MINERAL CONSTITUENTS OF METEORITES. 
193 
The retort is next transferred to a bath of paraffin and carefully heated in it. At first 
hydrogen fluoride conies over; and at this point of the process the flow of the hydrogen 
requires a little attention. At about 132° C., with the silicates described in this memoir, 
the silica first becomes visible in fine flocks in the ammonia solution, and in another minute 
the whole is cloudy. In from five to ten minutes the temperature has risen to 142°-145°, 
and so much of the fluoride has come over that the contents of the tube are of a semi- 
solid consistency, and nearly the whole of it has in fact passed over. The temperature 
is allowed to increase to 150°, and the retort then permitted to cool. The process is 
repeated by introducing a fresh charge of hydrogen fluoride into the retort and of 
ammonia into the test-tube, and again heating in the paraffin bath. If the quantity of 
the silicate taken be not more than 02 grm., twice charging the retort is sufficient; 
if it amount to 05, three or four repetitions of the process are required. The process 
must in fact be repeated so long as any fresh flocks of silica can be seen to form in the 
ammonia tube. Finally, 0-75 cub. centim. of sulphuric acid are introduced into the 
retort, and the temperature raised to 160°, the stream of hydrogen being continued as 
before*. 
The several ammoniacal charges of the ammonia tube are now brought together into 
a platinum dish with all the washings from the test-tubes and the connecting tubes ; 
and these are now slowly evaporated in a water-bath with continual stirring. 
At a certain point of the evaporation, just before the solution becomes neutral, and 
the ammonium fluoride begins to become acid, all the silica in the dish is dissolved by the 
fluoride. The process is gradual, but the moment is easily determined when it is com- 
plete. Then the dish being removed from the water-bath, potassium chloride is added 
in slight excess; and absolute alcohol equal in bulk to the liquid in the platinum 
vessel is poured on it. Potassium fluosilicate precipitates, and after standing twenty- 
four hours it is filtered and washed with a mixture of equal volumes of absolute alcohol 
and water, and then dried and weighed. The results are accurate. 
In the platinum retort are the bases, in the form of sulphates, the treatment of which 
calls for no further remark. 
A specimen of diopside, pulverised and analysed by the method here described, yielded 
53*46 per cent. ; on treatment by fusion with potassium and sodium carbonate, it gave 
in two analyses 53*51 and 53*54 per cent, of silicic acid. 
III. The Busti Aerolite of 1852. 
Among the meteorites with ingredients sufficiently large in the grain to offer an 
opportunity for isolating and determining their constituent minerals, is a stone that fell 
* In a series of analyses made witli a view to determine the degree of energy with which the acid attacks 
various silicates and the forms of silica itself, it was found that the first turbidity of the ammonia will, if suffi- 
cient time he allowed, commence at 120° C. At this temperature to 121° C. twenty-three minutes were requi- 
site for the bulk of the silicium compound to come over from the retort. It was not found, however, that the 
action of the acid was more energetic on one silicate than on any other, or on quartz. 
