14 CALIFORNIA ACADEMY OF SCIENCES. 



intermediate between the extreme types, and both very often give 

 faultless or nearly faultless reflections. 



i (P oo ) is always very small on the edge h (2 P oo) : g (0 P) 

 (Fig. 1, PL I) or the edge y (P) : y (P) (Fig. 10, PL III), and 

 is often absent. 



A was observed but once (Fig. 10, PL III) as a lozenge-shaped 

 face in the angle of 6, b, s and s. It is worthy of note that its 

 trace is the only one passing through the zonal point (Fig. 11, 

 PL IV) a = — £, b = 0. 



I have carefully measured the angles in the more important 

 zones on twenty-eight crystals with a large Fuess-Babinet goni- 

 ometer, furnished with two telescopes and Websky's slit as sig- 

 nal. The results, together with calculated values, are collected 

 in the following table. 



In making up the mean values of the measured angles, all 

 readings were rejected that were taken from reflections not sin- 

 gle, or blurred in outline to such an extent that successive read- 

 ings failed to agree to minutes. When, however, readings were 

 to be had from imperfect reflections only, such have been given, 

 and the fact indicated. 



It will be observed that the mean value of each of the funda- 

 mental angles, s : s, h : g, and g : s, agrees remarkably closely 

 with that obtained from the crystal first measured, and which 

 had been used in determining the axial elements some time 

 before the remaining crystals came to hand. Furthermore, 

 while considerable variation exists between the different values 

 of each angle obtained from the various crystals, the larger num- 

 ber of values of each are near the mean, as will be seen from the 



following: 



s : s = (oo P : oo P) =-. 107° 50' 15" 



107 51 



107 55 



Mean 107 56' 16" 



