35 



common with the Narsarsuk mineral. The twins have the base 

 as composition-face, and one of the individuals is revolved 60° 

 about the vertical axis with reference to the other individual. 

 A twinning of this kind would not, of course, be possible, if 

 the mineral were holohedral hexagonal. The twins are easily 

 recognizable by alternating re-entrant and salient angles at the 

 line of twinning. Sometimes the twinning is repeated, several 

 lamellae in twin position having grown together, which gives 

 the crystals a step-like or strongly striated appearance. If both 

 negative and positive rhombohedrons occur, the twins will re- 

 semble hexagonal bipyramids. It might be supposed that the 

 South American parisite were formed in such a manner and 

 that, consequently, it is only seemingly holohedral. 



In the table on the following page I have brought together 

 the results of measurements from which the different forms have 

 been determined, and also the corresponding calculated values. 



The specific gravity of the mineral from Narsarsuk I found 

 by weighing in benzole, to be = 3,902. The specific gravity 

 of the parisite from Muso is according to Dam our = 4,3ù8, 

 according to Vrba = 4,36i. By way of control I have also 

 determined the sp. gr. of a crystal of the Muso mineral by the 

 same method as I used for the mineral from Narsarsuk; I 

 found it to be == 4,3915. The material used for both deter- 

 minations was pure and homogeneous. The hardness of the 

 Greenland mineral is the same as that of the Muso parisite, 

 viz. 4,5. The mineral is brittle and easily pulverized. The 

 fracture is subconchoidal to splintery. In quite fresh material 

 i. e. in the small crystals and in the end portions of the 

 larger ones no cleavage is observable. The middle portions 

 of the larger crystals, on the other hand, may often be divi- 

 ded pai'allel to the base. Cut parallel to the c-axis they also 

 show in their middle part interrupted partings parallel to 

 the base. These partings, however, are probably not to be 

 regarded as true cleavages, but as partings of secondary origin, 



3* 



