Steel, a malleable compound of iron and carbon, which may be hardened and tempered. Considerable confusion in the use of the word has arisen in late years, owing to the introduction of improved metallurgical processes, whereby wrought or malleable iron may be melted and cast into ingots. These ingots, having the appearance of ordinary cast steel and some of its properties, have likewise received the name of steel, although they lack the capacity of hardening which hitherto was regarded as the essential characteristic of steel. Pure or wrought iron possesses a high degree of malleability and ductility, is difficultly fusible, may be welded at high temperature, but below fusion, and is soft enough when cold to be readily wrought with tools. By the gradual addition of carbon to iron we notice an increase in fusibility, hardness, and resiliency, while malleability and ductility decrease. The smallest proportion of carbon which will distinctly produce these effects is about 0.25 per cent., and the largest amount of carbon which can exist in iron without destroying its malleability is about 2 per cent. Within these limits the compounds of iron and carbon possess the property of becoming soft when heated to redness and slowly cooled, and of becoming hard again when heated and quickly cooled.

These processes of hardening and annealing may be repeated indefinitely, or until the carbon is burned out by the successive heatings. Iron with more carbon than 2 per cent, (say 2 to 5) is known as cast iron. It is more fusible than steel, but is not at all malleable, and while it may be hardened by sudden cooling, it is brittle and does not possess the resiliency or " spring " of steel. Soft or wrought iron has been until within the last 20 years worked by rolling or hammering when in a plastic condition at a red or white heat, owing to the impracticability of fusing pure iron. Steel was worked in the same manner as wrought iron until Huntsman succeeded in melting it in crucibles during the latter half of the last century, since when cast steel has replaced welded steel for most purposes, on account of its greater homogeneity, since all welded products consist of layers or fibres of metal separated by cinder, which, though it may be largely extruded by rolling or hammering, yet is always present to a sufficient extent to prevent the absolute contact of all the particles of metal.

Since the idea of perfect homogeneity combined with malleability has so long been associated with our notions of steel, it was natural that when malleable iron, or iron low in carbon, was melted and cast in moulds, it should receive the name of steel without regard to the amount of carbon or the capacity for hardening. It is thus that the products of the Bessemer converter and the Siemens furnace have all been classed as steel, although the content of carbon may vary from 1.50 to 0.10 per cent.; and owing to the very large production of metal by these processes, far exceeding in amount ordinary cast steel, this classification has become well established in iron metallurgy. The uncertainty and confusion that has arisen from classing together products of widely different physical and chemical properties, has led to an active discussion of the definition and classification of steel. The classification of Greiner of Seraing is as follows :

While the simplicity and convenience of this classification from a manufacturing point of view must be admitted, its adoption is opposed by Gruner and others on the ground that it takes no account of the capacity for hardening.

Among the elements other than carbon met with in steel are phosphorus, silicon, sulphur, and oxygen among the non-metals, and manganese, copper, tungsten, titanium, and chromium among the metals. Some of these are invariably present in the materials used for steel making, and are usually regarded as impurities in the steel, while others are added to produce certain specific effects. The modifications of the properties of steel by the above named elements have been already treated partially under Iron. Steel is more susceptible to the action of impurities than is wrought iron. This is especially true with regard to phosphorus and silicon, and is readily accounted for by the similarity of action of these substances with carbon. Recent experiments have shown that an amount of phosphorus which would be highly detrimental to steel containing say 0.50 per cent, of carbon, may be present with safety when the carbon is as low as 0.10 or 0.20 per cent., or in other words when the steel passes into soft iron. The effect of this formerly much dreaded enemy of iron and steel has been so thoroughly studied that "phosphorus steels," so called, are manufactured and sold.

Phosphorus makes iron hard, brittle, and coldshort (see Iron), and this is also true in a modified degree of carbon and silicon; hence, when two or all three are present together in iron, the effect is cumulative. The contradictory statements as to the maximum percentage of phosphorus that Bessemer metal will bear find here their explanation. It was formerly said that Bessemer steel with more than 0.05 per cent, of phosphorus was unfit for rails, but later experience has shown that if the amount of carbon does not exceed 0.15 per cent., phosphorus to the extent of 0.35 per cent, may exist without seriously impairing the strength and ductility of the metal. This fact, recently brought into prominence by the manufacture in France of phosphorus steel on a large scale, was recognized in this country as early as 1870. Samples of boiler plate and tough steel made at Trenton, N. J., by the Martin process, showed on analysis the following composition :

While it appears from the above that phosphorus may in a measure replace carbon in steel, the effect of these two substances is not identical, and the limit of rigidity is much sooner reached with the former than with the hitter. The use of phosphorus steel is solely a question of economic advantage, since its manufacture permits the use of impure and consequently cheaper materials; but as far as is at present known, the compounds of iron and phosphorus possess no properties that give them a superiority over the carbon compounds for industrial applications. The effect of silicon on steel appears to be similar to that of carbon, as the general analogy of the two elements would suggest; but to produce a given degree of hardness, the amount of silicon necessary is very much greater than that of carbon - the reverse of the case with phosphorus. The most contradictory statements exist regarding the effect of silicon on steel. The best established data are summarized by Turner as follows: A small amount of silicon is not necessarily injurious to steel, and may be an advantage in those varieties which are to be used without hardening, and where there is no special demand for tenacity and strength.

On the other hand, where steel must be hardened for use, as for tools, silicon can only be injurious, and that in proportion to the quantity present. This is one reason why Bessemer steel cannot generally be used for purposes requiring a line, hard steel; for it is usually made from highly silieious pig iron. But some of the Swedish Bessemer steel, made from pure manganiferous pig iron low in silicon, approximates in quality to ordinary cast steel. A puddled steel made with the addition of a highly silieious iron ore has been brought into prominence under the name of "silicon steel;" but there is no evidence that it derives any of its properties from silicon, or indeed that there is any more silicon in it than in ordinary puddled steel. The effect of sulphur on steel is entirely different from that of the elements already mentioned. It makes it "red-short," that is, brittle when hot; but unlike phosphorus, it does not sen-sibly affect its malleability when cold. The largest amount of sulphur that steel will bear without serious impairment of its malleability is said to be about 0.10 per cent. Oxygen produces the same effect on homogeneous iron as sulphur, as might be inferred from the close chemical relations of the two elements.

It can never exist in the harder steels prepared by fusion, for it would then combine with the carbon; but it is frequently met with in the Bessemer low steels and iron, and makes them red-short. Bed-shortness, formerly ascribed exclusively to sulphur, has been found in very many instances to be due to oxygen. Considerable importance has been attached to the presence of nitrogen in steel, and Fremy considers it an essential ingredient. Numerous analyses do not support this view, and it is probable that its presence in steel is entirely accidental and due to the property which many metals possess of absorbing or occluding gases. - The compounds of iron with the metals, or the true alloys of iron, have not been as closely studied as its compounds with the non-metals, and but little can be said with precision of the physical characters of these alloys as such, or as modified by the presence of the non-metallic elements. The properties of iron are not as radically modified by the addition of small quantities of metals as is the case with the non-metals. Manganese is closely allied to iron in its chemical properties; and it plays a very important part, and one in many cases not well understood, in the preparation of steel. Ores containing considerable manganese are often known as steel ores.

The beneficial effect of the addition of this metal or its compounds in the manufacture of crucible steel was discovered by Heath in England (patented in 1839), but the nature of its action is still somewhat obscure. Recent experiments by Caron show that sulphur is at least in part removed by manganese, and it is also probable that the presence of this metal in the steel prevents the injurious action of sulphur, although as a rule manganese added as oxide in the crucible does not enter into the composition of the steel. The part that it plays in the Bessemer and Martin processes is well understood, viz.: it removes the oxygen that the molten metal has absorbed, and thus corrects red-shortness, and it is probable that its favorable effect may be in many other instances referred to this action. The affinity of manganese for oxygen is much greater than that of iron, and therefore the reduction of metallic manganese from its oxide is accomplished with more difficulty; consequently a manganese cinder, unlike an iron cinder, protects the carbon of steel from oxidation. Further, when manganese is present in molten metal, the oxidation of the iron is prevented until all or nearly all the manganese is oxidized.

Below a certain amount, say 1 per cent., manganese has been shown, contrary to former opinions, to exert no disturbing effect on the properties of steel. Greiner says that manganese possesses the property of rendering steel very malleable and weldable, and that hard steels containing sulphur, phosphorus, and carbon (as high as 1.50 per cent.) can be forged with ease if they contain sufficient manganese. Nearly all observers agree that it is a corrective of red-shortness. Manganese steel, so called, was for some time made in Leoben, Austria, but its manufacture was abandoned, owing to the uncertainty of the product. The combinations of steel with chromium, tungsten, and titanium have attracted much interest from the fact that they appear to have peculiar and valuable properties. They are often represented to be steels in which the metals named replace carbon. This view is inadmissible from a chemical standpoint, and • it is probable that these compounds are carbon steels modified in their properties by the presence of other metals.

An analysis of Mushet's "special" steel, generally supposed to be made from titaniferous ores, showed the presence of tungsten and no titanium, viz.: tungsten, 7.98 per cent.; carbon, 1.40; silicon, 0.24. This compound is excessively hard under ordinary conditions; on sudden cooling it breaks, and it can only be worked at a very low red heat. Tungsten has also been added in the Bessemer converter, and the resulting steel, containing as high as 0.558 per cent, of tungsten, was found to combine a high degree of toughness and capacity for hardening. Tungsten likewise increases the power of steel to retain magnetism. Titanium seems to produce effects similar to those of tungsten. Chromium also appears to confer valuable properties on steel, somewhat resembling those produced by tungsten. The "chrome steel" manufactured in Brooklyn, N. Y., combines in a high degree tenacity and ductility, and is capable of bearing a high heat for rolling, hammering, and welding. It is highly carburized, one sample giving 0.98 per cent, of carbon, and another 1.23. The amount of chromium found in one analysis was 1.66, and in another it did not exist in appreciable quantity.

Determinations of the tensile strength of this steel by Kirkaldy of London, on bars 5 in. in length, varied from 115,780 lbs. to 167,320 lbs. per square inch, with an elongation in the first instance of 11.6 per cent., and in the second of 7 per cent. Determinations made at the West Point foundery ranged from 173,770 lbs. to 198,910 lbs. per square inch. When hardened at a very low heat, it acquires great hardness; a high heat renders it brittle, as might be expected from its large percentage of carbon. Copper is sometimes an accidental ingredient in steel. It seems to make it red-short, and its presence even in small amount is believed to be highly deleterious. Faraday and Stoddart have experimented on alloys of steel with the noble metals. They found the compound of steel with a small amount of silver to have valuable properties, but its expense would be a barrier to its introduction. Many analyses of line steel have shown the presence of aluminum; and it is not improbable that this metal exerts a favorable action on steel, but the subject has not been investigated. - Classification and Manufacture of Steels. For the purposes of description of steels and the processes of their manufacture, a classification based on the mode of production may he found convenient.

When iron ore is used, the process is one of deoxidation and subsequent carburization; with pig iron it is one of decar-burization; and with wrought iron it is one of carburization. The following outline of processes is arranged on this plan:

1. Steel from the ore direct, by reduction and carburization. Ore steel, direct steel. Example, bloomary steel.

2. From pig iron by decarburization.

a. By solid oxidizing agents, as iron ore, saltpetre. etc, without fusion. Examples, puddled steel. Ileaton steel.

b. By solid oxidizing agents with fusion. Example, Uchatius steel.

c. By the oxidizing agency of air, with fusion.- Example, Bessemer steel.

d. By oxidizing and reducing gases. Example, the Berard process.

3. From wrought iron by carburization.

a. By fusion with pig iron. Example. Martin steel.

b. By fusion with coal or carbonaceous substances. Example, Indian steel or wootz.

c. By heating in charcoal without fusion. Example, cement steel.

d. By heating in earburctted hydrogen, without fusion. Example, Mackintosh or Baron steel.

The distinction between crude and fine steel is not now so sharply defined as formerly, but in general the term fine steel is reserved for those products made by fusion of the purest materials in a crucible, and particularly for the cast steel made by fusion of cement steel. Shear steel, produced by welding and rolling cement steel, may also be classed here. In many instances two or more reactions or processes for steel making are combined, but in following the above classification the principal feature only of the process is considered. - 1. Steel direct from the Ore. The process for obtaining wrought or soft iron direct from the ore (see Bloomary, and Iron) affords, with some slight modifications of charging and manipulation, a product containing sufficient carbon to entitle it to rank among the steels; but the steel thus produced is always low in carbon, and may be classed with puddled steel. In the Catalan or bloomary forge the circumstances favoring the production of a steely product are : a slow process, that the reduced iron may have time to absorb carbon; the protection as far as possible of the mass of plastic metal from the direct action of the blast and from the action of rich iron cinder; and also the use of manganiferous ores, since the oxid© of manganese, as already explained, does not oxidize carbon readily.

The character of the steel produced by the bloomary depends on the nature of the ores and the skill of the workmen. The impurities of the ores are more completely eliminated in the direct processes than in the blast furnace process, a gain obtained at the expense of a considerable loss of iron. Titaniferous ores can be successfully worked in the bloomary, and are said to give a superior steel. The product of the bloomary generally lacks uniformity, a defect which can be remedied by repeated heatings and hammering. The bloomary process is rarely used now for the direct production of steel, but the iron made by this process is, on account of its purity, advantageously employed for conversion into steel by the cementation process. - 2. Steel from Pig Iron. 2a. Puddled Steel. The process of puddling for steel does not differ essentially from that for iron. (See Iron Manufacture.) The operation is stopped before complete de-carburization, or when the desired hardness is attained. The conditions favoring the production of steel in the puddling furnace are as follows: 1. Pure pig iron is necessary, since the refining is not carried as far as with wrought iron, and also because a less basic cinder is employed. 2. The pig iron should be highly carburized, that it may not come to nature too quickly. 3. It should not contain too little silicon, or the cinder will be too basic or "rich." 4. The presence of manganese is favorable, as it produces a fluid and non-oxidizing cinder. 5. The cinder should be "poor " or highly silicious, since rich cinder decarbn-rizes the metal. 6. The process should go on slowly, that it may be the more under control. 7. During the balling the temperature must be reduced as much as possible, and a smoky flame produced, to prevent oxidation.

The nature of puddled steel has already been considered in the foregoing. It possesses a degree of hardness proportional to the amount of carbon (which rarely exceeds 0.5 per cent.), and a fibrous or welded structure. It has been largely used for the heads of rails, being much more readily welded to iron than ordinary cast steel. Puddled steel made from pure pig irons is also much used for melting in crucibles for the production of fine cast steel. Saltpetre (potassium nitrate) has been used for the conversion of pig iron into wrought iron and steel. Its action is both oxidizing and purifying, the former through the large amount of oxygen of the salt which is readily given off, and the latter through the strong base, potassa, which combines with the silicic and phosphoric acids produced. The apparatus devised by Heaton for the reaction between molten cast iron and saltpetre, for the conversion of poor pig iron into good steel, has been abandoned on account of the expense and uncertainty of the process. 2b. Uchatius Steel. Steel produced by the reaction of pig iron and iron oro may be obtained in the molten condition, if the temperature of production is high enough. This is effected either in a crucible or in a Siemens regenerative furnace.

Uchatius steel, named from its inventor, is made by fusing a mixture of granulated pig iron, iron ore, and oxide of manganese in crucibles. Its manufacture is now confined to Sweden (though originally introduced in Austria), where the pure irons and ores are admirably adapted to the process. At the Siemens works in Lan-dore, Wales, the manufacture in the open hearth is regularly carried on by mixture of pig iron and iron ore. The process differs from that of Uchatius in that the ore is added in successive portions, and that to the decarbu-rized metal spiegeleisen is added, as is usual in the open-hearth processes. Scrap iron is also sometimes added, but its use is not essential to the process. The charge consists of 5 to (5 tons of Bessemer pig iron and 30 cwt. of pure ore. The product is used principally for rails, and averages 0.40 per cent of carbon. 2c. Bessemer Steel. The Bessemer or pneumatic process consists in the removal of the carbon, silicon, etc, from pig iron by means of a blast of air forced through the molten metal.

The reactions involved are in many respects the same as those in the puddling process; that is to say, the silicon is first oxidized, and the silica thus formed combines with the oxides of iron and manganese (if present) to form a cinder, and the carbon is subsequently oxidized to carbonic oxide. Owing, however, to the rapidity of the process and the large amount of pig iron employed, the heat developed in the oxidation of the silicon, carbon, etc, is sufficient to retain the resulting steel or iron in a fluid condition, so that it can be cast directly into moulds. The history of this remarkable process is briefly as follows : On Oct. 17,1855, Henry Bessemer patented a process of blowing air or steam through molten pig iron in crucibles, until the metal was decarburized to any desired extent. At this time he recognized the fact that while steam cooled the metal, air increased the heat from red to white. A patent in December of the same year specified a circular or elliptical vessel provided with a refractory lining and hung on trunnions, which could be filled and emptied by means of a lipped opening. In this patent the essential features of the process were fully developed.

A patent of Feb. 12, 1856, indicated that the heat developed in the process was sufficient without the additional use of fuel, and that, according to the duration of the blowing, steel or soft iron might be produced. In July, 1856, Bessemer read a paper before the British association at Cheltenham on the " Manufacture of Iron and Steel without Fuel," which created an intense interest. But the subsequent trials did not yield uniformly satisfactory results, and except by the inventor the process was practically abandoned. Patient and careful experiments showed that not all pig irons were adapted to the process; that sulphur and phosphorus were not eliminated, and consequently pig irons containing a notable propor-* tion of these substances could not be used. Again, the interruption of the process at the precise point of decarburization desired was found to be impracticable, owing to lack of trustworthy indications. Further, it was found that the process was not adapted to making the finer and harder steels, but had its chief application in the production of low steels or soft iron.

The absorption of oxygen, and the consequent red-shortness of the metal when the pig iron was blown to nearly complete decarburization, was overcome by the addition of spiegeleisen, a white pig iron containing from 7 to 12 per cent, of manganese. This was a suggestion of Robert Mushet, and • to it the practical success of the Bessemer process is largely due. After conquering all the obstacles to success, Bessemer did not find a ready acceptance of his process owing to the distrust caused by his previous failures. He therefore started in 1859 a small establishment of his own in Sheffield for the regular manufacture of his steel. His commercial success soon led to the general adoption of his process throughout the civilized world, more particularly at first in Sweden, where the pure ores and fuels furnished a pig iron admirably adapted to the process. In 1867 there were in England 52 Bessemer converters, in Prussia 22, in France 12, in Austria 14, in Sweden 15, and in Belgium 2. In 1873 Germany alone had 70 converters, and the number had risen in England to 105. The production in England has increased from 6,000 tons in 1867 to 540,000 tons in 1874. - The Bessemer process consists, first, in melting the pig iron; second, transferring the molten metal to the converter, where it is subjected to the action of the blast of air; third, pouring the finished product into a ladle; and fourth, pouring from the ladle into the mould.

The metal when solid, but while still hot, is taken from the moulds and worked by rolling or hammering into the desired form. Pig iron is in some cases used direct from the blast furnace, but remelting is generally found advantageous. The furnaces now used for this purpose are generally cupolas, which melt quicker and are more economical, although the direct contact of the iron with the fuel may cause a deterioration of the metal if the fuel is impure. The reverberatory furnace is not open to this objection, but the pig iron may here suffer a loss of silicon and manganese, owing to the oxidizing atmosphere. The molten metal is either run in troughs directly from the furnace to the converter, or is first run into ladles where it can be weighed, and thence carried to the converter. The latter is a pear-shaped vessel, sometimes called the retort or simply the vessel, consisting of an iron mantle lined with a refractory silicious material. It is usually made in two parts, upper and lower, for convenience of lining. The bottom, which contains the tuyeres, is made in a separate conical piece, and inserted from below. The size of the converter is usually calculated for a charge of five to six tons of pig iron.

This amount of metal occupies but a small part of the vessel, as is indicated in the accompanying figures. The greatest external diameter is about 8 ft., with a total height of from 12 to 15 ft. The silicious material of the lining usually contains a little alumina. The so-called "ganis-ter" used in England for this purpose is a ground sandstone found in the coal formation, containing 93 per cent, of silica, 4 per cent, of alumina, and 1 to 2 per cent, of oxide of iron. The lining is made by ramming the material in a moist condition around a form placed in the converter. It is usually about 12 in. thick. The greatest attention must be paid to the selection of the material for the lining and to its thorough consolidation, for upon the lining the success of the process largely depends. The tuyeres, from 7 to 12 in number, are made of fine clay in the form of truncated cones, each perforated with 7 to 12 holes about three eighths of an inch in diameter. They are arranged on the bottom plate, and ganister or other material stamped around them; and the finished bottom, after drying, is inserted in the converter. The bottom lasts generally for 6 to 10 heats, while a carefully made lining may endure 1,000 or more heats.

The converter is mounted on trunnions, one of which is hollow and conveys the blast to the tuyere box below the tuyeres, and to the other is attached the mechanism by which the converter is revolved. Figs. 1 and 2 give sectional views of the converter in two positions. Fig. 3 is a plan of the converter with the rotating machinery. The ladle into which the steel is poured from the converter is shown in figs. 4, 5, and 6. Fig. 4 is a vertical section of the ladle crane and elevation of the ladle. Fig. 5 shows the platform on which the ladle moves, and fig. C is a partial section through the ladle, showing the loam-coated rod which acts as a stopper in pouring. By this latter arrangement the fluid steel is discharged in a thick stream, and the cinder remains on top. The steel is usually cast in long ingots ahout 12 to 14 in. square at the base and tapering from 1 to 14¼ in., each ingot being rolled into two or three .rail blooms. When the steel is intended for other purposes than rails, moulds of special forms are used.

To obviate the occurrence of air bubbles in the steel, caused by the falling of the stream from the top to the bottom of the mould and spattering against the sides, bottom casting is employed; that is, pouring the steel down a central sprue and causing it to enter the bottom of several moulds at a time through fire-clay distributors. The blowing engine for supplying the blast is usually double, and should be able to deliver 8,000 to 11,000 cubic feet of air a minute at a pressure of 25 lbs. to the square inch. Probably no other invention of the magnitude of the Bessemer process ever came from the hands of its inventor in as complete a form. But while the accumulated experience of 15 years has added nothing to the essential features of the apparatus and machinery, yet in the minor details of construction improvements have been made which have increased the capacity of the process four fold. The highest perfection of apparatus and working has been attained in the United States, where there are now (1876) ten works with two converters each, of five to six tons capacity. The improvements in American practice have been largely due to Mr. A. L. Holley, who has superintended the construction of most of the works in this country.

In 1868 an output of 500 tons a month from two five-ton converters was barely reached. The production had gradually increased to 4,200 tons of ingots a month in the best works, in others to 3,800 tons, and in one instance to 5,000 tons. In the nominally five-ton vessels 5½ to 5¾ tons are sometimes produced at a heat. The improvements which have rendered this large and regular production possible in this country, far ex-ceeding that of European works, have been summed up by Mr. Holley as follows: 1, improved cupola furnaces and method of working; 2, the means used for quickly and soundly renewing the vessel bottoms, and the use of fire brick around the tuyeres; 3, more roomy and convenient arrangement and distribution of the working parts and spaces; 4, filling the ingot moulds from the bottom by improved and convenient apparatus. - The converter, after being lined, is thoroughly dried and heated to redness, and pig iron is run into it while turned to the'horizontal position. On tipping up the converter, it is necessary that the blast should be started before the metal reaches the tuyeres. This is effected automatically by a cam on one of the trunnions. When the converter has attained the upright position, the roar of the air rushing through the metal and escaping from the mouth is heard.

In this stage a large part of the oxygen is absorbed by the silicon and manganese (or iron in the absence of manganese), and the flame is short and not highly luminous. The spectroscope shows at this time a continuous spectrum without lines. Soon the escaping flame increases in size and brilliancy, assuming an orange or yellow color with blue streaks and a white edge, intermingled with sparks of metal. The spectroscope now shows the sodium line, and generally those of potassium and lithium, accidental ingredients of the metal or lining. This constitutes the first period of the conversion, and is known as the slag or cinder-forming period. The action now becomes more violent, and the flame more intensely luminous, and large masses of iron or cinder are often ejected from the vessel, probably from the energetic action of the oxide of iron in the cinder on the carbon of the metal. The spectroscope now shows bands of dark lines in the green, which have been proved to be produced by manganese, though their appearance is dependent on the oxidation of the carbon. This violent stage of the process passes gradually into the third and more peaceful period, in which the flame increases in heat and brilliancy and assumes a purple or violet tint.

At this high temperature the carbon appears to be directly oxidized by the blast. When the carbon is all removed the flame suddenly drops, which is the indication for tipping over the converter and stopping the blast. Coincident with the dropping of the flame is the disappearance of the dark bands from the spectrum. The length of the process up to this point may vary from 5 to 45 minutes, according to the heat of the metal, the amount of silicon and manganese, and the amount of pressure of blast. A "blow" usually lasts 15 to 20 minutes, of which the first or slag-forming period generally occupies one half. Pig irons with little silicon often pass directly into the second period. The metal in the converter after complete decarburization contains considerable oxide of iron in suspension or solution, and in that condition is worthless, since it breaks up under the hammer. There is added to it, therefore, metallic manganese, as before explained, which combines with the oxygen and passes into the cinder. Spiegelei-sen is generally used for this purpose. At the end of the blow the converter is tipped over, and the spiegeleisen, previously melted, is run in. An energetic action at once manifests itself by the escape of abundant gas and flame.

About 7 to 10 per cent, of the weight of the charge is used, according to the hardness of steel desired. Spiegeleisen contains about 4 to 5 per cent, of carbon, and the amount that can be used is therefore limited, for the carbon, taking but little part in the reaction, enters into combination with the metal. This has been an obstacle to the preparation of extra soft metal by the Bessemer process. Ferro-manganese, a combination of manganese and iron with a little carbon, containing 50 per cent, more or less of manganese, was early used with success in the process, but its manufacture was abandoned owing to its expense. It has recently been revived and its use resumed in the Bessemer, but more particularly in the Martin process, under analogous conditions, for the preparation of steel or homogeneous iron containing phosphorus. The employment of ferro-manganese is also becoming general for making soft iron of fine quality for construction of ships, bridges, etc. The addition of spiegeleisen or ferro-manganese is not universally practised. Where the pig iron contains considerable manganese, the process may be interrupted at the desired stage of decarburization, and even metal very low in carbon, which is not red-short from oxide of iron, may be successfully cast.

This method is followed in Sweden and some parts of Germany. The loss of weight in the conversion of pig iron by the Bessemer process, including scrap, is from 10 to 15 per cent. The heat produced in the process, formerly supposed to be mainly caused by the oxidation of the carbon, is now known to be mainly due to the oxidation of silicon and manganese, and also of the iron. Silicions pig iron is therefore generally demanded for the process. From 1½ to 2 per cent, of silicon is the amount generally desired, but pig irons with more and less are often used. The use of more silicions pigs is disadvantageous owing to a lengthening of the process, and also to the large amount of silicon remaining in the steel. When highly manga-niferous pig is used, the silicon may sink be-' low 1 per cent., and the resulting steel is of a much finer quality. Much of the Swedish Bessemer steel, celebrated for its purity and strength, is made from pig iron of this character. - The heavy ingots of Bessemer steel intended for rails are either hammered or rolled (bloomed), becoming thereby condensed and elongated, and then cut into lengths suitable for rolling into rails. Blooming is now generally conceded to make the best and most uniform, product.

The American blooming train consists of three rolls 30 in. in diameter and 5 ft. in length, which are adjustable in housings by means of steel screws. Ingots 12¼ in. square are reduced by four grooves and 17 passes to 6 or 7 in. square in four minutes. Special appliances for manipulating these heavy masses of metal by machinery are attached to the rolls and greatly facilitate the operation, which in some cases is nearly automatic. The rail trains are ordinarily three high rolls. (See Iron Manufacture.) A 21-inch train for rolling 7-inch ingots into rails in 13 passes is divided into three lengths. The product of a steel rail mill, working on 7-inch blooms, is about 1,000 tons of rails a week. The consolidation of steel usually accomplished by hammering or rolling may also be effected by the application of a heavy steady pressure. This latter method is applicable not only to the forging of masses of steel, but also to the compression of the metal while in the molten state. Bessemer embodied this idea in one of his earlier patents.

Originally practised in France, the compression of liquid steel has attained its greatest development in England, where Sir Joseph Whitworth has an extensive plant for this purpose, which includes four hydraulic presses capable of exerting a pressure of 2,000 to 8,000 tons. The pressure usually applied is six tons to the square inch, by which an ingot is reduced one eighth in length. To small castings a pressure of 20 tons to the square inch is sometimes applied. Mild steels treated by this process have shown a tensile strength of 40 tons to the square inch, with an elongation of 30 per cent. A tube of this compressed steel 26 in. long and 7.83 in. in diameter, with a bore of 2.56 in. (being that of a nine-pound field gun), sustained 48 explosions of 1½ lb. of powder with the bore closed by a screw plug, the only escape for the gases being through the touch hole, 1/10 in. in diameter. The expansion of the bore increases at every explosion, but without rupture. Forging steel by means of hydraulic pressure was first introduced by Haswell in Vienna in 1861. Heavy ingots are forged by this method more effectually than by hammering, and smaller articles of irregular or intricate outline, up to 150 lbs. or more, may be directly formed by pressure of the white-hot metal into moulds. 2d. The Berard Process. The conversion of pig iron into steel or soft iron by means of oxidizing and reducing gases, in this process, is carried out on the hearth of a reverberatory furnace heated by gas.

The pig iron is decar-burized by means of air in connection with hydro-carbon gases, which are expected to remove the sulphur and phosphorus. The resulting iron is recarburized by the reducing gases. This process has not yet proved a commercial success. - 3. Steel from Wrought Iron. The above described processes under the second division of the classification, to which many others of minor importance might be added, all use pig iron as the principal material for the preparation of steel; and as it is a substance of complex and variable composition, the quality of the steel derived from it will depend on the composition of the pig iron used. In none of the processes using pig iron is there a complete elimination of all the substances associated with the iron; hence only the purer varieties can be used where a good product is desired. In the third division wrought iron is the principal material used, and as this may be made in a state of great purity even from moderately pure pig irons, the steel made from it is as a rule superior to that made from pig iron.

Wrought iron when imperfectly worked contains considerable cinder, which holds the greater part of the phosphorus originally in the pig iron; and when steel is made from such wrought iron by fusion, the phosphorus enters the steel. 3a. The Martin Process. The principle of manufacturing steel by the reaction of wrought iron upon melted pig has long been known. Rinmann, Vana'ccio, and even Agricola (about 1550) describe processes of this kind. Reaumur (1722), Chulut, and Clouet (1778) published experiments in which steel was produced by the simultaneous fusion of cast and wrought iron, or of cast iron and iron oxide. But these experiments, and many others of subsequent date, were successful only so far as the manufacture in crucibles was concerned. It was only in closed vessels, heated from without, that the necessary high temperature, combined with exclusion of air, could be maintained. Vitreous fluxes were early used, to protect the surface of the molten metals; and the idea of employing a reverberatory furnace is found in the work of Hassenfratz (1812). Several English and French patents of the early part of this century show that metallurgists were actively engaged with this problem.

The most important historically, though at the time without commercial results, .was that of Heath (1845), which indicated the fusion of material in a hearth, the maintenance of an extremely high temperature, and the employment of gaseous fuel. In a former patent (1839) Heath had claimed the addition of carburet of manganese. The oxides of manganese had been previously used in metallurgy; but the introduction of metallic manganese, alloyed with carbon, was an important novelty, which prefaced the employment by Mushet, Bessemer, and Martin of the "triple compound" of iron, carbon, and manganese (spiegeleisen). The chief difficulty with all these attempts to manufacture steel by fusion in the reverbera-tory was the lack of efficient and economical means for the maintenance of the intense temperature required. This was supplied by the important invention of Siemens, the regenerative gas furnace (see Furnace), in which the use of gas as fuel was perfectly realized. The effect of this invention was great and immediate in every branch of metallurgy involving very high temperatures, and nowhere more signal than in the remelting and subsequently in the direct manufacture of steel by fusion.

Sudre, Alexandre, Attwood, and Brigues, Ram-bourg and co. (the last at Montlucon, under the advice of C. W. Siemens himself) attempted with the aid of the new system of heating to fuse cast iron with wrought iron or oxides of iron for the production of steel, and procured patents on the strength of their experiments. But the first practical success was that of Pierre and Emile Martin, whose method is set forth in their patents of 1865 and 1867. These metallurgists, after a series of experiments extending over many years, arrived at a combination of features, most of them separately known before, but constituting as a whole a new process, by which they were enabled to manufacture open-hearth steel of all grades, from the homogeneous metal approaching wrought iron to the hardest varieties, on a commercial scale and with profit. Naturally their claims as inventors, among so many eager competitors, were for a time contested; but the report in their favor of MM. Jordan and Burat, made in November, 1874, after an investigation extending over many months, for the tribunal of the Seine, will probably be accepted as conclusive.

The Martin process is now widely employed in England, on the continent of Europe, and in the United States, and constitutes the only rival of the Bessemer method for the production of cheap steel. It consists essentially in the decarburization of cast iron by fusion with wrought iron, iron sponge, steel scrap, or iron oxide, in the hearth of a reverberatory furnace, heated with gas, the flame of which assists the reaction, and the subsequent recarburization or deoxidation of the bath by the addition at the close of the process of white iron, spiegeleisen, or ferro-manganese. The period of fusion and decarburization lasts from four to eight hours; the amount of spiegeleisen or ferro-manganese added depends upon the condition of the bath, the grade of steel desired, and the percentage of manganese in the alloy used. The first of these, elements is determined by samples taken from time to time during the process and tested. The advantages claimed for the Martin as compared with the Bessemer process are: its less expensive plant; the greater duration of the operation permitting by means of sampling more complete control of the quality of the product, and also conducing to greater uniformity of result; and, as a consequence of the foregoing, the practicability of employing materials which have not hitherto been considered suitable for the Bessemer converter.

The greater variety of materials available for the Martin process also renders the direct conveyance of. the molten pig from the blast furnace to the steel furnace an easier matter in this process than in the other, since the initial quality of the pig is of less importance. Yet this direct conveyance of the cast iron has thus far been practised in certain Bessemer works alone. The Martin process has been employed at Terre Noire in France, and by Mr. Slade at Trenton, N. J., for the production of phosphoric steel mentioned above. The production of Martin steel in this country has risen from 3,000 net tons in 1872 to 7,000 tons in 1874. The number of establishments using the process in 1874 was 13, and its introduction was in progress in other works. 35. Indian Steel, or Wootz. This is produced by fusion of wrought iron with coal or carbonaceous substances in crucibles. Small pieces of iron made in the small native furnaces are put into a clay crucible with some dried wood and leaves, and covered securely with tempered clay. The crucibles are then heated until fusion is complete, when they are broken open, and a conical mass of steel weighing 2 or 2½ lbs. is obtained. This steel is generally very highly carburized, and requires to be worked at a low heat.

It is much esteemed for its purity, but the production is small in amount. About the beginning of this century David Mushet carried out an extensive series of experiments on the fusion of wrought iron and charcoal in crucibles, and determined the amount of charcoal necessary for the production of steel of different degrees of hardness. Since then numberless patents have been secured for mixtures for fusing in crucibles, comprising mainly the different varieties of pig iron, wrought iron, carbon, and oxide or other compound of manganese. The crucible steel of the present day is largely made from such mixtures, the quality of the product depending on the materials used. 3c. Cement Steel. The production of steel by heating wrought iron in charcoal without fusion (cementation) is a very old process, but its origin is unknown. It was described by Reaumur in 1722, and has not been materially changed since. Notwithstanding the introduction of modern processes, this method is still employed for the manufacture of the higher grades of steel for tools and other fine purposes. The iron is in the form of fiat bars about £ in. thick. These are arranged in layers in long boxes or chests of fire brick, each layer being covered with charcoal about ½ in. thick.

When the box is full, it is covered with clay or other impervious material, and heated to bright redness for seven to ten days, according to the degree of carburization required. Trial bars are inserted with their ends protruding, which may be withdrawn from time to time and the progress of the conversion judged from the appearance of the fracture. When the desired end has been attained, the fires are withdrawn and the boxes allowed to cool slowly for several days. The bars after conversion generally have blisters on the surface, apparently formed by the pressure of some gas from within the bar; hence the name " blister steel." The bars, originally soft and tough, are found after conversion to be hard and brittle, and the freshly fractured surface shows a steely appearance. Analyses of successive layers of the bar after conversion show that the carburization proceeds gradually from the surface to the interior, the iron near the surface being much more highly carburized than that at the centre. In order to obtain uniformity in cement steel, it is therefore necessary either to weld several bars together by repeated rolling or hammering, or by melting the bars in crucibles. The former process is adopted only for the softer cement steels, and furnishes shear steel.

The use of this welded steel has been generally superseded by cast steel, but it is still employed for many purposes, particularly for welding to iron. The melting of steel is usually effected in covered crucibles capable of holding 40 to 80 lbs. of metal. They are made of refractory clay or of graphite with sufficient clay to give it coherence. These crucibles are placed in furnaces arranged in a straight line, with their tops or openings on a level with the working floor of the casting house. Each furnace is a rectangular chamber of fire brick, capable of holding two crucibles, and has a separate flue. Siemens's regenerative furnace is also largely used for heating crucibles for steel melting. "When the steel is thoroughly melted the crucible is drawn out of the furnace, and the molten metal cast in the form of rectangular ingots or into special moulds. Where large castings are to be made of crucible steel, the metal from several crucibles is first poured into a common receptacle, and thence into the moulds. Case-hardening of wrought iron consists in a superficial conversion of the iron into steel by heating it with animal charcoal or organic matters in the same manner as that employed for cement steel, but for a shorter time.

Or the iron to be hardened may bo simply heated to redness and covered with a carbonaceous substance like prussiate of potash or cyanide of potassium, which will cause a superficial carburization. Case-hardening is employed for objects which should have a hard and steely surface combined with the toughness of wrought iron. 3d. Mackintosh or Baron Steel. The carburization of wrought iron by means of gaseous hydrocarbons without fusion was proposed in 1824, and was patented in England in 1825 by Charles Mackintosh. It has recently been revived under the name of the Baron process, but has not been made practically successful.

Fig. 1.

Fig. 2.

Fig. 3.

Fig. 4.

Fig. 5.

Fig. 6.

The conversion is effected at a white heat, and is said to be complete in a few hours. - The .limits of this article would not admit of even the enumeration of all the steel-making processes which modern inventors have suggested or endeavored to carry out. A large number of these inventions deal with the direct production of steel from the ore by processes similar to those described under Ikon Manufacture, and a still larger number with the direct conversion of pig iron into steel. - Properties and Treatment of Steel. The physical properties of steel vary according to its composition, structure, and treatment. Thus the specific gravity of blister steel was found by Kirkaldy to vary from 7.7080 to 7.7327; of puddled steel, from 7.6237 to 7.7345; and of cast steel, from 7.8110 to 7.8303. The effect of the amount of carbon, and also of hardening, on the specific gravity, is shown in the following series of Swedish Bessemer steels:

The appearance of the freshly fractured surface of cast steel depends likewise on the amount of carbon and on the degree of hardening. The more carbon present, the closer and more highly crystalline is the grain of the steel, and the lighter the color; effects which are all enhanced by hardening. Steel, unlike soft iron, has the property of retaining.magnetism, its capacity in this regard increasing with the amount of carbon. - Hardening, Tempering, and Annealing. Steel is hardened by suddenly cooling it from a red heat through immersion in water, oil, or other liquid. The degree of hardness thus produced is proportional to the amount of carbon in the steel and the rapidity of its cooling. Hardened steel heated to redness and allowed to cool slowly recovers its original softness and malleability (annealing); but when hardened steel is heated to a temperature considerably below redness, and cooled, it is only softened to a degree inversely proportional, generally, to the temperature of the previous heating. This process is called tempering. For temperatures considerably below red heat, it is practically indifferent whether the cooling be slow or rapid.

The operations of hardening and tempering are dependent on many conditions, such as the composition of the steel, the temperature to which it is heated, the temperature, specific heat, boiling point, mobility of particles, and heat-conducting power of the liquid in which it is cooled, etc. The following table shows the temperatures employed in tempering for different purposes, and the color indicative of each temperature, which appears on the surface of polished steel as it reaches the given degree. ' From these colors, probably due to superficial oxidation, the experienced workman judges of the temper which the steel will assume:

Polished articles may be heated for tempering over or between iron plates, in a gas flame, in molten lead, or in various other ways, until the proper color appears. For articles not polished, the temperature must be otherwise determined, as by heating in oil or tallow or in alloys of known fusibility. When oil or melted tallow begins to smoke, its temperature corresponds with that indicated by straw color on the polished steel; darker and more abundant smoke corresponds with brown; black and still more abundant smoke rises at 580°, the temperature of purple; when the vapor takes fire from a lighted taper, without continuing to burn, the temperature is about 580°; and finally, when the oil burns and rises in the vessel, the point of dark blue has been reached. The following table shows the fusing point of several alloys of tin and lead:

Steels containing other substances besides carbon appear to require different treatment from pure carbon steels. Thus tungsten and titanium steels, so called, if heated bright red and suddenly cooled, are said to become excessively brittle; they must therefore be manipulated at low temperature. Too little is known of these compound steels to permit inferences as to their physical behavior. The hardening of large or irregular masses of steel requires great care. Unequal cooling causes fracture. Generally the more massive portions are first dipped in the liquid, and the thinner portions last; or, in case of any great disparity, special means are adopted to retard the cooling of the smaller parts. The causes of the phenomena attendant upon hardening and tempering steel were long involved in mystery, and are not yet all known with certainty. What is clearly known on the subject may be briefly stated. The degree of hardness assumed on cooling by a given steel is dependent on the rate of cooling. Caron says the degree of hardening is inversely proportional to the square of the time. The liquids which favor rapid cooling are those having a high specific heat and a low boiling point.

Water fulfils these conditions in an eminent degree, while oil has a much lower specific heat and a much higher boiling point; consequently cooling in oil is a more gradual process than in an equal volume of water. Increasing the volume of the liquid and maintaining agitation, so as to diffuse rapidly the heat received from the steel, of course hastens cooling. The most rapid cooling is produced by mercury, by reason of its high conducting power. It is sometimes used to produce extreme hardness. But obviously the initial temperature of the cooling liquid is an essential point; so that heated mercury or fusible alloys could be used to effect slow cooling. Ordinary tempering is a partial annealing; that is, excessive hardness having been imparted to the steel, the excess is removed to the degree desired. It has been found in most cases practically easier to attain an accurate result in this way than by a single process of hardening, arrested at the desired point. But recent experiments by Caron have shown that it is possible, in some cases at least, to effect the hardening in one operation by carefully adjusting the amount and temperature of the water. Water at 181° F. was found to give results with some objects equal to those produced by the most careful hardening and tempering.

Caron has further found that hardening of steel with 0.2 to 0.4 per cent, of carbon in warm, or still better in boiling water, was accompanied by an increase of its tenacity and elasticity without a material impairment of its hardness. - The toughening of large steel objects, such as cannon, is effected by heating them to redness and immersing in oil, where they gradually cool. This process has been recommended for steel rails. The hardening of steel is probably due both to a chemical combination of the carbon (present partly as graphite in soft steel) with the iron, and to a state of tension among the particles, conditions which are both removed by annealing. The tension in a bar of hardened steel is shown by cutting it in two lengthwise, when each piece assumes a curved form, concave on the cut side. Soft iron does not harden when suddenly cooled, but acquires increased rigidity and tensile strength; while cast iron, containing more carbon than steel, becomes under the same treatment extremely hard (chilled iron), often harder than steel. The freshly fractured surface of hardened steel shows a fine grain, often velvety in appearance; that of soft steel presents facets. In the former, analysis shows no uncombined carbon; in the latter, a small amount of graphite is almost always present.

Steel expands on hardening, and loses specific gravity. Eisner found one sample to change in gravity from 7.9288 to 7.6578, and another from 8.0923 to 7.6578. Caron found a decrease from 7.817 to 7.743. The latter found that hammered steel on hardening lost in length and gained in other dimensions, while rolled steel gained in length. The effect of hardening on the tenacity of steel is discussed under Ikon, and also further on in this article. Steel over.heated becomes brittle, and is said to be burnt. Whether this impaired cohesion is due to oxide of iron, or, as has been suggested, to carbonic oxide (either of which might be formed at high temperatures with access of air), or to a crystallization of the particles, is not certainly known. Many fluxes have been suggested for restoring burnt steel. They usually contain easily fusible substances, such as alkalies, borax, etc, combined with carbonaceous compounds, such as prus. siate of potash. Hammering at a high heat is said to restore burnt steel. . The working of steel requires great skill and judgment. It cannot be wrought at very high temperatures; and the more carbon it contains, the lower must be the heat of working. The harder steels are generally hammered at a cherry. red heat.

On the other hand, working at too low a temperature seriously impairs the tenacity of steel, as is abundantly shown by experience with steel rails. Welding steel to steel or steel to iron is difficult, except with the softest or least carburized varieties. Fluxes to facilitate such welding are largely used with good effect; they add nothing to the intimacy of the weld, their action being mechanical only (cleansing, exclusion of air, &c), as in the case of iron welding. There is always danger of the separation of iron and steel at the weld, unless the latter is very soft. Special devices, such as causing one of the metals so welded to overlap and enclose the other, counteract this tendency in part. Or fluid steel may be cast directly around white.hot wrought iron, the weld being promoted by subsequent rolling or hammering. Sometimes the iron and steel to be welded are enclosed in a case of thin wrought iron and exposed to a welding heat, the enclosure preventing an access of air and oxidation of the surfaces of the metal. . Strength of Steel. The cohesive force of steel is usually considered under the different heads of absolute strength, or the force required to produce rupture; the elastic limit, or the least force by which a permanent alteration of form is effected; and the extensibility, or the amount of elongation under a breaking stress.

The experimental data are referred, for convenience of comparison, to bars or rods of one square inch section. The above named properties are dependent, first, on the chemical composition of the metal; secondly, on its homogeneity; thirdly, on its molecular structure; and fourthly, on the temperature. (For comparison of the strength of cast iron, wrought iron, and steel, see Iron.) 1. The effect of the amount of carbon on the properties of steel is shown in the following tables compiled from Knut Styffe's work on the " Elasticity, Extensibility, and Tensile Strength of Iron and Steel:"

The last sample was homogeneous iron prepared with ferro.manganese. To interpret correctly results like the above, it is necessary to eliminate all disturbing influences of composition and treatment. While these figures do not show a uniform change of properties with gradually increasing amounts of carbon, they nevertheless show decidedly that the effect of carbon on iron is to increase its absolute strength and elastic limit, and to decrease its extensibility. An increase of carbon beyond 1.2 per cent, is not accompanied, as a rule, by an increase in absolute strength. When reference is had to the fractured area, it will be seen that the force required to produce rupture does not differ as widely in different steels as when the original area alone is considered. The effect of melting, or in other words of the homogeneity of steel, is strikingly shown by a comparison of the two preceding tables, the former referring to puddled or welded steel, and the latter to Bessemer or homogeneous steel. The effect of molecular structure on the physical properties of steel has been partially treated under Iron. The table, vol. ix., p. 374, shows that the effect of hardening is to increase greatly the strength and elastic limit in steel, and to decrease its extensibility.

The data given by J. Barba ("Memoir on the Uses of Steel") show that as the proportion of carbon decreases, the effect of sudden cooling becomes less marked, but even the softest iron is made somewhat more rigid by this treatment. The effect of hardening and tempering is, further, well shown by the following results of experiments on bars of steel cut from the same mass and submitted to a different treatment, made with reference to the use of steel for the construction of the bridge over the Mississippi at St. Louis:

-The change of molecular structure resulting from working steel when cold has lately demanded attentive consideration from engineers, owing to the increased use of steel for construction and for the permanent ways of railroads. All violent mechanical treatment of steel after it has become cold, such as rolling, hammering, punching, notching, etc, is found to impair its strength seriously. Sand-berg has stated that the.strength of steel rails notched on the flange was decreased from 50 to 97 per cent.; the former where the notch was semicircular, the latter where the notch was square. It is evident that this decrease of strength is not alone due to the removal of so much material, but that there must be a local tension of the particles which leads to rupture, and annealing is found to remove this tension. -The variety of opinions entertained by engineers as to the principal causes of fracture of steel rails is shown in the following summary of answers recently obtained from the administrations of 24 German railways in response to the request of a commission appointed to investigate this subject.

The figures in parentheses indicate the number of administrations mentioning the prefixed cause. 1. The employment of too brittle metal (8). 2. Manufacture at too high temperature (2). 3. Rolling at too low temperature (3). 4. Cooling irregularly or too rapidly after rolling (5). 5. straightening cold, producing fissures which enlarge and result in fracture (15); producing a change of structure (1). G. Notching the flange (14) (only two denied this cause). 7. Manner of piercing the holes (6). 8. Reduction of area of section of rails by the holes (1).

9. Bending the rails for laying on curves (3).

10. Rough handling of rails, such as throwing from cars to the ground, giving rise to fissures which result in fracture (9). It will be noticed that the majority of answers agree in attributing the fracture of rails to improper treatment of the steel when cold. The cause of the brittleness and impaired strength in steel and iron consequent upon punching has been investigated in Lorient, France, by J. Barba, who has found that cold punching induces a local hardening and tension of the metal, in a zone less than 0.04 in. wide, around the hole, and that when this hardened portion is filed or cut away, or softened and relaxed by annealing, the metal regains its original strength and extensibility. He thinks the hardening due to the combination of carbon and iron, as is also supposed to be the case when steel is hardened by heating and sudden cooling. Indeed, this heating and cooling is what undoubtedly occurs to the immediately adjacent metal in punching. The diminished strength of punched plates is caused by this narrow hardened portion, which, owing to its decreased extensibility, receives the full effect of the stress, a rupture being produced in this portion and then extending throughout the whole mass of metal. The same effect, in an enhanced degree, would follow blows or shocks.

The following are among the results obtained by Barba:

The effect of annealing after punching is shown in the following:

The effect of temperature on the strength of steel has already been considered under Ikon. More recently Joule has experimented on the tensile strength of steel bars, and confirms the result of previous investigations, that the tensile strength is not impaired by reduction of temperature. In determining the effect of blows at reduced temperatures, he experimented on cast-iron nails, and found that as many nails broke at ordinary as at freezing temperatures when exposed to a falling weight. These results must not be regarded as contradicting those of Sandberg on iron rails, nor does it follow that the same effect would have been produced had steel bars been used instead of cast iron. - Uses of Steel. The industrial applications of steel, formerly confined mainly to tools, weapons, and springs, have been widely extended since the introduction of the Bessemer and Martin processes. Among the principal modern uses of steel are rails, boilers, machinery, bridge construction, and ship building. The fact must not be overlooked that the term steel is now generally applied to all homogeneous, malleable compounds of iron, and includes products of all degrees of hardness and rigidity.

The adoption of steel for any particular purpose must, therefore, be intelligently based on its composition, structure, and treatment. - Production of Steel in the United States. The following statistics are compiled from the report of the secretary of the American iron and steel association, of January, 1875:

Of the Bessemer production there was made into rails: in 1872, 94,070 tons; 1873, 129,-015; 1874, 144,944. The importations of Bessemer rails for three years were 149,786, 159,-571, and 100,486 tons, valued at $8,207,013, $8,984,103, and $6;838,875,.gold, respectively. The average price in currency at which American steel rails have been sold at the works since the establishment of the industry is as follows: 1867, $160; 1868, $158½; 1869, $132¼; 1870, $106¾; 1871, $102½; 1872, $112; 1873, $120i; 1874, $94¼; 1875, $75. Of the steel other than Bessemer produced in 1874, 34,128 tons was crucible steel, the remainder puddled, open hearth, and blister steel.