J. W. Mallet—Selenide of Bismuth from Guanajuato. 295 
tempting to settle the above questions by careful repetition of 
the chemical analysis. The already pulverized specimen was 
chiefly used, but was supplemented by a portion of the other 
—neither was altogether free from the hydrous silicate of 
aluminum which constitutes the gangue. 
The method employed was the following. Water having 
been driven off by careful heating in a slow stream of carbon 
dioxide gas, collected and weighed, the mineral was mixed 
with ten times its weight of potassium cyanide and fused in an 
atmosphere of hydrogen. The mass on cooling was trea 
with water, and the solution filtered ; the residue on the filter 
dried and again fused with the cyanide to ensure complete de- 
composition, repeating the treatment with water and filtration. 
From the mixed filtrates selenium was thrown down by addi- 
tion of hydrochloric acid in excess, filtered after thirty-six 
hours on a weighed filter, cautiously dried and weighed ; it 
was then burned, and a minute amount of silica left behind 
was determined. The solution from which the selenium had 
used) an inum were now precipitated by ammonium sul- 
phide, and separated by barium carbonate, the alumina being 
deter The original residue of bismuth, left on the filter 
fusion with potassium cyanide, and weighed as me 
ay 
Silentniss.o: cuue cede aoe 
ulphur ......-.-------+-+--- “61 
Biante is fo ss TCE OS 59°92 
abe ee Ec nae 2°53 
Webra OF kkk aii cee trace 
4 ee ee een 3°47 
Ne ee | OO 
99°63 
Zinc was specially looked for, both in the general analysis and 
using a separate portion for this purpose alone, but none could 
