140 On Cet^iain Cavities in Quartz, ^c. 



formly rhombic, than in any other that I have seen. The gene- 

 ral figure of this specimen is oblong. It is compressed upon two 

 opposite surfaces, and possesses additional interest by having two 

 translucent crystallized surfaces capping the one end. These 

 faces incline at an angle of about 120°, are 8 inches long, and 

 3 1 wide, having a ragged periphery, as if broken, and constituting 

 two of the planes of an hexagonal prism. Another smaller speci- 

 men has two sides of the prism and two sides of the pyramid 

 extending across the centre of the mass ; while a third is com- 

 pletely surrounded by innumerable cells, with acicular crystals of 

 actynolite pervading one end. 



In order to ascertain the particular angles of these cavities, 

 several casts were made for me by my friend Dr. E. Parry. Tak- 

 ing these as correct models of the crystals that preoccupied these 

 cavities, it will be easy to ascertain their exact figure. The fol- 

 lowing diagrams will represent the angles of several of the casts, 

 viz. 



These angular and tabular impressions vary in size from the 

 smallest pyramidal point and rhombic table, to rhombic cavities 

 of three inches parallel diameter, and two inches in depth. And 

 you will perceive that the above diagrams are all portions of 

 rhombic figures. 



A geological inquiry naturally arises as to the cause of the pe- 

 culiar cellular structure of these specimens. Before inquiring 

 into the particular agents which prepared the cjuartz to receive 

 these impressions, and caused the dispersion of the crystals that 

 produced them, it will be necessary to examine into the character 

 of the solids, which, at one time, were encased by these peculiar 

 cavities. Your attention has been called above to the rhombic 

 form of the casts. Now there are no mineralogical specimens, of 

 a rhombic character, occurring in our vicinity, that so exactly 

 correspond with these casts as those of calcareous spar. This is 



