262 SENARMONT ON THE OPTICAL CHARACTERS OF 



means of some one of the processes of compensation for which 

 science is indebted to M. Biot and Sir D. Brewster. 



The rings in convergent polarized hght were observed upon a 

 plate cut perpendicular to the axis, and I examined what modifi- 

 cations were effected in them by the interposition of a plate of 

 mica in which the axis of minimum elasticity had previously 

 been determined, and the thickness of which produced a retard- 

 ation amounting to half an undulation. 



Sometimes the great breadth of the rings rendered this mode 

 of proceeding inapplicable, and I then placed a plate of gypsum 

 giving a sensible colour, and the principal sections of which 

 formed an angle of exactly 45° with the plane of polarization, in 

 a bundle of parallel rays. No change takes place when the cry- 

 stalline plate is interposed perpendicularly to the path of the 

 rays issuing from this plate of gypsum, to the axis of which it is 

 likewise perpendicular ; but when it is inclined, so that the angle 

 of incidence is parallel to one of the principal sections of the 

 lamina of gypsum, the colour of the latter is altered in an inverse 

 manner, according as the optical axis of the plate is that of maxi- 

 mum or minimum elasticity. 



In order to determine the character of the optical axis in a 

 plate parallel to that axis, it was placed upon a plate of quartz 

 of convenient thickness, and cut parallel to its axis. This ar- 

 rangement, placed perpendicularly in a bundle of parallel polar- 

 ized rays, so that the axis of the quartz forms an angle of 45° 

 with the plane of polarization, developes colours when the axes 

 of the crystals, the one of maximum and the other of minimum 

 elasticity, are parallel, or when both the axes of minimum elasti- 

 city are perpendicular. 



It has long since been shown by Brewster, that in crystals 

 referable to a right rhombic prism, the directions of the three 

 axes of optical elasticity are parallel to the height of the prism 

 and to the two diagonals of its base. When therefore the iso- 

 chromatic curves are observed in convergent polarized Hght, it 

 is easy to recognise at the first glance the position of the plane 

 of the optical axes and of the bisecting line. 



The perfect determination of the three indices requires opera- 

 tions which can rarely be executed with precision on the more 

 or less imperfect crystals that were at my disposal ; I have there- 



