISIBLE REGION OF THE SPECTRU.\r. 
281 
interference fringes, were then seen, some of the specimens showing two series of bands 
running in directions inclined to one another. 
As showing the extreme dilEculty of procuring satisfactory material it may be 
mentioned that a pair of plates of dextro- and Imvo-quartz to take the place of the 
defective cylinders was procured from a Continental firm under a guarantee that the 
specimens should be absolutely free from flaw when tested in the most stringent way ; 
although each plate had an area of 25 sq. cm., the utmost that could possibly have 
been done with them would have been to drill out a single cylinder from one plate 
and possibly four or five of smaller diameter from the other, the whole of the rest 
of the material being shown to he grossly defective when viewed in the way described 
above. 
Finally, Messrs. A. Hilger and Co. were able to provide a plate of dextro-quartz, 
51 mm. in thickness after grinding and polishing, which showed some feathery patches 
and also groups of acicular crystals running through the block ; but by careful 
examination, it was possible to pick out four 15 mm. circles which showed no flaw of 
any description. Two of these, unfortunately, were splintered in drilling and had to 
he reduced to about 40 mm. in length. Each of the four cylinders satisfied the 
following tests :— 
(1) No flaw or irregularity could be seen on viewing the crystal between crossed 
Nicols with monochromatic green light. 
(2) No differences could be detected between the optical rotations with posjtNe 
and negative half-shadow angles ; of fifty pairs of readings, taken without special 
precautions, only two pairs showed a difference greater than 0‘05° and these were not 
cpnfirmed on repetition. 
(3) No differences greater than 0°'05 were found when the cylinders were rotated 
about their axes, each cylinder being set with the reflecting eye-piece and tested in 
six different positions. 
(4) No differences could be detected in the rotatory powers of the four cylinders 
beyond those which might be ascribed to variations of temperature due to imperfect 
jacketing. The actual values at about 20° were :— 
Ki . . 25-5366, K, . . 25-5369, . . 25-5368, E, . . 25°-5365 per mm. 
In order to increase the lengch of the quartz column, which had been reduced some 
20 mm. by breakages, a bridge-piece. Eg, separating two of the cylinders in the block, 
was tested and cut. Examined in three positions, with positive and negative half¬ 
shadow angles, it showed an extreme variation of only 0°-03. But the rotation per 
millimetre was higher, at 25°-5378, a relative difference that was confirmed after 
regrinding; the piece (which had a clear aperture of only about 8 mm. instead of 
15 mm.) was not used, therefore, in the final series of observations. 
As the dextro-quartz, although withstanding all the new tests that had been 
devised during four years of continuous work, had been cut from the immediate 
2 o 
VOL. CCXII.—-A. 
