308 



C. ANDERSON. 



Segments. 



Meas. b a b 



I, II 

 I, v 



57° 17' * 



57 10 fCalc. for r-twin 57 18 



I, III 

 III, IV 



62 38 jCalc. for m-twin 62° 46' 

 62 43 J 



and there is no 



marked divergence from the true twin 



position. 





Conclusion, 



Generally speaking the variations from the true twin 

 position are in the same direction as in the crystals inves- 

 tigated by Goldschmidt and Hubrecht, but this is not 

 always the case, and it is doubtful whether any far- 

 reaching conclusion can be drawn on the basis of these 

 observations. 



