3 1 8 Progress of Foreign Sckrice. 



at the same time thatthe iron combines with the carbon, it absorbs 

 also a small quantity of silicium. If this fact, which at present 

 is somewhat doubtful, were established by a great number of 

 analyses of iron, before and after cementation, we could conceive 

 the utility of certain substances which are added to the charcoal 

 in the chests of cementation*. By exposing a pure iron, which 

 contained no silica, to a melting heat, in contact with lime and 

 clay, he obtained a portion of the steel of Clouet, which consist- 

 ed, by analysis, of iron 99.2 silicium 0.8. From 100 parts, 1.6 

 parts of silica were obtained. There was no carbon in it. This steel 

 has the property of hardening with water in the usual way, but 

 it is not trusted to by workmen. Ten grammes of slender iron 

 nails (which yielded no residuum after solution in sulphuric) 

 were fused in a Hessian crucible. The brilliant melted button 

 was more difficult to hammer and to file than the iron employed. 

 It consisted, in 100 parts of iron 99.46, silicium 0.54 ; which 

 last yielded 1.08 of silica. 



We cannot therefore judge of the degree of fusion of iron 

 when this operation is conducted in a Hessian crucible, since it 

 seems to be sufficiently demonstrated, that at a high tempera- 

 ture, the iron reduces the silica, combines with the silicium, and 

 forms a compound more fusible than the metal. The same thing 

 happens to platinum, when it finds silicium ready reduced ; and 

 if it does not fuse in a Hessian crucible like iron, it is because 

 from its slight affinity for oxygen, it cannot, like iron, decompose 

 silica. In a creuset brasqnc, steel melts first, then silicated pla- 

 tinum, and lastly manganese, in the same slate of combination 

 with the other metals. M. Boussiiigault infers that pure manga- 

 nese is more difficult of fusion, than pure platinum. 



On some Combinations of Gold, by M. J. Javalf. The object 

 of this paper is to prove by experiment, that M. Pelletier's late 

 notions concerning the equivalent number and combinations of 

 gold are erroneous. M. Javal shews by varied researches on 

 the oxide of gold, obtained by adding barytes to the chloride 

 of that metal, that it consists of 11.909 oxygen united to 100 

 of metallic gold ; a result which approaches nearly to 12.077, 

 that obtained by Berzelius, by decomposing the neutral muriate 

 of gold with mercury, and estimating from the quantity of gold 

 obtained, of mercury which supplanted it, and rate of oxidation 

 of mercury, the rate of oxidation of gold. M. Oberkampf had 

 shewn that the sulpauret of gold obtained by precipitating this 

 metal from the muriate, by sulphuretted hydrogen, consisted of 

 24.39 sulphur + 100 gold. Now SA^ = 12.195, is very 

 near the preceding number. 



The yellow crystals which are obtained by evaporating the 



* There is nothing added at Mouklandto the cbarcual ; it may however 

 contain a little silica. 



t Ann. de Clilm. et de Phys. Vol. xvii. p. 337. 



