NO. 3 SMITHSONIAN EXPLORATIONS, IQI5 PON 
have been noted on the specimens, although cleavage planes, beveling the 
edges of the surface form, are visible here and there. The faces of the nega- 
tive rhombchedron are not smooth nor simple, but are built up of a series of 
innumerable smaller faces, belonging to individuals in approximate parallel 
position, though with enough divergence to cause the curvature of the faces; 
while at the thin edges of the crystals the individuals are still more distinctly 
Fic. 30.— Calcite rosette showing somewhat different structure from that of 
Fig. 29. Dr. Edgar T. Wherry, of the United States National Museum, 
explains their origin as due to starting of crystallization at numerous closely 
crowded points, the crystals being perhaps at the start quite parallel, but as 
they grew crowding one another out of parallelism, although not enough to 
prevent approximately equal growth of every individual. He states that this 
type of crystallization of the mineral calcite is comparatively rare. Photo- 
graph by Mr. T. W. Smillie, U. S. National Museum. 
separated, so that the effect is that of superposed layers of curved sheets, 
resembling the petals of the rose to such an extent that the first word to 
describe them which suggests itself is “ rosette.” 
