10 CALIFORNIA ACADEMY OF SCIENCES. 



P(x P \%) was found on a small cleavage-fragment as a very 

 narrow face on the edge s (go P): m (go P oo), the latter face be- 

 ing also very small as a cleavage surface. The reflection from P 

 took the form of a very broad band of light that entered and left 

 the field so gradually that no precise points could be fixed upon 

 as limits. The maximum culmination was symmetrically in the 

 centre, but neither very bright nor sharply defined, so that an 

 error of 15' in the recorded reading is quite possible. The fol- 

 lowing angles will show that |g is probably the correct value of 

 the coefficient: 



s (co P) : P = 162° 9' (observed) 



s : P ( = co P }-l) = 161 55 8" (calculated) 



8 : P ( = co P £) = 163 24 12 



s : P ( == co P 2) = 160 32 10 



s (go P) is invariably the only form largely developed in the 

 zone of the vertical axis. The faces are always bright, but the 

 reflections often highly complex from the slight want of parallel- 

 ism in the numerous sub-individuals that make up the crystal. 

 From four crystals only could faultless reflections be obtained, 

 and the mean value of s : s from these crystals is 107° 55' 17". 

 Four other crystals furnished reflections that were single and 

 only very slightly hazy in outline. Measurements from these faces 

 were accurate to within one minute, and averaged with the above 

 result, give as a mean from eight crystals, s : s = 107° 56' 17". 

 This differs only 8" from the value obtained for this angle in the 

 crystal first measured, that is, from the value used in the deter- 

 mination of the axial elements. 



I (go P 2) is very generally present with bright faces, but 

 usually very small. 



c (P oo) is never absent and varies from a very small plane 

 replacing the edge g (0 P) : a (2 P oo), when g (0 P) is large, to 

 one of the dominant terminal faces. 



a (2 P oo) is small in only one instance. On the crystal rep- 

 resented in Fig. 3, PI. I, it occurs as an exceedingly minute tri- 

 angular face in the angle of the three faces c, oo, and d. It was 

 overlooked in the drawing, and in any case would have had to 

 be considerably enlarged for a drawing on that scale. It was 



