34 BELL SYSTEM TECHNICAL JOURNAL 



frequently as the natural faces of the most common form of the diamond 

 crystal, the octahedron. The disposition of the stone in the above manner 

 would be logically expected to take place in the mounting methods generally 

 used by the American die manufacturers since the stone would probably 

 rest on a flat side, an octahedral face. This placement for very obvious 

 reasons also facilitates cutting true the starting cone for the drilling 

 operations. 



A few stones had been drilled normal to the 110 and 100 planes or at an 

 angle to the 111 plane. Holes normal to the 110 plane, which is the face of 

 the rhombic dodecahedron, another diamond crystal form, are along a face 

 diagonal of the cube. Holes normal to the 100 plane, which is the cube face, 



- o ^ X 



< 



d 

 h 



A B 



Fig. €> 



A — Diamond crystal showing grain and faces (crystallographic planes) 



h. Cube face 100 plane 



d. Rhombic dodecahedron 110 plane 



o. Octahedron Ill plane 



B — Diamond octahedron 



are along a cube edge. Very few of these were found. This is as expected 

 since natural crystals with faces parallel to the 100 plane are rare. This 

 face is usually produced by sawing or "bruting." 



As previously noted, diamond die life is restricted by the breakage and 

 "chipping out" of the stone. The manufacture of the dies with due con- 

 sideration of the peculiar crystallographic structure of the diamond to 

 minimize this condition should greatly enhance its performance. Findings 

 to date substantiate this conclusion. Seventy per cent of the dies which 

 were drilled normal or at a small angle to the 111 plane failed due to cracks 

 parallel to this plane and normal to the drawing axis. The crack usually 

 occurred at the pressure ring and its plane coincided with the cleavage plane, 

 the weakest one in the diamond crystal. The stone in many cases on re- 



