392 K T. Allen, F. K Wright and J. K Clement- 



Mg-pyroxene. 



0*l mm long and 0'04: nim wide, from a solution in calcium vanadate, 

 was first measured. A single fine twinning lamella after the 

 s orthopinacoid was observed under the 



microscope, but could not be detected 

 on the goniometer. The reflection sig- 

 nals were very faint and the recorded 

 angles fluctuated somewhat in conse- 

 quence. The largest and best devel- 

 oped crystals (fig. 3) were produced by 

 heating magnesium metasilicate with 

 magnesium chloride in a stream of dry 

 hydrochloric acid gas. Six different 

 crystals obtained by this method were 

 measured and the forms : c (001) (?), 

 a (100), I (010), m (110), n (120), 

 I (250) (fyjfc (310), r (210) (I),* (111), 

 i (121), p (101), s (111), e (121), (103) 

 (?), and (103) (?) observed. Several of 

 these forms, c (001) (?), r (210) (?), 

 I (250) (?), (103), (103), were noted 

 only once, gave poor reflection signals, 

 and are therefore uncertain. The faces in the prism zone were 

 much better developed than the terminal forms and, although 

 small, usually gave sharp reflection signals and concordant 

 angles. The prismatic cleavage angle (110 : 110 = 88° 08') 

 thus obtained has a probable error of only =b 3'. The terminal 

 faces, on the other hand, were much smaller, less clearly 

 defined and seriously interrupted by intergrown twinning 

 lamellae, so that the axial ratio obtained for the vertical axis 

 is of a lower order of certitude. 



In order to show graphically the variations in the angles 

 observed in the different crystals, a gnomonic projection of the 

 results from the six crystals measured is presented in figure 4. 

 The irregularity in the position of many of the projection 

 points is due in large measure to the indistinct reflection sig- 

 nals obtained. Transitional faces were observed in several 

 zones and have been indicated in the figure by the shaded 

 portions of the zone lines. They furnish an excellent criterion 

 for the primary zones and nodes which dominate the develop- 

 ment of the forms bounding the crystal. 



On comparing these crystallographic measurements with the 

 angles given for enstatite and diopside, it was found that the 

 prism angles, and therefore the axial ratio a : h for the three 

 minerals, were very similar, while the angles for the terminal 

 faces differ so noticeably that they cannot be ascribed to experi- 

 mental error alone. The axial ratio a :h: c for 



