STEEL.—Sulphur. Phosphorus. Manganese.

879

This steel could be heated and rolled, and wasvery fine in the grain. The tensile strain of therolled bar was 48 tons per square inch, but it onlyshowed 3 per cent, elongation. Steel containingmuch silicon will not take a coating of tin. Ananalysis of a sample of steel containing -833 percent, silicon, and very little carbon or manganese, isgiven in the table. It was found to be impossibleto clean the surface of this steel so as to enableit to take a regular coating of tin. Sulphuric acidof various strengths, sulphate of copper, andchloride of zinc were tried for cleaning the surface ;but in all cases the tin only adhered in patches,while there appeared to be a film on the surface ofthe metal, which was doubtless silica, and this in allprobability prevented the tin getting to the cleanmetallic surface. The presence of silicon in rail steelmay be detected during the process of rolling; asif there is even a very small quantity of siliconpresent, the fine scale produced in the finishingrolls adheres to the rail, or falls off in flat dryflakes, while if the steel contains only a trace ofsilicon this oxide leaves the steel freely, and falls offin curly scales. It is very probable that silicon maytend to prevent the oxidation of steel by forming aninsoluble coating on its surface, in the same waythat oxide of zinc and carbonate of load protect thesurfaces of zinc and lead respectively from furtheroxidation.

The presence of an excess of silicon in steel mayalso be detected by an examination of the fracture,which gradually becomes of a light chalky colour andfine grain as the silicon increases. Silicon is oftenleft in the metal when all the carbon has beenburnt out, both by the Bessemer and Siemens process,if the pig iron contains an excess to start with. Thusit is important to keep the silicon in pig iron downto about 2 - 2 - per cent, in pig iron to be melted in thecupola for the Bessemer process; and if taken fluidfrom the blast-furnace, even this quantity is toogreat, while the iron will blow sufficiently hot in thelatter case if it contains no more than lj per cent.It is pretty well proved that as carbon and man-ganese increase silicon decreases in pig iron.

Sulphur .—This element is extremely deleterious insteel that has to be worked, as the most minuteproportions cause decided red shortness. One-tenthof 1 per cent, is so active that it is very difficult toget the steel containing that quantity to roll soundly.It does not appear to be of so serious consequencein tool and cutlery steel, as it does not producebrittleness. It is very objectionable in steel forcastings, as it conduces to unsoundness in the metal.Its effect on the strength of steel has not been wellmade out.

Phosphoms is as active in producing, cold shortnessin steel as sulphur is in causing red shortness, and toabout the same degree, while to a certain extent oneelement appears to counteract the other. Phos-phorus is much more active in producing hardness insteel than carbon is; 1 part of phosphorus producesabout the same absolute hardness that is caused by2 or 3 parts of carbon. Still, if carbon be nearly

absent, it is possible to have about -25 per cent, ofphosphorus present in steel intended for rails with-out its being unduly brittle, and even in plates acertain quantity is admissible provided carbon benearly absent. When carbon is present to theextent of - 3 to -5 per cent., T per cent, phosphorusproduces very decided brittleness. Steel for toolsand cutlery should be very free from phosphorus, asit is not only extremely difficult to temper phos-phorus steel (as may be seen from the tabulatedanalysis of steel punch), but for fine cutlery it isimpossible to have a keen cutting edge with steelthat contains much over '025 per cent, phosphorus.On the other hand, there is no objection to a smallquantity of phosphorus in steel for castings, as ittends to cause the metal to run solid without greatlyreducing its strength.

Mr. HoLLEy has recently made a careful set ofexperiments on the phosphoric steel introduced atTerre Noire, following the material all through theprocess of its manufacture. The following tableshows the composition of the steels, and how theyrolled, &c. Professor Thurston carefully tested allthese steels mechanically, and the following are someof his notes:—“ They are not as ductile as commonsteels. Their higher elastic limit is accompanied bya lower degree of toughness. They are harder, andrather less tough than the standard steels. Com-paring these steels with common steels of similartemper, it would seem that these steels excel usuallyin strength, and in some cases are greatly superiorin tenacity to ordinary steels. When exposed tosudden and excessive shocks they are more liable tobreak. The fractures are peculiar in appearance.”(See Table XIV.)

Manganese is the most useful metal we have as analloy of iron. While it adds slightly to the hardnessof steel, especially when carbon and silicon arepresent, it does not induce brittleness unless otherhardening elements are present. Even with 1 percent, manganese steel may be very tough. But thegreat value of manganese arises from its tendency toneutralize red shortness. Where this is due to theuse of excess of oxygen, the action of manganeseis undoubtedly to remove it, owing to its greateraffinity for oxygen. The oxide of manganese formedtends to the production of a very fluid slag, whichpromotes the soundness of the steel. In the case ofred shortness due to sulphur it is doubtful whetherthe manganese actually removes, or neutralizes, theeffect of the sulphur, or whether it acts only byremoving the red shortness due to oxygen. Itappears that, in addition to removing or neutralizingthe red shortness, an alloy of iron and manganese upto 1-j per cent, of the latter actually works betterthan steel, in other respects similar, would do withthe manganese absent. It was formerly the practiceto add only sufficient spiegel to leave about -25 percent, manganese in the steel for rails, but in Eng-land steel makers find that by using richer spiegel,and more of it, so as to leave about 1 per cent,manganese in the steel, they can use rather lowerand therefore cheaper grades of iron ; the excess of