PROCEEDINGS OF GEOLOGICAL SOCIETIES. 
179 
mineral. It does not at all follow that his process should be exactly that 
which nature herself has adopted, but still it is remarkably illustrative of 
what may occur, and on that account is worthy of notice in detail. We take 
a porcelain tube, and place it vertically in a furnace. It is open at both 
ends, and we introduce through the upper part of the tube the gas, fluo- 
ride of silicon. At the top we have a layer of alumina; then there is a 
layer of silica ; and so on in alternation— alumina, silica, alumina, silica- 
beginning with alumina, and finishing with silica. We will heat that tube 
to red-whiteness, and then note the result. The gas, fluoride of silicon, 
is a compound of fluorine and silicon. AVhen it comes in contact with 
the alumina, a curious reaction takes place. A portion of the silica is 
deposited here, and we get the mineral silicate of alumina, a corre- 
sponding proportion of the aluminium displaced and volatilized in the 
form of fluoride of aluminium. It is important clearly to understand this 
part of the operation. You perceive that we bring fluoride of silicon in 
contact with alumina, that is, w ith a compound of aluminium and oxy- 
gen. A reaction takes place between those two substances at this tem- 
perature, whereby silicate of alumina is produced. The mineral, stauro- 
lite, happens to be this particular silicate of alumina, and fluoride of 
aluminium is evolved. That descending, and coming in contact with the 
stratum of silica beneath, undergoes a change similar to that which the 
fluoride of silicon underwent in the first instance, and we get the same 
mineral formed in the second layer from silica as was formed in the first 
]a3'er from alumina — in one instance by the action of fluoride of silicon, 
and in the other by the action of fluoride of aluminium. At the second de- 
composition there is fluoride of silicon evolved, which acts upon the alumi- 
nium below exactly as in tlie first instance ; and so it goes on in succession, 
and at length you obtain the tube full of the mineral staurolite — this sili- 
cate of alumina, and yet you have only employed a small amount of fluoride 
of silicon in the first instance to eflect this transformation. Thus, by a 
small amount of this body, you can convert an indefinite quantity of alu- 
mina and silica into this mineral. As much fluoride of silicon finally 
escapes from the tube as entered it in the first instance. It is one of the 
most beautiful and striking experiments in the whole of this department 
of science. The staurolite produced in this way is crystallized. 
AVe pass on to a few remarks on another curious and important mineral, 
— namely, topaz. This is a compound, the precise rational composition of 
which does not appear to be very clearly understood even at the present 
time. It is essentially a silicate of alumina containing fluorine, but some 
doubt is entertained as to the exact mode in which that fluorine exists in 
the compound. 
Now, one might have reasonably anticipated that topaz might have been 
produced in some such way as that just explained with regard to the pro- 
duction of staurolite ; but, according to Deville, who tried the experiment, 
topaz cannot be so formed. There is, however, some little discrepancy in 
the statements which have been made on this point ; for Daubree states 
that it is formed by heating alumina to redness in a current of fluoride of 
silicon. We must wait for further information, seeing that these two 
chemists disagree on the subject. But if topaz cannot be so formed, a pe- 
culiar mineral termed " zircon," which is a silicate of zirconia (zirconia cor- 
responding in foi-mula with alumina), may be produced in beautiful crystals 
when the fluoride of silicon is passed over zirconia. Take a tube contain- 
ing zirconia, and heat it to a good red-heat, or more than that, and then 
pass over the gas fluoride of silicon, and you obtain the mineral crystallized 
