102 



Messrs. D. Herman and F. Rutley. 



marked crystalline rods, apparently composed of piles of octahedra 

 like those of alnm, and where they tonch the margin of the crystal 

 they usually pass beyond it, forming little spicnlar crystallisations 

 like fir-trees or like the crystals formed in cast iron. They throw out 

 branches at right angles to the main spicule. The crystallite is also 

 traversed by other crystalline rods of a like character, but at right 

 angles to the first set, and these also pass out in little fir-tree-like 

 crystallisations. There are also small rods which run in two direc- 

 tions obliquely to the former, and which intersect in an angle of 

 about 87°. The form therefore is not cubic, as might at first sight be 

 thought. The larger spiculae also show double refraction. There is 

 some depolarisation in the glass around this crystallite due to strain, 

 but no cracks are developed. The spicule extinguish parallel to and 

 at right angles to their longest axes. At least they appear to do so, 

 but it is difficult to tell, and the colour difference is so slight when a 

 Klein's plate is employed that it is impossible to speak with any 

 certainty on this point. On the whole we are inclined to regard these 

 crystallites as belonging to the rhombic system. The one last de- 

 scribed is a twinned form, and exhibits several re-entering angles. 

 A rough sketch of it (fig. 2) is appended. 



Fig. 2. 



