336 POLARISATION. 



a diagonal one), then we obtain as the combined effect of both for 

 the interval of acceleration a colour of the second order (yellow 

 II.), and for the interval of retardation a colour of the first order 

 (light blue I.), which in regard to the yellow is one grade lower 

 than red. 



We shall presently return to these acceleration and retardation 

 colours, and tabulate them for a series of combinations as they 

 commonly occur in practice ; here it is only a question of estimat- 

 ing them for the determination of the relative magnitudes of the 

 axes of elasticity, and for this purpose the facts we have mentioned 

 amply suffice. It is obvious that the comparison of any given 

 medium with a compressed glass plate, whose ellipse of elasticity is 

 known, must present a simple means for correctly explaining the 

 unknown ellipsoid of elasticity. We will denote its three axes, 

 whose directions we assume to be known, by a, I, and c ; then it is 

 only necessary to combine the sectional surface passing through a b 

 with the glass plate : the position in which acceleration or retarda- 

 tion takes place then decides whether a or b is the greater. In like 

 manner we determine in sectional surfaces, which are cut parallel 

 to b c, the ratio of b to c, and in others cut through a c, that of 

 a to c. Thus, as nearly as possible according to this method, our 

 problem is solved, we know what directions correspond to the 

 least, the mean, and the greatest elasticity. 



The examples already adduced will serve here for further 

 elucidation. If the tabular crystal of selenite (Fig. 192) is so 

 placed upon the compressed glass plate, that the 

 direction a a is parallel to the major axis of elas- 

 ticity in the glass, we observe an increase of the 

 interference colour the effects of the two media 

 are added together. On the other hand, if we 

 make a rotation of 90 retardation sets in, and 

 the interference tint is lowered. Hence of the 

 two axes of the ellipsoid, which are parallel to 

 the directions a a and c c, the latter is the shorter. 

 The discovery of the third perpendicular axis is 

 FIG. 192. combined with some difficulties. From the un- 

 symmetrical position of the axes a a and c c it is 

 evident that the crystal is biaxial, which excludes the possibility 

 of the third axis being equal to one of those already determined ; 

 y.et we have always to discover whether it is the major, the minor, 



