OP ARTS AND SCIENCES. 65 



V. 



CONTRIBUTIONS FROM THE CHEMICAL LABORATORY 

 OF HARVARD COLLEGE. 



By W. H. Melville. 



Presented, July, 1881. 



CRYSTALLINE FORM OF CRYOLITE. 



The crystalline form of Cryolite was described by Dana in hia 

 " System of Mineralogy," 1868, as Trimetric, but, as he distinctly stated, 

 the system in which the mineral crystallizes was considered doubtful. 

 Des Cloizeaux has since investigated the optical characters, and deter- 

 mined the system of crystallization to be Tricliuic, which Websky 

 corroborates by his measurements. 



The difficulties which are presented in the determination by measure- 

 ments are two-fold. In the first place, the angles which the three most 

 prominent faces make with each other are very closely right angles, 

 thus suggesting one of the three orthometric systems ; and, secondly, 

 these angles are rendered uncertain by the presence of striations, a 

 habit almost invariable. Minute crystals (Figs. 1 and 2), varying 

 from three to five hundredths of an inch in length, were employed in 

 the following determination, and these were taken from the surface of 

 a pure white specimen of Cryolite. By mounting a large r>umber of 

 crystals, it was observed that the angles made by the three pinacoid 

 planes with each other were quite constant, — the greatest difference 

 from the means amounting to about 3', — so that all suspicioii which I 

 entertained as to the probability of their being right angles, was removed. 

 A few colorless transparent crystals, absolutely free from stria?, and 

 showing no evidence of twinning, were found, and from these the funda- 

 mental angles used in calculating the elements of the crystalline form 

 were taken. They are represented in Fig. L The plane (001) was 

 chosen for the basal section, because it formed the termination of the 

 crystals, and was not commonly striated. 



