f[ 
i 
! PLATINUM FOUND IN GADOLINITE. 123 
; CHEMICAL ANALYSIS OF GADOLI- 
NITE, TOGETHER WITH AN EXAMI- 
NATION OF SOME OF THE SALT OTF 
' YTTRIA AND CERIUM. 
, By Thomas Thomson, M.D., F.R.S., 
! L. AND E., &c., 
j Regius Professor of Chemistry in the Univer- 
[ sity bf Glasgoio ; and Andrew Steel, M.D. 
\ The specimen of gadolinite which furnished 
1 materials for the experiments contained ia 
I this paper, was purchased several years ago 
from a German mineral dealer. He stated 
that he had accidentally observed it in a col- 
lection of minerals in Sweden, from the pro- 
prietor of which he procured it, and that its 
orignal locality was not certainly known. It 
weighed several ounces. 
It was an amorphous mass, having a very 
deep green colour, so as to appear to the eye 
almost black. 
The lustre on the fresh fracture was vitre- 
ous and splendent. But when broken in cer- 
tain directions it presented a sui’face almost 
dull, having a whitish aspect; but when viewed 
ed throxigh a microscope no extraneous mat- 
ter could be observed. The colour and want 
of lustre was probably owing to long exposure 
to the air, which had acted through certain 
natural rents in mineral. For it broke with 
much more facility, so as to exhibit the dull 
than the splendent surface. 
Hardness 6'5 ; specific gravity at 60° from 
4*1493 to 4*1795. The mean was 4*1607. 
Its other chai-acters being the same as those 
of common gadolinite need not be described 
here. Twenty grains, by ignition, acquired a 
brownish colour, and lost 0*198 grains of 
weight, or almost one per cent. This loss was 
doubtless owing to the escape of water. 
During the pounding of the mineral for 
analysis small metallic looking grains were 
observed, which were carefully picked out and 
submitted to the following examinations. They 
were malleable, infusible before the blowpipe, 
not acted on by muriatic acid, but dissolving 
slowly in aqua regia. The solution had a deep 
orange colour, a few blackish grains remain- 
ing at the bottom. It aftbi-ded a yellow pre- 
cipitate with sal ammoniac and salts of potash 
and was obviously platinum. From 120 grs. 
of the mineral 2*33 grs. of platinum were 
picked out. But the quantity was found to 
vary in different pieces. 
We attempted to analyze gadolinite by the 
processes employed by Berzelius. But we 
found that the peroxide of iron, wEich we had 
precipitated by benzoate of . ammonia, con- 
tained also'glucina, from which we could not 
separate it by means of benzoate of ammonia, 
however carefully added. This led us to a 
careful investigation of the properties of 
cerium, yttria, and glucina. The facts ascer- 
tained suggested the following method of 
analyzing gadolinite : — 
A. 25 grs. of the mineral, finely pounded, 
were boiled in a flash with aqa regia, till the 
whole was decomposed. The gelatinous silica 
remaining, collected on a filter, and well 
washed, weighed, after ignition, 6*22 grs. 
B. The residual liquid was evaporated to 
dryness, and the solid matter remaining, 
digested for some time in distilled w^ater. A 
solution of oxalic acid was then added as 
long as it occasioned a precipitate, and until 
this precipitate was of a perfectly white 
colour. The oxalic acid retained in the solu- 
tion the iron, glucina,. &c., and threw down 
the yttria and cerium imthe state of oxalates. 
The whole was thrown upon a filter, and the 
white insoluble matter was well washed. 
When dry it was a beautiful light, snow-white 
powder, and weighed, after ignition, 12*617 
grs. It had now assumed a light yellow 
colour and was a mixture of oxide of cerium 
and yttria. 
Various methods of separating these two 
substances from each other were tried, but 
without success, on account of the perfect 
correspondence of the two in all their pro- 
perties. We were, therefore, obliged to have 
recourse to the common method of putting a 
quantity of solid sulphate of potash into a 
neutral solution of the yttria and oxide of 
cerium, which threw down the cerium and 
left yttria. But as the quatities w^ere never 
absolutely the same in two successive expe- 
riments, this method is certainly suspicious. 
By this method the 12*176 grs. were resolved 
into 
Yttria 11*087 
Peroxide of cerium l*530=143protoxide. 
12*617 
C. The oxalic solution from (B.) was eva- 
porated to dryness, and ignited to destroy the 
oxalic acid. The residue was dissolved in 
muriatic acid, by the assistance of heat, with 
the exception of a few blackish grains. These 
were at first supposed to be charcoal from 
the oxalic acid, but after being collected and 
ignited they proved to be platinum, weigh- 
ing 0*45 grs. 
D. The muriatic acid solution was now 
evaporated to dryness, and to insure perfect 
neutralization, heated, so that a small portion 
of the iron became insoluble on its being 
again digested in water. Into this muddy 
liquid a quantity of benzoic acid was thrown, 
and well stirred. After standing 24 hours 
the whole iron was found to be precipitated, 
while glucina remained in solution.* The 
benzoate of iron was collected on a filter, 
washed, dried, and ignited. The peroxide of 
iron obtained weighed 3*75 grs. =3*375 pro- 
toxide. On trial it appeared perfectly pure. 
E. The residual solution was now mixed 
with ammonia, which threw down a white 
floculent precipitate. It was collected on a 
filter, well washed and dried. After ignition 
it had a dirty brownish white colour, and 
weighed 3*182 grs. 
It dissolved in acids, forming very sweet 
tasted salts. The sulphuric acid solution 
* It had been ascertained by experiment that 
benzoate of ammonia throws down glucina, 
which benzoic acid does not; and that, with 
care, the peroxide of iron may be completely 
thrown doAvn by benzoic acid. 
