42 Scientific Proceedings, Royal Dublin Society. 



readily extend the method above the specific gravity 2-77, that of 

 Thoulet's solution (the mutually-saturated solutions of biniodide 

 of mercury and iodide of potassium). The solutions necessary to 

 extend the range above this are either costly or difficult to work 

 with : some can only be used by maintaining them at a high 

 temperature (as lead chloride at 400° C). Ir any case the range 

 of density hardly passes that of garnet, 3*4 - 4*3, and I am not 

 aware of any other published method of dealiug with small frag- 

 ments of minerals of a specific gravity exceeding this. In short, 

 if a few milligrams of any of the host of minerals ranging above 

 4'5 in density — about 90 per cent, of the unsilicated mineral 

 species — be presented to the mineralogist, he is unable to deter- 

 mine this characteristic. 



In the second case — the case of porous bodies — the impossi- 

 bility of freeing the body from contained air, when immersed in 

 liquids of the nature of those to which we are restricted, renders 

 the method fallacious. The air-pump or the application of heat 

 will generally be found of little avail. In this case we must again 

 seek a large quantity of the substance, so that we may be able to 

 weigh it in a liquid of low-surface tension, or of a ' creeping ' 

 nature, such as turpentine or alcohol. Small quantities cannot be 

 dealt with. 



The method to be now described enables the specific gravity of 

 substances to be determined under both these conditions — that is, 

 whatever their density or whatever their state of aggregation — in 

 extremely minute quantities, with an accuracy limited only by the 

 sensitiveness of the chemical balance, and by the aid of solutions 

 of a density varying from about that of water to say twice that of 

 water : but this is under our own control. Unfortunately, it is 

 inapplicable to the purpose of effecting the separation of bodies of 

 different specific gravities. 



Briefly, the theory of the method is as follows : — The mineral 

 by itself will not float in any known solution, suppose. If, how- 

 ever, we mix it with another substance of much lower specific 

 gravity, there is easily found such a proportion for the constituents 

 as will enable the mixed bodies to be equilibrated by dilution of 

 the specific gravity liquid. We may, in short, adjust the specific 

 gravity of the mixed substances to be as close to that of either of 

 them as we please. 



