Gold schmidt and Nicol — New Forms of Sperrylite. 453 



2. Curvature of faces near the outlines. This is a considera- 

 tion of importance in connection with- the origin of the crystal 

 and demands special study. S perry] ite itself may perhaps 

 furnish suitable material for this on account of its perfect 

 metallic lustre, which makes it possible to secure good signal 

 reflections from even minute faces : on account of the simplic- 

 ity of the crystal system and the certainty of the elements and 

 calculated angles : on account of the perfect formation of the 

 principal faces, the great variety of kinds of faces and the large 

 number of individual faces, which are incident to the regular 

 system. It may be said : the degree of perfection of a face is 

 a function of Us ahsolute and relative size. Very large faces 

 are usually imperfect since their parts and the parts of the lay- 

 ers underneath have a development history exposed to different 

 influences. The faces of smaller crystals are, as a rule, better 

 formed than those of larger crystals. Below a certain size the 

 faces again become more imperfect. This may be assigned to 

 two causes : 



1. The relatively small faces are those derived from higher 

 complication. They are weak and more easily displaced. They 

 are mobile, like the leaves and- branches in the wind, while the 

 trunk is unmoved. 



2. In the case of small faces the marginal parts are relatively 

 large. The marginal parts, however (on edges and corners), 

 are the places at which the particles are bounded only on one 

 side by securely situated particles which are, as well, oriented 

 themselves and have an orienting influence on their neighbors. 

 This is the place of the manifold formation of varying result- 

 ants by complication.* The midfield of not too great a face is 

 therefore usually the smoothest. Convexity readily occurs at 

 the perimeter and corners. Small faces lying close together 

 arid very little inclined towards one another, frequently shade 

 into each other by rounding. It may happen that only the 

 border part is present, no smooth middle part has been formed. 

 If, now, the border part is large relatively, it furnishes much 

 light for the signal reflection. In this way the signal reflection 

 from the inner portion of the face, when this is small and of 

 low brilliancy, may T become darkened and eclipsed by that of 

 the border parts. If the faces are too small it is not possible 

 to shut off the border portions by means of the shutter at the 

 eye-piece of the telescope. It is not advisable at this place to 

 go more deeply into these interesting points. They should 

 form the ground of a special investigation. 



The discussion of the position of the faces and the symbols 

 of the subordinate ones must be undertaken with caution on 

 the above mentioned grounds. Considerable experience is 



* Compare Zeitschr. f. Kryst., xxix, 47, 1897. 



