COLORS m CRYSTALLINE PLATES. I33 



When the principal plauc of the prism coincides with the plane of polarization 

 of the light, and the azimuth of the lamina is varied, the maximum brilliancy 

 of coloring is found in the azimuth of 45° between the principal section of the 

 lamina and the plane of polarization. When the azimuth is 0° or 90°, the 

 color entirely vanishes, and the light appears entirely unchanged. At inter- 

 mediate azimuths the color has an intermediate intensity, regularly increasing 

 and diminishing between the positions of minimum and maximum. These 

 variations, as well as the thickness of the laraina3 themselves in Avhich the phe- 

 nomena appeared, satisfied Mr. Arago that the colors could not be owing to the 

 same causes which produce the colors of NcAvton's rings. Still they had evi- 

 dently some relation to the thickness; for it was not difficult to remove them 

 entirely, either by considerably increasing the thickness or by excessively 

 diminishing it. In the rotation of the lamina as just described, the colors 

 which appeared between the successive positions of minimum were always the 

 same in the same image. But when the lamina itself remained fixed, while the 

 prism at the eye was rotated in azimuth, the two images interchanged their 

 colors in passing each successive position of minimum. 



If, instead of a doubly refracting prism as an eye-piece, a mirror, presented 

 to the ray at the polarizing angle, be employed, only one of the images is re- 

 flected; but the other, if the mirror be transparent, will be seen in the light 

 transmitted. In consequence of this separation of effects, Mr. Arago was led 

 to distinguish the mirror xA\cn lased in this way as the analyzer. In observa- 

 tions with the analyzer, then, it appears that when the lamina is rotated in azi- 

 muth, the same colors come and go in the successive quadrants ; but when the 

 analyzer itself is rotated, the colors in the alternate quadrants are comple- 

 mentary to each other. 



The colors thus seen in crystalline lamince recur in several successive series, 

 as the thickness of the lamince is increased. Accordingly, if in a ])late of sele- 

 nite we hollow out a spherical cavity of very large radius, Ave shall find it to 

 exhibit several orders of rings resembling those of NcAvton, and folloAving the 

 same laws; though the thicknesses at Avhich the colors of the same order occur 

 are A^ery much greater. According to the determinations of Biot, the compara- 

 ti\^e thicknesses at Avhich the same colors appear in air, in Iceland spar, in 

 quartz, in selenite, and in Siberian mica, are as the numbers 1, 13, 230, 230, and 

 440; the thickness for selenite and quartz being sensibly the same. 



The limits of absolute thickness below Avhich crystalline plates fail to give 

 colors in polarized light, are, for selenite, 0.017 in.; for mica, 0.0323 in.; and for 

 Iceland spar, 0.001 in. The maximum thickness for this last cr3'stal is but six 

 or seven one-thousandths of an inch. Jllica and selenite are therefore prepared 

 Avith facility for this class of chromatic experiments ; but this is not equally 

 true of Iceland spar. If a lamina of selenite — a mineral Avhich is very easily 

 Avrought — be secured by transparent cement of any kind to a plate of glass, 

 A^ery fanciful effects may be produced by grinding it away unequally in different 

 parts according to any definite pattern. Figures of various kinds, images of 

 insects, floAvers, gothic AvindoAvs, &c., &c., may thus be prepared, which Avill 

 come out in polarized light in very brilliant colors. 



When laminas are presented obliquely to the polarized ray and the inclina- 

 tion A'aried, the colors change Avith the obliquity; sometimes ascending in order 

 Avith an increase and sometimes Avith a decrease of obliquity, according to the 

 character of the crystal and the direction in Avhich the lamina has been taken 

 from it. For these experiments it is best to cut the laminse parallel to the 

 optic axes of the crystals. 



If tAvo lamina;, either or both of Avhich exceed the limits at which colors arc; 

 seen, but Avhose difference of thickness is Avilhin those limits, be placed one upon 

 the other Avith their principal sections crossed — that is to say, placed at right 

 angles to each other — colors Aviil be seen corresponding to those of single 



