50 



sharp line can hardly be drawn between them. The least distinctly 

 developed crystals measure less than V2°"" in diameter and are 

 nearly spherical in form. Single crystals of the other kind reach 

 ^mm across, and they are in general somewhat more distinctly 

 developed. Though their faces are always strongly curved, it may 

 however, easily be observed that the crystals have an octahedral 

 development. It may also usually be seen without difficulty that 

 they are not regular octahedra, being shorter between the opposite 

 corners in one direction than in the two other directions. Another 

 easily observable fact is that two pairs of faces , namely those 

 meeting at two opposite middle edges, if the aforesaid shortest 

 direction is given a perpendicular position, differ as to their 

 physical character from the two other pairs. From all this it is 

 evident that the mineral cannot belong either to the cubic or 

 to the tetragonal system. The optical investigation also shows 

 that we have here a rhombic mineral. As, however, the faces 

 present are of two different kinds they cannot form a rhombic 

 bipyramid. The crystal form of the ancylite must consequently 

 be considered as a combinatian of the prism of the first and 

 that of the second order, viz. (Fig. 3, Plate III) 

 d == {101} and e = {OU}. 

 It has not been possible to find a single crystal with faces 

 plane enough for measuring the angles on the reflecting gonio- 

 meter. That the middle edges of the prismatic faces make 

 angles of very nearly 90° with each other may be perceived 

 without measuring. This is, indeed, all that could be ascer- 

 tained by the attempts made to measure the angles on the 

 reflecting goniometer, for the values obtained for the same angle 

 vary up to 20°. I was therefore obliged to use another method 

 for measuring the angles, by which, however, only approximate 

 results could be obtained. On microscopical sections orientated 

 parallel to the first and the second pinacoids the angles of the 

 outlines were measured under the microscope with the greatest 

 possible accuracy. As, however, the faces are curved, the 



