an 



KECOKDS OF THE AUSTRALIAN MUSEUM. 



If we are correct in regarding these specimens as pseudouiorplis 

 after glaiiberite, the original crystals must ha^e been acute pyra- 

 midal in habit, with elongation in the direction of the clino-axis 

 (PI. vii., figs. 7, 9). 



As the amount of error in measurements does nut exceed 2'^, 

 the disagreement between the values obtained for the same angle 

 on different crystals must be due to the varying amount of curva- 

 ture and imperfection of form. 



We next proceeded to confinn our results, and to explain, if 

 possible, the divergence of the measurefl angles from the true 

 values by determining the terminal angles between the edges A 

 and B, and C and C '^" (fig. 4), and the terminal pyramidal angles 

 Hii" and .s'';^'" (using the lettering of glauberite for corresponding 

 faces of the pseudomorphs). The results are tabulated below : — 



The difference between the measured and the true angles is 

 considerable, and some explanation must be forthcoming if our 

 conclusions are correct. Now, on several specimens it is apparent 

 that the cleavage is not exactly parallel to the plane of the two 

 opposite edges C and C. (fig. 1). This would be explained 

 (assuming the oi'iginal mineral to have been glauberite in which 

 the cleavage is parallel to this plane) by a curving downward of 

 those edges towards the cleavage. This downward curving could 

 be accounted for by oscillatory combination of the s faces with a 

 form hkh (k>h). No such form is recorded in Dana, but 

 observation reveals the presence of a set of striations on the faces 

 s and s' running parallel to the edge A. These striations would 

 be a natural result of such an oscillatory combination, which 



A is the edge between -v and -s'; B tlie edtje between j/' and «'"; C 

 and C the edges between s and u" and s' and w" respectively. 



Taken fioiu stereos^ram by Penfleld's protractor. 



