Peckham and Hall—Thomsonite from Minnesota. 129 
These figures prove conclusively that we were dealing with 
varieties of the same mineral. On comparing these percentages 
with those given in Dana,* the water and silica were found to be 
high 
igh. 
Computing the oxygen ratios and formula for Number III, we 
a 
Per cent. Metal. Oxygen. Atoms. 
FeV 0°40 03111 0-0889 0064 
K,O 0-49 0°4068 0:0832 “0052 
Na.O 4-055 3°0118 1:0432 0654 
CaO 10°37 7:4070 2°9630 1852 
4:1783R. 2622 
Al,O, —30°215 16°1343 14-0807 2933 
Sid, _-40°605 18949 21 656 676 
H,O 13°75 1528 12°222 165 
Dividing the oxygen percentages by 5, we have 
RO:R,0,:8i0,: H,O=1: 3:4: 24, 
which is the ratio for thomsonite, given in Dana’s Mineralogy, 
with the bases low and the silica and water high. Dividing the 
atoms by ‘005 we have the formula 
(2(FeO+K,0+Na,O) +2Ca0)Al,0,, (SiO,),(H,0),, 
with the protoxide bases low, and the silica and water high.t+ 
Computing the ratios after Rammelsberg, we have: 
(Na+K) : (Ca+Fe) ::1: 1°35 
Ca nre. :Adl::1:1°54 
Ca+Fe+K+Na)=R: Al: Si:: 1°13: 1: 2°03 
Si: H ::.1 22°26 
Rammelsberg deduces from these ratios a formula which he 
calls a half silicate (Halbsilicat), according to the expression 
m(2Ca AlSi,O,+-5aq) ) f 
n(2Na,AlSi,O,+-5aq 
in which m indicates a certain proportion of a hydrous silicate 
of aluminum and a dyad protoxide, and n a certain proportion 
of a hydrous silicate of aluminum and an alkaline or monad 
protoxide. The ratio between m and 2 varies in different 
Specimens. Number I and Number II, without the excess of 
silica, approach more nearly the thomsonite of Elbogen in com- 
tae (in which the ratio of m ton=2:1) than any mentioned 
y Rammelsberg. While the ratio of Si to H is about the 
same as given by Rammelsberg, the percentage of both in these 
Specimens is higher than in the analyses quoted by him. 
* System of Mineralogy, fifth edition, p. 425. + 5th ed., p. 425. 
Rammelsberg Min. Chem., Ed. 1875, p. 637. 
