830 



Mr. J. C. McConnel. 

 Fie. 5. 



[Mar. 1-2, 



It will be noticed that the directions of the optic axis in different 

 parts form a series of straight lines. This is an immediate conse- 

 quence of the hypothesis of the existence of sliding surfaces, and 

 may be shown in the following way. In the part of the crystal 

 beyond the dotted line, however, this rule does not hold good. 



In the original unstrained crystal the optic axis is in the same 

 direction everywhere. Hence layers perpendicular to it are of equal 

 thickness throughout. Their subsequent bending and slipping does 

 not affect their uniformity of thickness. 



We need only consider one of the principal directions of 

 curvature. 



Draw PP', QQ' normal to the surface at P, Q, to meet the next 

 surface at P'Q'. On PP' drop perpendiculars Qn, Q'n'. All the quan- 

 tities in small distances are small except the radii of curvature p, p' 

 at P and P'. Since the thickness of the layer is uniform, PP' = QQ' 

 = nn'. Thus to the second order of small quantities Pn = P'n'. 



FIG. 6. 



But since Pn is normal to the curve PQ at P, Pn = Qw 2 /^/' = 

 QV-/2// to the second order. Now this would have been the 



