T. I. Tikhonov.

Metallography

and

Its Tasks

Tomsk.

Steam Typo-Lithography of P. I. Makushin, Blagoveshchensky Lane, own house.

1908

Metallography and Its Tasks.

T. I. Tikhonov.

Studies of the structure of metals under the microscope created a new branch of science, known under the name of metallography. The first works in this direction were carried out by the English metallurgist Sorby in 1864, and then, in 1878, Prof. Martens published his microscopic studies of iron and steel. In 1880 Barba introduced the microscope at the Creusot works and thus gave the first impetus to the labors of Osmond, who, in fairness, may be called the founder of modern metallography. At the present time metallography is already passing beyond purely scientific bounds and is spreading more and more in factories, where it is indisputably becoming a useful assistant to chemical analysis and mechanical testing.

Metallography chiefly pursues the following two tasks:

1) The determination of the separate constituent parts of alloys (by the appearance of their structure), the determination of their chemical and physical properties, as well as their purpose; the study of the relation between these latter factors and the properties of the metal as a whole.

2) The determination of those changes in structure that are caused by various treatments of alloys (heating, hardening, and mechanical working in general).

For solving the above-mentioned tasks, at first all the methods of physical chemistry were used, insofar as it made it possible to draw any conclusions concerning the internal structure of an alloy; but for metallography at the present time, the indications given by the microscope are more valuable, as are also studies of the circumstances of solidification of alloys. These two branches of metallography may be called microscopy and cryoscopy; the latter name is used in physical chemistry for the study of the solidification of solutions of salts.

Preparation of metallic sections. Let us first turn to the microscopy of iron. A favorite method of observation in petrography—

under the microscope by means of light passing through thin sections—here, because of the opacity of thin metal plates, this cannot be used; therefore, for the examination of iron under the microscope, one must have a polished surface, so that reflected light can be used. The preparation of a polished metal surface is one of the most difficult and troublesome tasks in the examination of metals under the microscope, but at the same time the success and accuracy of the results depend entirely on the preparation of good specimens.

Before proceeding to describe the methods of polishing itself, I consider it useful to make a few remarks concerning the polishing substances used in preparing specimens. The most convenient and commonly used substance for polishing a specimen is ordinary commercial emery in powder form, but only after washing and careful sorting by elutriation. Engineer A. Rzeshotarsky*) in his work used a very simple method of elutriation, consisting of the following: into a bottle with a capacity of 1–2 liters, filled with water, one places one pound of fine-grained emery; then the liquid is shaken and left at rest for a certain period, so that the coarse particles of emery settle to the bottom. After this the liquid is carefully poured into another vessel and, after allowing it to settle for some time, is again poured into a new vessel, and so on. By repeating this operation 5–6 times, several grades of emery of different fineness are obtained, which are then either carefully dried in a water bath at a temperature not above 70°, otherwise the particles of emery again gather into grains, or the emery is simply left in the water solution, with labels only being put on the jars, 1, 2... according to the fineness of the powder.

Martens**) carries out the separation of emery powder somewhat differently, namely: 1/2 pound of emery powder is mixed with water in a porcelain mortar in order to obtain a liquid paste. The paste is then washed into another vessel, 1/2 liter of water is added, and after vigorous stirring it is poured into a special separator (Fig. A, Pl. V). The separator consists of a bottomless bottle A, closed below with a stopper through which passes a brass tube B, provided at its upper end with a small dish. When the liquid in vessel A has stood for some time and the coarse particles of emery have settled to the bottom, then

*) A. Rzeshotarsky, Microscopic Investigations of Iron, p. 9, 1898.

**) Sted. Method for preparing polished surfaces. Artillery Journal, 1895, No. 11.

the tube with the dish is lowered downward and the mixture is poured into the lower vessel C; the settled emery is dried and used for polishing.

For producing one’s own assortment of emery paper, the dried grades of emery obtained by the above-mentioned method should be mixed with albumin glue, which is used for making sensitive plates, and this mixture should be applied with a brush to sheets of strong paper.

Prof. Heyn, in his work, uses commercial emery paper from the Hubert factory in Paris of the following numbers: 000, 00, 0, 1ᵐ, 1, 1°, 2, and 3.

Besides emery, crocus (“crocus for steel” and “crocus for gold”), “Vienna lime,” and the highest grades (50–60) of powdered carborundum are sometimes used for polishing. Finally, it should be noted that all polishing substances must be kept in closed vessels, so that particles of room dust, and still more of factory dust, under no circumstances can get into them. The area of the metal to be polished should not be large (1/2 □ cm/cm and 1/1 □ cm/cm), since it is very difficult to polish a large surface; first the surface is made smooth and even by ordinary means, and then it is treated with emery paper of various grain sizes, glued onto rotating wooden disks, until there are no scratches at all on the surface of the section. The subsequent treatment of the section is carried out on wooden washers covered with cloth, by means of the finest powder and water; with some skill, it is possible to obtain a perfectly mirror-like surface with no scratches whatever. A surface treated in this way very rarely makes it possible to see the structure of the metal under the microscope, and therefore it must undergo still further treatment, which may be done in 3 substantially different ways, each separately or successively one after another.

a) “Relief” polishing (introduced by Sorby and considerably improved by Osmond, Martens, and Wedding). The section is polished further, under slight pressure, with soft rubber or parchment sprinkled with a little crocus or very fine emery. The harder constituent parts of the section resist the action of the elastic polishing surface more than the softer ones, and protrude in the form of reliefs *) (Fig. 8. Vol. I. 9, 10, and 12. Plate II. 18 and 19, Vol. III).

*) The raised parts of the section can easily be distinguished if one remembers that, on the side opposite to the direction of the ray of light, the raised parts of the section always have a black stripe—a shadow.

b) Polishing with etching (introduced by Osmond). The section prepared according to point a is polished further with rubber or parchment, with the addition to the water of licorice extract or sugar; in this process certain constituent parts of the structure of the section acquire coloration, which provides a greater possibility of still further distinctions. But different coloration of the constituent parts of the section can be obtained not only by oxidation of the section, but also by heating the latter, if only the constituent elements of the alloy are subject to oxidation to different degrees; thus, for example, when a section of alloys of copper and silver is heated, the constituent parts of the section with a high silver content remain shiny, whereas others, with a high copper content, acquire various colors.

Note. Licorice root is sold under the name coco (coco), and an extract from it is prepared in the following way: 10 grams of finely crushed root are soaked for 4 hours in 100 grams of water and then the infusion is filtered. On the third day the infusion already begins to ferment and is fit for use.

Prof. Heyn prepares sugar water of the following composition: 10 grams of sugar per 100 grams of cold water.

c) Etching. The section polished on cloth is thoroughly cleaned of grease and placed in a bath with the etchant. This is either a solution of iodine in potassium iodide and water, or very weak nitric acid, or hydrochloric acid strongly diluted with alcohol, or bromine water, or ordinary water with a few drops of salicylic acid in alcohol, etc., reagents. Owing to etching, a certain change in the coloring of the section is obtained, and the relief of the latter becomes more distinct (Figs. 1 and 3. Vol. I. 15. Vol. II. 26, 27 and 31. Vol. IV).

Note. Wedding *) proposes, for washing sections, a whole series of different liquids. First the section is washed in water, then in chloroform, carefully cleaning with a brush all pits and pores of the section of accidentally remaining emery or sandpaper grit. Then the section is successively immersed for a short period of time in the following liquids:

1) A mixture of chloroform with alcohol (1 : 1).
2) Chloroform.
3) Sulfuric ether.
4) Alcohol.

) Wedding. Eisenhüttenkunde* B. I. S. 867.

5) A mixture of alcohol with water (1:3).

5). Distilled water.

If the specimen, taken out of the distilled water, is evenly covered with water, it is then lowered into the etching liquid. If, however, water gathers in drops on the surface of the section, this proves that the surface of the section is not yet clean, and the section must be washed again. If it is desired to examine the specimen under the microscope without etching in acid, then, after removal from the distilled water, it must again be immersed in liquid No. 5, and then in No. 4 and for several seconds in No. 3. The section must be immersed in the acid with tweezers (and not with the hands, which always leave greasy spots), and no deeper than 1–2 cm below the level of the liquid (with such immersion of the section in the acid, the appearance of the pattern on the section can easily be observed through a magnifying glass). A. Rzheshtarsky*) advises, for etching hard grades of iron, taking 2% nitric acid and a 4% solution for soft grades of iron, immersing the specimens for 10–15 seconds. For very gentle etching he advises taking only 5–10 drops of nitric acid to a glass of water.

Etched specimens must be kept in a dry place protected from dust. It is best to wrap each specimen in tinfoil and store it in a closed room in the presence of caustic potash, or simply to store it under a glass bell in the presence of sulfuric acid.

The section thus prepared for further observations and photography is placed in the microscope. Illumination of the object for microscopic observations and photography is achieved in the following manner: the object is placed obliquely with respect to the optical axis of the microscope and is illuminated by daylight (Fig. B in Table V); but for strong illumination of the section it is necessary to give it a large inclination, as a result of which a clear image in focus is obtained only in the form of a narrow strip. Magnification above 200 times is rather difficult to obtain, since, when the objective lens is brought closer to the object, the objective mount prevents the light rays from illuminating the section.

In drawing C of Table V the section is arranged perpendicular to the optical axis. A ray falling from an artificial light source falls on the reflecting glass pl, inclined to the optical axis at an angle of 45°, is reflected from it onto the polished surface and returns from the latter in the direction of the optical axis,

) A. Rzheshtarsky. Microscopic Investigations of Iron*. p. 27. 1898.

passing through the reflecting glass pl. Part of the light is expended quite uselessly—owing to its passage through the glass pl.

In Fig. D, Plate V, the section is again placed perpendicular to the optical axis. The prism pz completely reflects the ray 1 falling upon the objective f. The latter serves at the same time both to concentrate the incident bundle of rays and to obtain a microscopic image. In addition, Frémont’s device *) deserves special attention: in it the rays of light reflected by mirror D (Fig. E, Plate V) pass through an opening into the tube of the microscope, where they meet the concave movable mirror C. By changing the position of mirrors D and C, one can make the light ray pass through the objective. Prism K directs the rays of light parallel to the axis of the microscope. Through mirror C and prism K passes the conical tube P, which, with the aid of an eyepiece, serves for observing the image of the section illuminated by the rays of light.

The source of light may be an ordinary or an electric lamp, Drummond light, and the like. The latest microscope attachments make it possible to examine sections at any magnification and to take photographic pictures, but only on the condition that the plane of the section be set strictly perpendicular to the axis of the microscope; to facilitate setting the section in such a position, the microscope stage is provided with a foot with a ball-and-socket joint (the system of Prof. Martens, manufactured by the firm of Zeiss in Jena).

Microscopic images can be reproduced in two ways: either by drawing them on paper, or by means of photography. The first method is now used only at high magnifications, when the images are not clearly visible, or when the section has been poorly prepared—has a spherical surface **). But in order to draw a section successfully, one must have great skill; photographing a section involves far less trouble and requires less time. For this purpose a photographic camera is attached directly to the eyepiece of the microscope; first an image is obtained on ground glass, and then it is replaced by a cassette with a sensitive plate. It should be noted that a horizontally attached photographic apparatus (the Martens system of the Zeiss firm) is incomparably more convenient than a vertical one for photographing sections.

) Rzehotarsky, Microscopic Investigations of Iron*. P. 30, 1898.

) More detailed information on drawing instruments may be found in Zimmermann’s brochure The Microscope. Trans. by Dr. A. Ilyich, 1896, St. Petersburg.

Microstructure of iron. The individual constituent parts of iron, discovered by means of the methods described, have received special names, namely:

a) Ferrite (after Howe). Iron free, or almost free, from carbon; the softest constituent part of iron. On relief sections it appears lower than the other constituent parts. By polishing with an etchant, or by careful etching, it is not colored at all, or is colored a pale yellowish hue. Very poor carbon iron consists of ferrite and under the microscope has the appearance of small grains, so that the structure of the section resembles marble. When the action of the etchant is somewhat stronger, or when a special agent is used in the form of a solution in water \((1/12—1/30)\) of cupric-ammonium chloride, grooves appear on the surface of the section, called etching figures (ätzfiguren); these figures very clearly bring out the ferrite grains and make it possible to determine precisely their number and average size (Fig. 1, Pl. I). The black network separating from one another, in this section, the ferrite grains consists of a special carbonaceous substance—pearlite.

In soft grades of steel the ferrite grains are arranged in the form of a mosaic making up star-shaped figures (Fig. 2, Pl. I), but as the carbon content in the steel increases these thin white plates of ferrite, forming a more regular network, become thinner and thinner (Figs. 3, 4, Pl. I). With free and slow cooling of iron (in the middle of an ingot), the ferrite grains are separated out in a powerful network along the periphery of the carbonaceous iron (Fig. 5, Pl. I). Forging and rolling strongly affect the change in the form and grouping of the ferrite grains: the former crushes them and scatters them in disorder among the carbonaceous iron (Fig. 6, Pl. I), while the latter draws them out in the form of fibers (Fig. 7, Pl. I); but one must not forget that the grouping of the ferrite grains in the latter two cases depends strongly on the final temperature of treatment.

b) Cementite (after Howe). According to Osmond, cementite is considered to correspond to iron carbide in the form \(Fe_3C\). Cementite is harder than the other constituent parts, as a result of which, on “relief” sections, it protrudes on the surface earlier than the other constituent parts of the structure of iron. After etching and polishing with an etchant it remains uncolored and shines like a mirror (Figs. 8, Pl. I; 9, 10, Pl. II). The structure of cementite can best be studied in cement steel, where it often forms whole accumulations in the form of straight bands (Fig. 11, Pl. II) and the finest fibers scattered among the remaining carbonaceous iron (martensite).

c) Pearlite (according to Howe), on “relief” polished sections, stands higher than ferrite and lower than cementite. When polished with etching, pearlite becomes yellowish and even yellow-brown and black. Pearlite always consists of two elements, most often having the appearance of plates (Figs. 12 and 13, Pl. II), and more rarely of tiny grains (Fig. 14, Pl. II), these elements alternating with one another. Both forms of pearlite differ from each other in hardness, and sometimes also in coloration. Pearlite apparently consists of ferrite and cementite. Sometimes a transitional form occurs, called sorbite by Osmond, which is then found in pearlite together with ferrite and cementite. Pearlite is found only in those grades of iron in which carbon is present predominantly in the form of carbide, which gives grounds for assuming that pearlite, in its hard parts, contains carbon in a form identical with carbide. In very soft grades of iron, pearlite in the form of a network borders the ferrite grains (Fig. 1, Pl. I), and conversely, in semi-hard and hard grades of steel the pearlite grains are bordered by a network of ferrite (Figs. 4, 6, Pl. I). In gray cast irons, pearlite is grouped in the form of fir-tree-like, discontinuous crystals, similar to those found in shrinkage cavities of cast-iron and steel castings (Fig. 15, Pl. II).

d) Martensite (according to Osmond) is found only in that iron which has been quenched at a temperature above the critical point A, according to Chernov, and therefore contains all the carbon in the form of quench carbon. The carbon content in martensite varies in such a way that it may be regarded as a solution of carbon and the carbide Fe₃C in iron. Owing to the inconstancy in the amount of carbon, martensite may be of varying hardness, but it is always harder than ferrite and considerably softer than cementite. When polished with etching it either is not colored at all, or is colored a faint yellow; and when etched, for the most part, brown. Coloration is a deceptive means of identification; characteristic only of martensite is its acicular structure: the needles intersect in 2–3 directions (Fig. 16, Pl. II; 17, Pl. III), whereby martensite differs essentially from pearlite, the grains of which are always arranged in parallel (Fig. 12, Pl. II; 19, Pl. III).

e) Graphite and annealing carbon metallographically differ very little from one another (Figs. 18, 19, Pl. III; 15, Pl. II).

Thus, comparing together all the characteristic features distinguishing the microscopic elements of the structure of iron, we obtain the following table:

From all the foregoing it is not difficult to observe that the elements of the structure of iron depend on the chemical composition of the latter and on the conditions of mechanical treatment in general, but, moreover, are also in strict interdependence with one another; therefore some authors, on the basis of their observations, have tried to express this dependence analytically. Thus, for determining the percentage content of slag, Sauveur*) gives, on the basis of his observations, the following formulas:

a) For iron with a carbon content less than 0.8%.

1) Annealed.

\[ \text{carbide }(\mathrm{Fe}_3\mathrm{C})=\frac{100\cdot C}{6.67}\,\infty = 15\times C; \]

\[ \text{pearlite}=\frac{100}{12}\times \text{carbide}; \]

\[ \text{ferrite}=100-\text{pearlite}; \]

2) Hardened:—martensite \(=100\%\).

Note. These equations are of value only for those grades of iron in which the carbon content is not less than 0.14%.

b) For iron with a carbon content greater than 0.8%.

1) Annealed:

\[ \text{carbide }(\mathrm{Fe}_3\mathrm{C})=15\times C; \]

\[ \text{pure iron (ferrite)}=100-(15\times C); \]

\[ \text{pearlite}=(100:88)\times \text{ferrite}; \]

\[ \text{cementite}=100-\text{pearlite}; \]

2) Hardened:

\[ \text{cementite}=\frac{100\cdot C-90}{5.77}; \]

\[ \text{martensite}=100-\text{cementite}. \]

Up to the present time the various compounds into which iron enters with carbon under the influence of heating and of foreign impurities have not yet been fully elucidated. As a result, many authors call the same elements of the structure of iron by different names, so that the beginner who studies metallography finds on this question a terrible chaos in the literature.—In order to make this task somewhat easier, A. Rzeshotarskii**) gives a table in which are grouped all the names given by different authors to one and the same element.

) “Trans. Am. Inst. Mining Eng.” Colorado Meeting Sept. 1896.
“Journal Iron and Steel Inst.” 1896. Vol. II. p. 191.
) A. Rzeshotarskii. Microscopic investigations of iron.* p. 64. 1899.

In concluding the metallographic structure of iron, one cannot fail to say a few words about the dependence between the microscopic structure of iron and its physical properties; the latter are wholly dependent on the character and type of the former. Thus quenching carbon increases hardness, raises the limit of elastic changes and strength, but reduces percentage elongation and contraction. Carbide carbon acts in the opposite way. In addition, it has already been fully established that the strength of iron increases with the content of martensite and, conversely, decreases with an increase in the content of pearlite. Voids, slags, cementite, and graphite reduce the density of iron, and consequently also its strength, while increasing elongation and contraction. The size of the crystals acts in the same way. In general, it may be said that the strength of iron increases:

1) from the presence of elements which themselves possess great strength (nickel, cobalt);

2) from the presence of elements which delay the separation of carbon into graphite or carbide (manganese up to a certain percentage content, chromium, tungsten, and titanium);

3) from the presence of elements which reduce the size of the crystals (an insignificant amount of silicon);

4) from the presence of elements which densify the mass of the metal (silicon and aluminum, if they are present in an extremely small quantity); and

5) from the presence of elements which promote the separation of hard, needle-shaped crystals.

The strength of iron is diminished:

1) By the presence of elements that cause the separation of graphite or carbide (silicon, phosphorus, and sulfur), or that only promote such separation (copper);

2) By the presence of elements that increase the size of the crystallites (phosphorus and a considerable content of manganese);

3) Through non-metallic impurities that separate the crystals of the metal (gases, a large quantity of silicon, compounds of copper and sulfur, manganese and sulfur, oxides).

A large size of the crystallites causes cold-shortness (phosphorus), while foreign impurities which themselves melt at a high temperature (sulfides), or which interfere with the welding together of the individual crystals of the metal (oxides), cause red-shortness.

Albert Sauveur *) devoted much attention to the question of the influence of the size of metal crystallites on its physical properties and arrived at conclusions of the following kind:

1) Slow, quiet cooling of iron from a certain temperature \(X\) causes a change in the structure of the metal.

2) Quiet, slow cooling of iron from a temperature below \(X\) causes no changes whatever in the structure of the metal.

3) The position of the temperature \(X\) depends on the chemical composition of the iron; thus the impurities—carbon and phosphorus—lower this temperature, although in different ratios.

4) Carbon and phosphorus impart a coarse crystallization to iron.

5) The purer the iron, the higher the critical temperature \(X\).

6) The higher the heating temperature from which the iron is cooled, the coarser the crystallization of the latter.

7) The more slowly the cooling of the iron proceeds, the more considerable is the growth of the crystallites.

8) During mechanical working the crystals may be comminuted, but it is sufficient to heat the iron to the temperature \(X\) or above it and to cool it slowly for the iron again to assume its former structure.

How the size of the crystallites affects the mechanical properties of iron may be judged from the following experiments of Sauveur **). Sauveur carried out experiments with rails and for this purpose took test bars from different parts of the rail—the head, the web, and the base; one

) Baumaterialienkunde, p. 252. 1900
) Stahl und Eisen*, p. 280. 1899.

one batch of rails was rolled at a very high temperature, and the other at a low one. The difference in mechanical properties is evident from the following table.

For greater clarity, the results of Sauveur’s observations on steel in which the crystal size does not exceed 0.0225 qmm are represented in the form of curves (Fig. 1, Table V), from which it is clearly seen how the principal mechanical properties of steel decrease with crystal size. These observations of Sauveur were fully confirmed by the experiments of N. N. Ljamin*).

From Sauveur’s observations H. v. Jüptner**) derived several formulas for determining the mechanical properties of iron as a function of crystal size, namely:

$\alpha)$ Rupturing force.

\[ \delta_B = 75.5 - 0.004 \cdot K. \]

$\beta)$ Elongation.

\[ \delta = 26.5 - 0.0735\,K. \]

$\gamma)$ Compression.

a) for $K < 130$—

\[ q = 47 - 0.2\,K. \]

b) for $K > 130$—

\[ q = 29 - 0.07\,K. \]

Note: $K$ is the mean crystal size in 0.0001 qmm; but these formulas may be used only in the case when the crystal size varies from 35 to 230, for beyond that the dependence of the mechanical properties of the metal on the crystal size takes the form of a straight line.

*) Chem. Ztg. 21 Rep. p. 205.
**) Baumaterialienkunde. S. 252. 1900.

For steel with 1.10% C, 0.02% Si, and 0.41% Mn, Howe*) gives the following relation between the magnitude of the crystals in qmm \((A)\) and the critical temperature \(T_{\max}\) (between \(722^\circ\)—\(1050^\circ\) C.):

\[ T_{\max}=680+281350\cdot A. \]

The relation between the size of the metal grains and its mechanical properties is not difficult to imagine: let us assume, roughly, that the whole metal consists of spherical crystals, the spaces between which are filled with some other substance. Then the mechanical properties of the metal will depend on the cohesive forces of the spherical crystals themselves; and the cohesive force of the crystals will be the greater, the more frequently the crystals are in contact with one another and the smaller the spaces between them, i.e. the smaller the crystals themselves are in size.

Theory of alloys. All alloys in general may be compared with crystalline rocks. In actual fact, homogeneous alloys, which in the solid state would be the same in different places both physically and chemically, are exceptions. The division of rocks into simple and complex ones is also admissible for alloys, and especially for various grades of iron. By simple crystalline rocks are meant those which consist of one and the same rock-forming mineral, as, for example, marble. Alongside them may be placed very low-carbon iron, which consists of small, separate crystallites of (almost) pure iron, clearly visible under the microscope (Fig. 1, Table I). Complex crystalline rocks are considered to be those which consist of a mixture of different minerals, for example granite. Comparable with these rocks are grades of iron rich in carbon, where besides iron there are also other constituent parts (Fig. 19, vol. III, 15, 11, vol. II).

Many metallurgists, and especially Ledebur, long ago pointed out that alloys, including alloys of iron with carbon, should be classed as solidified solutions, and the investigations of the most recent chemists have indeed confirmed Prof. Ledebur’s view.

If a 10% solution of common salt in water is gradually cooled, then, after a gradual fall in temperature, at about \(-8^\circ\) C we shall notice a sudden retardation, accompanied by the separation of ice crystals. Approximately at \(-22^\circ\) C there will be a second halt of the thermometer, during which the entire remaining liquid mass of the solution will solidify, and the temperature throughout the entire period of solidification remains unchanged (\(-22^\circ\) C). Only after the complete solidification of the whole mass does further

*) Journ. “Iron Steel. Inst.” 1891. 1 pag. 199.

decrease in temperature. If, for different solutions, the percentage content of common salt is plotted along the abscissa axis, and along the ordinate axis the corresponding temperatures at which slowings or arrests in the fall of the thermometer’s mercury column were observed, one obtains a “freezing-point curve” (Fig. 1a, Table V), first constructed by Guthrie in 1876.

Such a curve consists of 3 parts: 2 branches AB and BC, intersecting at point B, and the horizontal DE, passing through point B. Branch AB represents the gradual solidification of the solvent, i.e. water, and BC—the solidification of the dissolved body, i.e. common salt. The line DE corresponds to the lowest point in the freezing curve, which, for any salt content in water, occurs at −22° C. At this temperature, from all solutions of common salt in water there is always separated a dense mixture of crystals of salt and ice in the proportion: 23.5% common salt to 76.5% ice. The freezing point of a solution with such a relative quantity of constituent parts is only one (B) and is called the eutectic point, while the solution itself, corresponding to point (B), is the eutectic mixture. All solutions with a salt content less than that in the eutectic mixture separate, during the period represented by curve AB, crystals of ice. Owing to this the solution of common salt becomes enriched in salt, and when the salt content in the remaining liquid solution reaches the limit corresponding to the eutectic mixture, then this peculiar mixture of particles of ice and the eutectic mixture also solidifies. Analogously to this solution, solutions with a content of common salt greater than 23.5%, when changing along curve BC, separate crystals of common salt until again the remaining liquid part of the solution becomes the eutectic mixture and solidifies at −22° C.

A completely analogous picture, judging from the data of Heycock and Neville, is presented by the freezing curve of alloys of silver and copper (Fig. 2, Table V), where along the abscissa axis is plotted the percentage content of copper in the alloy, and along the ordinate axis the corresponding freezing temperatures. Here branch AB corresponds to the period of separation of silver crystals, BC—to copper crystals, and DE—to the freezing point of the eutectic mixture, whose composition is 28% copper to 72% silver at a freezing point of 778° C. In comparison with all the others, the eutectic alloy has the lowest freezing point, and, consequently, melting point, and is uniform in structure, for which reason it was called eutectic (well built). The eutectic mixture consists of small, well-mixed crystals of the mixed elements of the alloy. Alloys in which copper is less than 28%, in the solid co-

in the solid state consist of crystals of silver immersed in a mass of eutectic mixture, the amount of which keeps increasing until the percentage content of copper in the alloy reaches 28%; then the entire alloy will be eutectic. Alloys in which there is more than 28% copper appear in the solid state as crystals of copper immersed in the eutectic mixture, the amount of which decreases as the copper content increases.

A completely analogous picture to that for the alloy of copper and silver is observed during the cooling of many other alloys, such as: lead and zinc, tin and lead, lead and antimony, copper and cuprous oxide, and many others. But not all alloys have cooling curves similar to the preceding ones; thus, for an alloy of copper and antimony, the cooling curve, as the observations of Le Chatelier*) have shown, has a somewhat different form (Fig. 3, Table V). Such a curve may be divided into two: ABE and ECD, each of which in character fully resembles the preceding one. At point E the solidification curve makes a break, and at this time a chemical compound of copper and antimony, corresponding to the formula Sb Cu₂, separates from the alloy (the percentage content of such a chemical compound Sb Cu₂ is determined by the abscissa of point E, G).

Let us denote both constituent parts of such an alloy by K and L, and the third body, the chemical compound of K and L, by M; then the microstructure of the alloy whose solidification curve is ABE is a mixture of the elements K and M, while the microstructure of the alloy with ECD is a mixture of the elements M and L. At points B and C eutectic mixtures separate from the alloy—in B, a eutectic mixture of K+M, and in C, of M+L. Thus the microscopic structure of such a binary alloy of the elements K and L will have the following character:

1) Crystals of element K, surrounded by a eutectic mixture of K+M (corresponds to curve AB).

2) Crystals of element M, surrounded by a eutectic mixture of K+M (corresponds to curve BE).

3) Crystals of element M, surrounded by a eutectic mixture of M+L (corresponds to curve EC).

4) Crystals of element L, surrounded by a eutectic mixture of M+L (corresponds to curve CD).

Microscopic investigations by Charpy**) and Stead***) of alloys of copper and antimony fully confirm the validity of such reasoning.

*) Bul. d. l. Soc. d’Encour., 1895, p. 573.
**) Bul. d. l. Soc. d’Encour., March 1898.
***) Joc. chem Ind., December 1897.

Thus alloys similar to the alloys of copper and antimony have only a somewhat different form of the solidification curve, and in the character of the separation of the constituent parts of the alloy during the period of solidification they do not differ in any way from the preceding ones.

For observing the characteristic points of solidification of all easily fusible alloys, a very simple apparatus is used, schematically represented in Fig. 4, Plate V. The molten alloy is poured into cylinder B, which ends in a cone with a tube; the conical opening of cylinder B can be tightly closed by the spherical end of rod A, and at the end of the tube there is a platinum gauze E; around cylinder B there is the corresponding bath CC. During observations the entire apparatus is placed under the appropriate conditions and, from time to time, rod A is raised; then the liquid parts of the alloy pass through gauze E to the bottom of apparatus D, while the solid parts (crystals) remain on the gauze. The apparatus is extremely simple in construction, inexpensive, and gives excellent results.

It is well known that sulfur, at a temperature below 95.6°, a short time after melting, changes its structure and color and changes in specific gravity from 1.96 to 2.07. Such a change of a body from one state to another, in which only the physical properties of the body change, is called allotropy. As in chemical, so also in physical changes of bodies, the liberation or absorption of heat always takes place; consequently the critical points of a body, at which its change occurs, can be observed. It turns out that an alloy of iron with carbon (carbonaceous iron), upon slow cooling, undergoes changes entirely analogous to the allotropy of sulfur.

The simplest way to detect the critical points of an alloy is as follows: the molten alloy is placed in a well-insulated vessel, so that heat is not radiated, a pyrometer is lowered into it, and the time of cooling of the alloy is observed by the clock. During the passage of the alloy through the critical points, either absorption or liberation of heat will occur, and because of this an accelerated or retarded cooling of the alloy itself will be observed. Sometimes the critical points of an alloy proceed with such an abundance of heat evolution that heating of the alloy becomes noticeable even to the naked eye. In order to make observations of the critical points of alloys more precise, Roberts-Austen, chiefly for alloys of carbonaceous iron, used an apparatus of the following construction: (Fig. 5, Plate V) in front of the mirror of galvanometer G there is, in a wooden hood, an electric lamp A₁, the light from which passes through the ho-

horizontal slit with a vertical thread (the slit is made in the wall of the cap of lamp \(A_1\)), and strikes the mirror of the galvanometer. The light reflected from the galvanometer passes through a series of prisms placed at the galvanometer itself and strikes the photographic plate \(P\), leaving on it a trace in the form of a curved line, whose abscissas are proportional to time, and whose ordinates are proportional to the current strength of the galvanometer; the current strength of the galvanometer, as we shall see below, is directly dependent on the change in temperature of the alloy. The photographic plate \(P\) is fastened to the float \(F\), which can be lowered or raised depending on whether water flows out of or into the vessel \(H\). The piece of alloy \(B\) being observed is placed inside the porcelain tube \(A\); a hole is made in the piece of alloy and a thermoelement of 90% platinum and 10% iridium is placed there, and the conductor from the element is connected to the galvanometer \(G\).

Recently Roberts-Austen made a small modification in his apparatus, namely: he takes two thermoelements of Pt and Ir and connects them, as shown in Fig. 6 of Plate V, with two galvanometers \(G_1\) and \(G_2\), of which \(G_2\) is distinguished by great sensitivity. In this arrangement, one thermoelement \(A_1\) lies in a piece of alloy \(B\), the cooling curve of which has already been precisely determined, and the other \(A\)—in the specimen \(C\) being tested. With such an arrangement of the instruments, as indicated in Fig. 6, galvanometer \(G_2\) will show the temperature difference of the thermoelements \(A\) and \(A_1\). In Fig. 7 of Plate V are shown the cooling curves of carbon steel (with 0.1% C), the ordinates of which give the difference in the readings of the thermoelements, and the abscissas the corresponding time; as is evident from the character of the curves, the iron observed had three critical points: \(A_3\) at \(862^\circ\) C, \(A_2\) at \(767^\circ\) C, and \(A_1\) at \(690^\circ\) C.

With the aid of such a sensitive instrument Roberts-Austen investigated a whole series of iron alloys with carbon not far from the points of solidification. He presents his observations in the form of curves (Fig. 8 of Plate V), whose abscissas are proportional to the carbon content in the iron, and whose ordinates are the temperatures at which arrest points of the thermoelement were observed. The characteristic points of solidification of carbon steel are represented by the upper curve \(a b c d e\). Here the line \(ab\) indicates the temperatures at which carbonaceous iron alloys of various composition begin to solidify. Pure iron solidifies at \(1600^\circ\) C, and then the melting point of iron falls with increasing carbon content, and for an alloy with 4.3% C the melting point lies at \(b\) on the straight line \(de\), corresponding to a temperature of \(1130^\circ\) C. All carbonaceous iron alloys with a C content greater

1.2% on cooling always have two characteristic points: one on the line ab or bc and the second on de; only the alloy with 4.3% C has a single characteristic point of solidification—b.

Casting a general glance at the solidification curve of carbonaceous iron obtained by Roberts-Austen, and at the solidification curves of common salt and of alloys in general discussed earlier, one cannot fail to notice a striking similarity, but one inevitably asks what here is to be the eutectic mixture, the solvent, and the dissolved substance. The appearance for the first time of a second characteristic point in the alloy with 1.2% shows that, up to this moment, only a dense compound of iron with carbon separates from the alloy, and it has only one point of solidification. If, however, the carbon content in the iron increases to 1.2%, then, on cooling, particles of carbonaceous iron (approximately with 1.2% C) will first solidify in the alloy and give a point on the curve ab, while the remainder of the liquid alloy, owing to the separation of the first carbonaceous iron particles, will gradually become enriched in carbon until the carbon content in the alloy rises to 4.3%; then this remainder of the alloy will solidify and, in its solidification, will give the second point on the straight line de. Thus, in the solidification of carbonaceous alloys with a C content greater than 1.2% but less than 4.3%, solid carbonaceous particles of iron first separate from the alloy (let us call them Fe C) and, upon their solidification, give the first characteristic point on the curve ab; but as the Fe C particles separate, the remainder of the liquid alloy is enriched with carbon, and as soon as the carbon content in this remainder of the liquid alloy reaches 4.3%, it too solidifies and gives the second characteristic point of solidification on the straight line de. Such a picture of the phenomenon is observed until the carbon content in the alloy rises to 4.3%; the alloy with 4.3% C has only one characteristic point (b)—the point of solidification at 1130°C. If the carbon content in the alloy of iron is greater than 4.3%, then, during solidification of the alloy, graphite will first separate and give the first characteristic point on the curve bc; but, as a result of the separation of graphite, the remainder of the liquid alloy will lose part of its carbon until the content of the latter is lowered to 4.3%; then the alloy will solidify completely and give the second characteristic point on the straight line de. Thus, the role of the eutectic mixture, apparently, in both cases is played by an alloy of iron with 4.3% carbon. Up to a carbon content of 4.3% in the iron alloy, dense carbonaceous Fe C particles first separate from the alloy during solidification (cf. ice crystals), and after—4.3% C—graphite particles (cf. common salt), and consequently the solvent in both cases...

parts is the carbon steel itself. As a result, upon solidification of the alloy, we have: to the right of point \(b\)—a solid mass of eutectic mixture (with \(4.3\%\) C) with particles of graphite, and to the left of point \(b\)—the same eutectic mixture, but only with particles of \((\mathrm{Fe}\ \mathrm{C})\) carbon steel of another composition.

The character of the curve \(abc\) is influenced by the rate of cooling of the alloy; thus, some grades of white cast iron can be cooled rapidly without separation of graphite, and the particles of the solidified alloy will consist only of particles of cementite (carbide) and pearlite, or, with a slight separation of graphite, of graphite, cementite, and pearlite; but at the same time the final point of solidification of the eutectic alloy for white cast irons will lie much lower than for ordinary carbon steel. This latter circumstance gave Le Chatelier occasion to infer the existence of yet a second line \(XZ\)—a line of solidification points of the eutectic mixture for white cast irons; but the existence of the second line \(XZ\) is only hypothetical and has not been confirmed by the extensive experiments of Roberts-Austen. Thus, at temperatures above the curve \(abc\), all carbon alloys of iron are in the liquid state and are homogeneous in composition, while below \(abc\) they represent either eutectically solidified carbon steel (point \(b\)), or a solidified mixture of particles of dense carbon steel \((\mathrm{Fe}\ \mathrm{C})\) or graphite in the midst of eutectically carbon steel.

If we trace the cooling of carbon steel below a temperature of \(1000^\circ\mathrm{C}\) down to \(650^\circ\), we shall observe several critical points accompanied by the evolution of heat. Such evolution of heat, of course, depends not on the transition of the alloy from the liquid state to the solid, but on a change in the molecular structure of the alloy (on the allotropy of the alloy). The curve of such critical points, first depicted by Osmond and finally confirmed by Roberts-Austen, is shown in Fig. 8—\(ABCDE\), and in character fully resembles the freezing curve of alloys in general. Assuming, according to Osmond, that above the points of the curve \(ABC\) the alloy is present as a solution of carbide (a certain variety of carbon steel,—\(\mathrm{Fe}_3\mathrm{C}\)), we shall see that the curve \(AB\) corresponds to the separation of crystals of the pure solvent, i.e. ferrite, while the curve \(BC\) corresponds to the separation of the pure dissolved body, i.e. carbide (cementite). Point \(B\) corresponds to the simultaneous separation of both bodies in the form of a eutectic mixture—pearlite. All alloys with a carbon content less than that which corresponds to the eutectic mixture \((0.8\%)\), in the solid state consist of grains of ferrite surrounded by the eutectic mixture—pearlite. The quantity of pearlite in the mass

increases, and that of ferrite diminishes until the carbon content in the alloy becomes equal to 0.8%; then the entire mass of the alloy will consist only of pearlite. An alloy with a C content greater than 0.8% consists of crystals of cementite (carbide) that have separated out first, surrounded by pearlite. All this has been fully confirmed by metallography, so that the theory of Osmond and Roberts-Austen is well founded. The influence of the straight line FG on the structure of carbonaceous iron has not yet been sufficiently clearly established, and therefore I could, for the sake of brevity, refrain from considering it at all, so as not to have to enter into the purely theoretical and still disputed domain of Osmond’s allotropic theory. If we consider an iron-carbon alloy with a C content of about 0.4%, it will appear that such an alloy, at temperatures above 760° C, is present in the form of a solid solution of carbide in iron, i.e., in the form of martensite. On cooling below 760° C, ferrite begins to separate out, while the solvent—martensite—is enriched in carbon, and the carbon content will increase until, at a temperature of 680° C (690°), its amount rises to the content of the eutectic mixture (0.8% C); then, at the lower point corresponding to line DE, the whole alloy will transform into an intimate mixture of ferrite and cementite, i.e., into pearlite. The reverse case occurs when the alloy is heated, but then the critical point, for the most part, appears somewhat higher. If, however, this same iron-carbon alloy is quenched at temperatures above those indicated by line AB, then the martensite of the alloy will remain in the solid state, and the alloy itself will be in a state of supersaturation; a slight rise in temperature above that indicated by DE is enough for this unstable state of the alloy to be disturbed and to pass into another, stable state, in which the carbon separates out in the form of pearlite (tempering of steel begins). If the quenching temperature lies between AB and DE, then ferrite also forms together with martensite; quenching at a temperature below DE has the same effect on the structure of the metal as slow cooling.

A eutectic alloy with 0.8% C, having only one critical point, at temperatures above it must consist only of martensite, and below it—only of pearlite.

Summarizing all that has been said above about the theory of carbon alloys of iron in general, we obtain the following table:

Thus the proposition that carbon steel is entirely analogous to alloys in general and to a solution of common salt in water in particular, and that, moreover, like sulfur, it undergoes allotropic changes, is fully proven.

In most varieties of carbon steel, three critical points are observed: Ar₃, Ar₂, and Ar₁; the most important and characteristic point is Ar₂, for iron heated to Ar₂ rapidly loses its magnetic properties. The position of this point Ar₂ is determined differently by different investigators, thus:

The data of various authors, as is evident, differ greatly among themselves, but one must not forget that the position of the critical points in general, and in particular \(Ar_2\), depends on the chemical composition of the iron being tested, and the point \(Ar_2\) lies the lower, as Curie’s observations have shown, the more impurities the iron contains. But besides this, the position of the critical points in general is appreciably influenced by the external pressure under which the cooling of the metal takes place. External pressure lowers or raises the critical points of the metal; everything depends on what change of volume of the metal is observed in passing through the critical points. Thus Prof. Mallard and Le Chatelier, by artificially changing the external pressure, lowered the critical point of silver iodide by more than 100°). Roberts-Austen) in a similar way diminished the recalescence of metals and in general proved that compression of the external shell of steel during hardening has a tremendous influence on the structure of the internal parts of the object being hardened; Prof. Akerman**) holds the same view.

Application of metallography. As regards the application of metallography, “I should make a mistake,” says Prof. Heyn, “if I were to begin by asserting that it is called upon to replace the practically tested means for examining iron: chemical analysis and mechanical testing of materials. The task of metallography is not replacement, but supplementation.” Metallography is most important as a means for studying the nature of iron and is applied where chemical analysis, which in the present case is quite analogous to the analysis of rocks, does not give sufficient indications for conclusions. In this case metallography renders us an indispensable service, and if we were stubbornly to refuse it, this would be equivalent to the case in which, while seeking the causes of the properties of some rock, for example granite, we stubbornly refused to investigate what constituent parts granite consists of and what properties these parts possess. Moreover: strong magnification by the microscope reveals to us flaws and defects of the metal which are quite invisible to the naked eye, but very often entail substantial damage in practice. As an example I shall point to the case cited by A. Rzeshotarsky†), of the breakage of a shaft manufactured by a German plant for a steam engine in Moscow. Although the chemical composition of the metal was quite

*) Comptes rendus. 1884. t. XCIX, p. 157.
*) Proc. Inst. Mech. Eng. 1893. p. 124.
**) Stahl und Eisen 1897. S. 913,
†) A. Rzeshotarsky. Microscopic investigations of iron. P. 100. 1898.

satisfy the intended purpose and, apparently, appear to be free of defects; however, microscopic investigations revealed that the entire surface of the polished section was dotted with the finest lentil-shaped pits, which, without doubt, were the cause of the shaft’s fracture (Fig. 20, Plate III).

There is nothing surprising in the fact that slag interlayers can be seen very well with the aid of the microscope (Fig. 21, Plate III), as can internal tears produced during the rolling of the metal (Fig. 22, Plate III), gas bubbles of the smallest form (Figs. 23, 24, Plate III), completely invisible to the naked eye, and places where pieces of iron have been welded together (Fig. 25, Plate IV).

Some metallurgists of a later time have attempted to apply microscopy to determine the influence of foreign impurities on the structure and quality of iron. Thus, according to the observations of A. Rzeshotarsky, the improving properties of nickel consist in the fact that, under the influence of nickel, ferrite separates out in the form of a very complex network and, as if with nails, joins the pearlite grains (Fig. 26, Plate IV). The higher the nickel content in iron, the more strongly the connection of the pearlite grains increases—the branches of ferrite penetrate more deeply into the pearlite grains (Fig. 27, Plate IV). The appearance of ferrite grains in nickel steel and their relative hardness indicate that the ferrite grains consist not of pure iron, but of an alloy of iron with nickel. Such branching of ferrite grains within pearlite grains gives nickel steel great viscosity and toughness. Extensive chemical and microscopic investigations by Andrews*) of various fractures of locomotive axles, tires, rails, and propellers make it possible to believe that the main cause of these accidents was the presence of sulfur. It turns out that sulfur, by forming certain chemical compounds with the elements of iron, forms inclusions, like gas bubbles, among the structure of the iron and thereby, of course, considerably weakens the resistance of the metal (Figs. 28, 29, Plate IV). (The black spots on polished sections are sulfurous compounds of the metal—inclusions.)

Only thanks to observations in metallography was it possible to confirm the very serious hypothesis of Prof. Chernov on the hardening and annealing of steel. On the question of segregations, metallography can also provide very valuable indications. The phenomena of the separation of ferrite into cementite, in the broad sense of the word, must be classed among segregations. There is no carbonaceous iron, except eutectic iron, which would not manifest

*) Microscopic interval flaws inducing fractur in steel by Thomas Andrews London. 1896.

segregation. But in the narrower sense of the word, the segregation of iron is understood as the phenomenon of enrichment of the content of impurities in that part of the alloy which solidifies last of all; and this follows directly from the theory of solutions. In weak solutions the pure solvent first solidifies—in the present case carbon-bearing iron—while the liquid mass becomes still more enriched with foreign bodies, accumulates in the middle of the ingot, and solidifies. Analytically this has long since been proved, and metallography gives a vivid picture of this phenomenon by preparing sections for this purpose by etching (Figs. 30, 31, Table IV).

The phenomenon of segregation produces the most serious consequences in the casting of iron ingots intended for subsequent working, since the accumulation of foreign impurities in the middle of the ingot considerably lowers its homogeneity and mechanical properties. Analyses of such segregations in the middle of the ingot give the following results*).

Thus the internal inclusions of the ingot consist chiefly of oxides of the principal elements of iron; and, moreover, such a separation in ingots is most often encountered in the upper third of the ingot. Consequently, all this taken together makes it possible to put forward the following proposition: oxygen, dissolved in the molten metal during the cooling of the latter, gradually rises toward the top of the ingot and along the way oxidizes the principal elements of iron, Fe, Mn, Si, forming with them low-melting slags; these slags, as well as those which remained in the ingot from the smelting, owing to their lower specific gravity, float to the top of the ingot and collect in its upper third. The impurity in slags, CaO, S, and P₂O₅, is most often encountered in ingots of open-hearth steel when the smelting is conducted from a charge with a large content of pig iron and, in order to accelerate the course of the smelting process itself, ore and iron scrap are added to the charge; thus CaO, S, and P₂O₅ probably pass into the iron even during the smelting itself, as constituent parts of the charge.

Serious attention must be paid to the harmful consequences of steel segregation when casting critical shafts; the steel must not be greatly overheated before casting, since with temperature the susceptibility of steel to segregation increases. Thus, according to Arnold’s investigations**), the chemical composition

) Stahl und Eisen. S. 41. 1897.
*) Stahl und Eisen. S. 307. 1897.

one cast shaft, the fracture of which caused a great disaster, was, in different places, completely different:

It is evident that the chief cause of the shaft’s fracture was its inhomogeneity; the middle of the shaft, owing to the presence of a large quantity of sulfur and phosphorus, was weaker than its edges, and therefore the destruction of the shaft began gradually, imperceptibly, from its middle toward the periphery, which was fully confirmed by Prof. Arnold’s microscopic observations.

Note. To reduce the phenomenon of segregation in steel, one should add from 60 to 120 grams of aluminum per ton of open-hearth steel and from 80 to 210 grams of aluminum per ton of Bessemer steel. See Borchers, Metallhüttenkunde Aluminium. 1900.

Very thorough microscopic observations were made by Stead ) on the causes of the fracture of rails and tires even after an insignificant period of service. Rails that have been in use are very easily destroyed when bent, if only the head of the rail is thereby working in tension, i.e., is turned downward. Smith *) had already noted that a rail which had been in service again acquires its good properties and differs in no way from a new one if only the head of the rail is planed off by approximately 3 mm in thickness; but he could not explain the reason for this phenomenon. Only Stead, under the microscope, succeeded in noticing on the inner surface of the head of a rail that had worked for a long time an entire series of tears penetrating to a depth of 0.1–0.25 mm, and never more than 3 mm, into the rail (Fig. 32, Plate IV). Further microscopic observations by Stead showed that the head of a rail, under the influence of the impacts of the train wheels, changes its structure over time and, correspondingly, its mechanical properties. The particles of metal, under the influence of the impacts of the wheels, begin to move in the direction of the train’s motion; the metal becomes compacted, loses its toughness, and finally, under the influence of further impacts of the wheels, begins to tear from the surface. A rail that has been in use for only a short period has only a head whose structure has changed, without the presence of tears; it is enough to heat such a rail to a bright-red heat and cool it slowly for all the good mechanical qualities of the iron to return again.

) Stead, “Microskopic metallography,” “Micro-Mechanicas examination of old Steel rails and tyres,” Journl of the West of Scotland Iron and Steel Institute*, Vol. IV, pp. 23 and 24.

) Stahl und Eisen, p. 304, 1897.

Finally, metallography has the last word in explaining certain highly interesting phenomena in the mechanical processing of metals in general. In a word, the task of metallography is extensive, and metallography, as a science, will undoubtedly have a great future.

For lack of a sufficient number of polished sections, in my brief survey I also have not mentioned the very original works of the most recent period in the field of metallography; however, I cherish the hope that in the very near future I shall again return to these highly interesting questions.

Table I.

1. \(V = 300\). Mild basic steel with \(0.14\%\) C. \(f\)—ferrite, \(p\)—pearlite.

2. \(V = 100\). Mild steel with \(0.45\%\) C. \(f\)—ferrite.

3. \(V = 100\). Mild steel with \(0.6\%\) C. \(f\)—ferrite.

4. \(V = 100\). Steel with \(0.72\%\) C. \(f\)—ferrite.

5. \(V = 100\). Specimen taken from the middle of an open-hearth ingot containing \(0.69\%\) C and \(0.48\%\) Mn, heated to \(550^\circ\) C and slowly cooled. \(f\)—ferrite.

6. \(V = 100\). Forged steel, \(f\)—ferrite.

7. \(V = 100\). Mild armor steel, rolled under rolls.

8. \(V = 100\). Steel with \(0.95\%\) C. Relief polishing. \(c\)—cementite.

Table II

No. 9. Cement steel with 1% C. Relief polishing. c — cementite. \(V = 100\).

No. 10. Steel with 1.24% C. Relief polishing. c — cementite. \(V = 1000\).

No. 11. Steel with 1.5% C. Relief polishing. c — cementite and m — martensite. \(V = 100\).

No. 12. Steel with 0.45% C, heated to 800° and quenched at 720° C. Relief polishing with etching. \(V = 1000\).

No. 13. Hard annealed steel. Pearlite. \(V = 600\).

No. 14. Forged crucible steel with 0.92% C. Granular pearlite. \(V = 1240\).

No. 15. Gray cast iron, refined in a Bessemer retort. g — graphite, p — pearlite. \(V = 200\).

No. 16. Steel with 0.45% C, heated to 825° and quenched at 720° C. Relief polishing with etching. Martensite. \(V = 1000\).

Table III.

Cement steel, \(c\)—cementite, \(m\)—martensite, \(ш\)—slag.

No. 21. \(V = 100\).

Thomas cast iron after 108 hours of annealing in wood charcoal. Relief polishing. \(g\)—(black) temper carbon, \(c\)—cementite, \(p\)—pearlite.

No. 19. \(V = 270\).

Hardened steel with 0.85% C. Martensite.

No. 17. \(V = 800\).

Cast steel, forged.

No. 23. \(V = 100\).

Cast steel, forged.

No. 24. \(V = 100\).

Rolled steel.

No. 22. \(V = 100\).

Specimen taken from a broken shaft.

No. 20. \(V = 500\).

Gray cast iron with 0.72% Si, 1.43% Mn, and 1.8% Ph. Relief polishing. \(p\)—pearlite (convex), \(c\)—cementite (stands out), and \(g\)—graphite.

No. 18. \(V = 90\).

Phototype by A. I. Vilborg.

Plate IV

No. 25. Specimen of weld iron. Relief. Polishing (the weld areas have a shiny surface). \(V = 1.1\).

No. 26. Nickel steel with \(0.25\%\) C and \(2\%\) Ni; \(f\)—ferrite. \(V = 100\).

No. 27. Nickel steel with \(0.3\%\) C and \(3\%\) Ni; \(f\)—ferrite. \(V = 100\).

No. 28. Steel with a content of \(0.02\%\) S. \(V = 500\).

No. 29. Steel with a content of \(0.03\%\) S. \(V = 500\).

No. 30. Slowly cooled ingot of strongly overheated iron.

No. 31. Slowly cooled ingot of weakly overheated iron.

No. 32. View of the head of a used rail from above. \(V = 5\).

Table V.

Fig. 8.

Labels and inscriptions visible in the diagram:

• \(1\%\), \(1\frac12\% C\), \(2\%\), \(3\%\), \(4\%\), \(5\% C\)
• \(1600^\circ\), \(1500^\circ\), \(1400^\circ\), \(1300^\circ\), \(1200^\circ\), \(1100^\circ\), \(1000^\circ\), \(900^\circ\), \(800^\circ\), \(700^\circ\), \(600^\circ\), \(500^\circ\)
• “Hardening begins”
• “Cast iron begins to solidify”
• “Eutectically frozen”
• \(Fe_3(C)\)
• “\(\beta\) begins to precipitate”
• “\(\beta\) is released”
• “\(\beta\) eutectically precipitated”
• “\(\gamma\)”
• “\(\delta\)”
• “Possible solidification in \(\beta\)-phase or transition to \(\beta\)-phase”
• “Possible separation of cementite from molten cast iron”
• “IV”
• “Separation of cementite from molten cast iron”
• “Pearlite”
• “Metastable alloy of \(Fe +\) carbon”
• \(A\), \(B\), \(C\), \(D\), \(E\), \(F\)
• \(x\), \(y\), \(z\), \(e\), \(v\), \(c\)

Fig. 8.

Labels visible in the apparatus sketch:

• \(C\)
• \(D\)
• \(B\)
• \(K\)

Fig. 4.

Labels visible in the apparatus sketch:

• \(A\)
• \(B\)
• \(C\)
• \(D\)
• \(E\)

Water

Fig. A.

Labels visible:

• \(A\)
• \(B\)
• \(C\)

Fig. 7.

Labels visible on the curve:

• \(A r_3\)
• \(862^\circ C\)
• \(A r_2\)
• \(768^\circ C\)
• \(A r_1\)
• \(690^\circ C\)
• Vertical axis: “Magnitude of the change in thermoelectric current \((y)\)”
• Horizontal axis: “Time”

Fig. 5.

Labels visible:

• \(g_1\)
• \(g_2\)
• \(P_t\)
• \(P_t + 10\% Rh\)
• \(A\)
• \(B\)
• \(C\)

Additional visible diagram labels:

• “Beginning of solidification”
• “End of solidification”
• “Alloy”
• \(90\), \(93.5\), \(95\%\)
• \(C\)
• \(D\)
• \(E\)

Fig. 1.

\(\delta_B\)—coefficient of rupture, in kg/mm²
\(\delta\)—elongation, in %
\(q\)—content of transverse section, in %

Mean value of crystallites, in \(0.0001\) mm.

Fig. 2.

Tempering from the martensitic state

\(\%\,Cu\)

Fig. 3.

Weight part \(Z\) in the alloy, in %.

Fig. 5.

Water

Fig. A.

Fig. 1a.

Tempering from martensitic hardening

Martensitic alloy

\(\%\,NaCl\)

Fig. B Fig. C Fig. D.

\(O\)—object, \(f\)—line of intersection of the object, \(p^e\)—plane of the reflected section, \(p^z\)—prism, in the falling beam, and \(\alpha\)—angle of reflection.