eiger.J ACTION OF SILVEE NITRATE, ETC., ON SILICATES. 81 
The reliction here is incomplete, but the thorasonite ratios have 
sen fairly well preserved. 
As the water of this mineral is usually considered to form no part 
I the chemical molecule, in the following formula it has been repre- 
mted separately. 
For this and the two minerals to follow, the formulae proposed by 
. W. Clarke have been adopted, regarding them as mixtures of salts 
f orthosilicic and trisilicic acids. The analysis gives the subjoined 
itios: 
18Ag 2 0, 23CaO, 2Na 2 0, 40Al 2 O 3 , 100SiO 2 , 80H 2 O. 
Combining monoxide bases, we have 
R^ 3 Al 80 Si 100 O 363 .80H 2 O; 
hich gives a ratio of orthosilicic to trisilicic acids of 
7;3Si0 4 : 9Si 3 8 . 
Substituting these, we have 
R" 43 Al 80 (SiO t )„(Si 3 O 8 ) 9 . 80H 8 O; 
nd combining the two acid radicles under the indiscriminate symbol X 
r e have 
R" 43 A1 80 X 82 .80 H 2 = 1-2-2, 2H 2 0. 
The analysis of thomsonite reduced in the same way gives the for- 
iula 
R" 4a Al 86 (SiOJ 78 (Si 3 8 ) 8 . 98H„0, 
, condensed, 
R" 42 Al 85 X 84 .98H 2 = l-2-2, 2iH 2 0. 
CHABAZITE. 
The chabazite used in the following work was the well crystallized 
ish-colored material from Wassons Bluff, Nova Scotia. 
Two silver derivatives were made, as follows: 
A. Ammonium chabazite was heated in an open tube with four times 
is weight of dry silver nitrate for six hours at 300° C. 
I |B. Natural chabazite was heated in a sealed tube with four times its 
ight of dry silver nitrate for twenty-four hours at 250°-290 C. 
Bull. 262—05 6 
