1902.] on the Constitution of Crystals. 145 



dicular to the line of propagation. Consequently in these two 

 directions, which are termed optic axes, the crystal will exhibit no 

 double refraction. When the Y velocity approaches nearer to the X 

 velocity the inclination of the two circular sections to each other 

 naturally diminishes, and the angle between the optic axes becomes 

 proportionately less, until at length we have equality of X and Y, 

 the ellipsoid becoming a spheroid, with a single optic axis. This 

 is the special case which we observe in tetragonal and hexagonal 

 crystals, such as the well-known cases of calcite and quartz. 



We will now see some very beautiful phenomena in connection 

 with these optic axes, employing for the purpose a projection polari- 

 scope, which is furnished with a special set of lenses to render the 

 light strongly convergent while passing through the crystal. First 

 let us investigate the phenomena exhibited by one of the uniaxial 

 crystals we have just referred to. A section-plate cut perpendicular 

 to the single optic axis is now between the crossed Nicols, and you 

 see the beautiful spectrum rings and the black cross, characteristic of 

 uniaxial crystals, which are produced. 



We will next see the effect produced by a section-plate of a biaxial 

 crystal, similar to the salts we are discussing. The plate is cut per- 

 pendicular to that axis of the ellipsoid which is the bisectrix of the 

 acute angle between the optic axes. You observe the loci of the two 

 optic axes are marked by hyperbolic brushes, in the present position 

 of the section, and they are surrounded by separate rainbow-coloured 

 rings, which in turn are surrounded by lemniscates and eventually 

 ellipse-like curves. If we rotate the section 45°, the hyperbolae join 

 up to form a cross, but the loci of the optic axes remain marked by 

 the rings. These rings are large and brilliant, to render the demon- 

 stration clear, and are afforded by a very large section-plate. But 

 for research purposes we require the rings to be very small and the 

 hyperbolae very narrow, so that the measurement of their position 

 may be very accurate. 



We will put in another section-plate, much smaller because of the 

 impossibility of obtaining larger ones, of rubidium magnesium sele- 

 nate, and you see how small and sharp, although naturally fainter, 

 the rings and hyperbolae are. 



Next let us demonstrate how we measure the angle of separation 

 between the two optic axes. Another section-plate has been arranged 

 on a small goniometer, so that it can be rotated in the plane of the 

 axes. You see we are able to bring first one and then the other optic 

 axis up to the cross wires, which you see also focussed on the screen. 



If we note the reading of the goniometer circle while one optic 

 axis is so adjusted, and then rotate until the other is in position and 

 read the circle again, the difference between the two readings will 

 give us the apparent angle between the axes as seen in air. 



In order to arrive at the true angle within the crystal, it is neces- 

 sary to cut another section perpendicular to the bisectrix of the obtuse 



Vol. XVII. (No. 96.) l 



