The diallage has been studied in some detail, a large 

 crystal in the gabbro-porpbyry affording good material. It 

 was brown in colour and of a silky lustre. Partial chemical 

 analyses showed the presence of *33 per cent, of chromium 

 Besquiozide and '14 per cent, of oxides of nickel and cobalt 

 though the lower figures obtained in the analyses of the 

 diallagite ('24 and a trace respectively) show that the 

 diallage cannot be considered of constant composition in 

 regard to these minor constituents. This quantitative 

 work confirms the previous determination, by borax bead 

 only, of the presence of chromium in the diallage. 1 By 

 Shroeder van der Kolk's method the refractive index ft was 

 determined in red light to be 1*679. The position of the 

 acute bisectrix was given by C a c = 36°. On a cleavage 

 flake the angle between one of the emergent optic axes and 

 the c axis was determined, and using the above figures the 

 optic axial angle was calculated to be 59° 40' in red light. 

 This result is, of course, no more than a rough approxima- 

 tion. The dispersion of the optic axes is slight, p is greater 

 than I-. By grinding oriented sections of diallage with 

 plates of quartz cut parallel to their optic axis, and com- 

 paring their birefringence with a Babinet compensator, 

 the double refraction on (100) was shown to be '004 and that 

 on (010) which is the maximum birefringence (y - a) was 

 *023. In other samples of diallage the extinction angle 



Thin strips may be seen in section to run parallel to (100) 

 with an orientation different from that of the rest of the 

 crystal. It was not clear whether these were interpo- 

 sitions of hypersthenc or twin lamellae of diallage, but 

 owing to the high birefringence and oblique extinction of 

 some of the lamellae, the latter is considered the more 



' Professor David, Snieeth, and Watt, Proc. Roy. Soc. N.S.W. 1893, p. 



