584 Dr. Schafhaeutl 07i the Different Species of 



licon combine more closely with the carbon which they im- 

 bibe during cementation. The carburet of silicon partially 

 separates from the carburet of iron, and crystallizes during a 

 slow process of evolving carbonic oxide gas, particularly when 

 no more than one definite carburet of iron is contained in the 

 bar of steel. By these means the peculiar appearance by 

 which the Damascus steel is distinguished is produced*. As 

 Wootz, or Indian steel, is only a cast iron, and has therefore 

 not lost any portion of its silicon or aluminum, which a 

 malleable iron bar always has to a greater or less extent lost, 

 and which it never recovers during cementation, those cry- 

 stals of silicon or aluminum are more developed in Indian 

 steel, than in that made in the common way from malleable 

 iron. As, therefore, iron during cementation is only capable 

 of imbibing carbon, it is evident that the silicon and other 

 electro-negative metals must be already contained in the iron 

 in order to produce good steel, and for this reason certain 

 descriptions of iron only are capable of producing good steel. 

 All iron, and particularly English iron, has during the pro- 

 cess of puddling lost a much greater part of its silicon than of 

 its carbon, and its grains are intermixed with an extremely 

 thin layer of a supei'silicate of iron, which during cementa- 

 tion is only partially reduced, and leaves a different silicate 

 between the grains of the bar, which makes the steel thus 

 produced invariably red-short. Iron prepared from pure ox- 

 ide of iron in so-called German fires, or even by a different 

 process in a puddling furnace, is at large intervals only inter- 

 woven with almost pure protoxide and peroxide of iron, 

 which during cementation is entirely reduced. The carbonic 

 oxide produced occasions those large blisters from which 



* As a proof that a real separation takes place in melted steel between 

 the carburets of silicon and iron, I refer to an experiment made in Mr. 

 Wilson's steel manufactory at Sheffield, when I melted English iron with 

 lamp black and charcoal from sugar. The molten mixture was poured into 

 an ingot mould used for common cast steel, one side of which was hotter 

 than the other. After cooling the metallic parallelepipedon, one half of 

 its cross-fracture was observed to be crystallized, and the other granu- 

 lated. When drawn out under the tilting hammer, a bar was obtained ; 

 one half of its short axis, corresponding to the crystallized side of the in- 

 got, was perfect steel, and the other half the softest iron, just as if a steel 

 bar had been welded to a soft iron bar. The soft part of the bar contained 

 a great quantity of carbon, but merely traces of silicon, too small to be 

 weighed. 



In another similar experiment, after the liquid steel had been poured into 

 the ingot mould, a lump of metal the size of a hen's egg remained at the 

 bottom of the crucible. On perceiving this, I put the steel back into the 

 crucible, and re-melted it at the strongest heat I could obtain ; and on 

 again pouring it out, I found the lump, still remaining at the bottom of 

 the crucible quite unaltered ; 35 grains of this lump, dissolved in hydro- 

 chloric acid, left only 0198 grains of a grayish residuum. 



