| HARRINGTON | MINERALOGICAL CHEMISTRY 5 
gested by Retgers') will be found to be an admirable medium for the 
purpose in question. The double salt fuses at 75° C. and in the fused 
state is clear and mobile and has a specific gravity of 5. A great advan- 
tage here, too, is that the density may be lowered by the addition of water. 
and the salt regained by evaporation. 
In the separation of minerals for the purposes of analysis advantage 
is frequently taken of the fact that certain solvents may attack some of 
the constituents of a mixture and not attack others, or perhaps only dis- 
solve them very slowly. Some species, such as zircon, rutile and andalu- 
site, for example, resist the action of hydrofluoric acid (or HF with HC] 
or H,SO,), while others readily pass into solution. Woodward long ago 
took advantage of this fact in estimating the zirconia in certain crystalline 
rocks,’ and more recently, by this means, several chemists have succeeded 
in isolating rutile from slates. 
Certain sulphides such as pyrrohotite, stibnite, sphalerite and bornite, 
are, according to Bolton, readily attacked by citric acid, while pyrite, 
chalcopyrite, chalcocite and many other sulphides resist the action of that 
acid. The removal of carbonates, by means of dilute acids, from minerals 
previous to analysis scarcely requires comment here. 
While it is not my intention to go into minute details with regard to 
the analysis of minerals, there are a few questions in this connection to 
which | wish to refer. In the analysis of silicates, chemists are too fre- 
quently content with determining the total iron and then calculating it as 
ferrous or ferric oxide as the case seems to demand ; but in any important 
analysis the proportions of the two oxides, if both are present, should be 
carefully determined. Without such determinations it is impossible in 
many cases to draw any satisfactory conclusions with regard to the con 
stitution and formula of the mineral. An impression seems to prevail 
that the estimation of ferrous iron in insoluble silicates is a difficult matter, 
while in fact the operation is usually simple and rapidly performed. 
Tolerably accurate determinations may be made by decomposing the 
silicate with hydrofluoric and sulphuric acids in a platinum crucible with- 
out the employment of carbon dioxide, heating with a small flame.’ 

‘Jahrb. f. Min, 1893,i. p. 90. The double salt may be prepared by bringing 
together the two nitrates in the proportion of one molecule of each. Thallium nitrate 
is easily made by dissolving thallium in nitric acid. 
Cle 
? Rept. of the Geological Exploration of the Fortieth Parallel, vol. ii., 1877— 
Descriptive Geology, by Clarence King, p. 397. 
* By this method I have sometimes obtained higher results than by heating over 
the water-bath in an atmosphere of CO,. The higher results were no doubt due to 
more perfect decomposition of the silicate. In one case a mineral yielded 21:05 per 
cent of ferrous oxide, using CO, and the water-bath, and 21°97 p. c. over the flame 
without CO,. 
