362 Proceedings of the Royal Society of Victoria. 



2. The mineral, after powdering and passing through a fine wire 

 sieve, was treated with hot concentrated hydrochloric acid, and 

 it was found that ultimately everything dissolved except the 

 incrustation and the quartz. Since the incrustation is a silicate, 

 it was necessary, in order to determine its quantitative composi- 

 tion, to separate it from the quartz prior to analysis. The 

 great difficulty experienced in this separation was due to the 

 fact that the specific gravities of the two materials are very 

 close together, being 2.61 and 2.65 respectively, the difference 

 of specific gravity being the particular difference on which the 

 methods of separation depended. 



The first attempt at separation was made by use of a diffusion 

 column using Sondstadt's^ solution, of specific gravity 3.04, at 

 the bottom of a tube covered by a diluted Sondstadt's solution 

 of specific gravity slightly less than 2.6. Many modifications 

 of the method were tried, but all failed to bring about any 

 separation, as shown by examination of minute samples under 

 the microscope. 



The nex-t method attempted was to shake up the mixed 

 powder with Sondstadt's solution in a test tube and allow it 

 to stand. Both minerals then rose to the surface. A single 

 drop of water was then added, and the process repeated until 

 at last a point was reached when the powder was divided into 

 two -layers, one of which sank and the other came to the surface. 

 Samples of each layer were taken, and on examination 

 each proved to be a mixture of the two minerals practically 

 identical with the original mixture, both as to size of particles 

 and as to proportionate quantities present. Several trials of 

 this method always gave the same result. 



The next step was to put this same solution with the sus- 

 pended powder in the centrifuge. Two layers were obtained, 

 the top one of which was poorer and the bottom one richer in 

 quartz than the original mixture. The top layer was now 

 stirred up without disturbing the bottom layer, and on again 

 centrifuging more quartz left the top one. On repeating the 

 operation three or fouz' times the top layer was obtained quite 



1 Sondstadt's solution. An extremely concentrated solution of Mercuric Iodide in 

 aqueous Potassium Iodide solution. 



