138 P. SCHEI. 



determined with somewhat care, using the pycnometer method 

 and working with 3.46 grams of fresh homogeneous material in 

 grains of 2 — 5 milhmeter diameter. It was found by two diffe- 

 rent determinations to 5.681 and 5.683, medium. 5.682. 



At the first inspection, the mineral from this locality looked 

 very promising for measurable crystals and a great many loose 

 crystals were gathered. None were, however, in the end found 

 fit for measuring with the reflecting goniometer; some small 

 crystals showed pretty even faces, but unfortunately they were 

 as dull as those on the greater individuals, which were always 

 uneven and irregularly built. 



A set of four small Qiystals 2 — 8 millimeter thick were, 

 however, measured by gumming small pieces of glass to the 

 faces of jlllj ; only such individuals were chosen that showed, 

 at one end of the vertical axis, all four faces of this form. 



The result is given in the following table: 





1: 



2: 



3: 



medium. 



I 



78° 57 



78° 48^ 



79° 0' 



78° 55' 



II 



79° 6' 



79° 0' 



78° 52' 



78° 59' 



III 



79° 3' 



79° 6' 



79° 7' 



79° 5' 



IV 



79° 27 



79° 9' 



79° 1' 



79° 12' 



79° 3' 



The first table gives the values of the angles of jlllj on 

 the four different individuals I, II, III and IV as media of mea- 

 surements of the four angles of the pyramide. 1, 2 and 3 give 

 such media by three separate gummings. As will be seen, these 

 media in the case of each crystal agree very well, because of 

 the errors caused by the gumming on uneven faces being, at 

 least partially, ehminated by the measuring around the solid 

 corner. An idea of the errors involved, that might ensue from 

 a single measurement in this way, is suggested by the follow- 

 ing table, where the angles actually obtained on III are given: 



