and. on its Presence in Steel. 187 
dissolved: this was evaporated, and the dry salt redissolved in 
water, leaving a white powder behind. 
This powder was then heated in a silver crucible with three 
parts of potash, in which it readily melted, and the alkaline mass 
easily and totally dissolved‘in water, with the exception of some 
minute fragments of platina separable by the filter. Sulphuric 
acid poured into the filtered fluid gave a white gelatinous preci- 
pitate, which was evidently silex. It is probable, therefore, that 
the wood-charcoal (which yields by combustion 2 or 3 per cent. 
of ash, chiefly siliceous) furnishes the silex that unites with the 
platina during the cementation, probably in the form of silicium, 
every circumstance being, favourable for the reduction of the si- 
lex into its metallic basis. The silicium is not furnished by the 
crucible, for the cementation of the platina takes place equally 
well when a pretty large crucible is employed, and stuffed full 
of charcoal, with only a small cavity in the middle of it, to re- 
ceive platina. The increase of weight thus acquired by the pla- 
tina is very trifling. It is necessary not to use too much platina 
relatively to the quantity of charcoal, otherwise the fusion goes 
on very imperfectly, or not at all. 
To ascertain further, whether the wood-charcoal furnished the 
silex, I repeated the experiment, using Jamp-black instead of 
common charcoal; but the platina returned. from the crucible 
unchanged, and quite ductile. 
To judge of the quantity of silicium absorbed by the platina 
in the above-mentioned process, 1 took in one experiment ex- 
actly five grammes of platina, and after fusion the button weighed 
5-025. One gramme of this button gave on analysis ‘O10 of 
silex. If the silex were in the state of earth in the metallic but- 
ton, one gramme should have yielded only 005, and therefore we 
must admit that it alloys wit the platina in the state of silicium, 
and that it absorbs -005 (or its own weight) of oxygen by so- 
lution in aqua-regia, whereby it passes into the state of silex. 
These are the proportions which I have assumed in calculating 
that of silicium as it enters into the composition of steel. 
On Silicium in Steel. 
‘The conversion of iron into steel is attributed to carbon alone ; 
and this opinion, supported by the experiments of Monge, Ber- 
-thollet, and Vandermonde, has been generally adopted by all 
chemists who have turned their attention to this subject. It is 
true that carbon is always found in steel ; but another product, 
silex, which is as constantly obtained in the analysis of steel, and 
sometimes in as large a quantity as the carbon, has been usually 
considered as accidental. . I have, therefore, expressly sought Ne 
Aa2 the 
