Vol. 70.] TOPAZ-BEAKIXG HOCKS OF GTJNONG BAKATT. 379 



Let us suppose that in the depths of the igneous mass there was 

 a magma which, if undisturbed, would have crystallized out as a 

 rock composed chiefly of potash-felspar, quartz, and a mica rich in 

 iron. If this magma were invaded by a volume of gas as a huge 

 bubble rising from below, where it had collected in a part of the 

 magma on which it could not react, and if the gas were composed 

 of fluorine partly combined with tin, together with boron and 

 lithia in some form, then the molecules of ' nascent felspar ' etc., 

 would be attacked, violently where the bubble rose, and less 

 violently in its neighbourhood when the gases spread. They would 

 be forced to rearrange themselves in combination with the attack- 

 ing gases, and the heat generated would raise the temperature 

 sufficiently to lead to an immediate irruption of this part of the 

 magma into the consolidated granite above. As the molten mass 

 rushed through the comparatively cool rock the molecules would 

 begin to form crystals. The alumina of what had been nascent 

 felspar-groups crystallizes out in combination with fluorine and 

 silica to form topaz ; the molecules that might have been mica 

 without lithia crystallize as zinnwaldite and as tourmaline ; the 

 potash of what had been nascent felspar goes into both these 

 minerals ; and the tin ciystallizes out as cassiterite. The base in 

 which these minerals were formed was an excess of silica, and 

 before this solidified as quartz there would be nothing to prevent 

 some degree of segregation of the earlier-formed minerals. 1 This 

 will explain the abundance of zinnwaldite in Quarry M and its 

 absence elsewhere except in the mica-patches. It also explains the 

 irregular, sporadic distribution of cassiterite, and the absence of any 

 fixed relation between the amount of cassiterite and that of topaz. 



The chemical reactions in the ma^ma would certainlv be more 

 complicated than is indicated above, but the course of events 

 sketched will suffice to show on what the hypothesis is based. It 

 is hardly necessary to endeavour to work out complete formulae for 

 the magma, the gases that attacked it, and the resulting minerals. 



The reaction-borders of schorl-rock and greisen are easily ac- 

 counted for by this hypothesis. They are the effect of the surplus 

 gases entangled in the solidifying rock on the granite walls of the 

 veins, and the total absence of felspar in the quartz- topaz rock is 

 due to there having been an excess of these gases. 



Meanwhile, during the cooling of the quartz-topaz rock, the 

 remainder of the magma was being acted upon slowly by the smaller 

 quantities of gas that spread out from the bubble. There was not 



1 Without segregation one could not expect to have a rock very rich iu 

 topaz. In the case of a pure orthoclase-magma, the percentage of alumina 

 would be 18*4. Since in 100 parts of topaz there are 56'5 of alumina, with 

 18*4 per cent, of alumina available only 32 - 6 per cent, could be formed if the 

 magma were invaded by fluorine unaccompanied by more alumina. It is 

 unlikely that the magma would be a pure orthoclase-magma, therefore the 

 percentage of topaz possible would be lower than 32 - 6. In the case of greisen 

 bordering a fissure-vein with distinct walls, I do not see how segregation of 

 minerals could take place without the effacement of the walls of the fissure- 

 vein. 



