DISLOCATIONS OF GYPSUM

A. M. Kuz’min

Gypsum under natural conditions occurs in a variety of morphological forms and often as crystals (Fig. 1), whose common faces are \(\{010\}\), \(\{\overline{1}10\}\), \(\{111\}\), \(\{\overline{1}11\}\); less frequently the faces \(\{101\}\) and \(\{\overline{1}01\}\) occur. For gypsum a number of cleavage directions have been described \([1, 2, 3, 7]\). Parallel to \(\{010\}\) there is very perfect cleavage. Less perfect cleavage appears parallel to \(\{111\}\), the fracture along which is usually characterized as “fibrous.” The same less perfect cleavage runs parallel to the face \(\{100\}\), for which, however, a “fine-conchoidal” fracture is characteristic.

Fig. 1. Gypsum crystal.

O. Mügge established for gypsum the presence of translational slip \(T\) in the plane \((010)\) in the direction of the third coordinate axis \(Z\), while the axis of bending—\(f\) (crumpling) of gypsum plates lies in the plane \((101)\) and is perpendicular to the edge \((100):(010)\) \([5, 6]\).

According to Johnsen, the translational capacity of gypsum decreases with decreasing temperature and becomes insignificant at a temperature

liquid air. Under the conditions just described, no fibrousness appears on the \(\{\overline{1}11\}\) plane: the plane remains even and smooth [4].

A very important point for the correct setting of the gypsum crystals under study and of the position of cleavage cracks in them is that \(N_g\) makes an angle of \(14\)—\(20^\circ\) with the trace of the \((\overline{1}01)\) plane, and \(52^\circ 32'70''\) with \((100)\). The angle between \((\overline{1}01) : (100) = 66^\circ 9.5'\).

For the purposes of studying dislocations in gypsum, we are more interested in its relatively large crystals, which are known as “Mary’s glass.” The latter occur in comparatively large and giant crystals and are usually characterized by large dimensions (from 5 to 20 cm and more\(^1\)) and, as a rule, are colorless. Studies show that such crystals bear traces of deformations, some of which owe their origin to the results of extracting the crystals from pockets and splitting them along 010 into plates of various thicknesses. However, “Mary’s glass” under natural conditions for the most part proves to be deformed, and the deformation itself is manifested in the form of sliding in the plane of very perfect cleavage and in the appearance of cleavage cracks along \(\{101\}\) and \(\{100\}\) and other rare directions. Finally, the gypsum plates themselves prove to be, to varying degrees, wavy or intricately wavy, crumpled in one or in two different directions. The phenomena of translations, which for gypsum were described in detail by O. Mügge [5], deserve special attention.

The properties of gypsum just enumerated and its types of deformation, as it seemed to me, provide grounds for asserting that slip in it can develop only along the cleavage plane and that such slip causes the appearance of a fibrous fracture, while the cleavage along \(\{\overline{1}11\}\) may thereby be mistakenly taken as such. At least the surface corresponding to \(\{\overline{1}11\}\) on gypsum crystals is real; it usually looks more or less even, and more often proves to be longitudinally striated to varying degrees. The fracture surface parallel to it is always, to varying degrees, stepwise striated. However, cleavage parallel to this plane, during deformation, does not reveal itself either on the surface of the crystal or inside it in the form of cracks.

To confirm what has been said, collections of gypsum from many deposits, held in the museums of Tomsk higher educational institutions, were studied in detail. The study of large gypsum plates from deposits in the area of Kungur, from the area of Pavlodar, and from other places first cast doubt on the correctness of establishing the presence of cleavage parallel to \(\{\overline{1}11\}\). Thus, on many gypsum specimens there is an even or stepwise-even fracture surface without visible traces of fibrousness, which may be regarded as a clear cleavage making, with \(\{010\}\) on the crystals studied, an angle of from \(90\) to \(70^\circ\). Such a change in the edge angle\(^2\) could be observed on one and the same plane over a length of 6—10 cm; in a number of cases this angle at one end of an even cleavage was \(90\)—\(88^\circ\), at intermediate points was successively 85, 82, 76, and at the opposite end of the fracture became \(71^\circ\).

Finally, on gypsum crystals of the Mary’s-glass type from Uzbekistan, which had undergone deformation under natural conditions, one can observe the cleavage plane \(\{\overline{1}01\}\), which on all specimens was even and smooth and made an angle of \(90\)—\(88^\circ\) with the \((010)\) plane. Inside the specimens there were cracks of various extent parallel to \(\{101\}\). On other

\(^1\) In individual cases up to 50 cm across.

\(^2\) The angles were measured with a contact goniometer.

in the specimens this same plane made an angle of 89–85° with (010), and the surface itself appeared flatly convex or convex-concave, i.e., the observed cleavages parallel to \((\bar{1}01)\) are nothing other than perfect cleavage, whose position with respect to (010) changes as a result of the unevenly developing, during deformation, translation of parallel blocks.

When the same specimens of Uzbek gypsum were examined under the microscope, the following picture could be observed. During deformation, cleavage pieces of gypsum (Fig. 2a, b) are separated by parallel (010) planes, as shown in the figure, into individual plates—blocks, the \((\bar{1}01)\) surfaces of which, when viewed under the microscope,

Fig. 2. Schematic sketch of the cleavage surface of gypsum \((\bar{1}01)\): a, b — surface with traces of parallel blocks; c, d, e, f — sections perpendicular to \((\bar{1}01)\) and 010 of cleavage pieces, illustrating the relief of the \((\bar{1}01)\) surface.

appear concave and grooved, while at times the surface of some plates seems convex. Such figures may therefore be compared with frozen small waves with relatively sharply projecting crests of the ridges and gently concave intervals between them (Fig. 2c, d, e, f). The picture described, in all its details, very closely recalls the picture of deformation of rock salt described by me in the corresponding work [8, Fig. 9].

On the same fracture surface \((\bar{1}01)\), a very fine striation is visible, parallel to the trace of very perfect cleavage. A typically fibrous fracture on this plane can be obtained by breaking with bending to one side and the other, or by stretching (tearing). At the place where one part of a gypsum plate separates from another, plastic deformation with cylindrical bending of part of the (010) and \((0\bar{1}0)\) plane near the fracture edge is clearly visible on the (010) face. The greater the angle of inclination to one side and the other, the steeper the cylindrical surface of plastic deformation appears. When one part is torn away from the other, the same character of deformation is obtained, but the cylindrical bending of the (010) plane near the fracture edge is almost imperceptible. In this case the thin plates along the edge of the tear become slightly twisted.

The fracture surface obtained by one method or another is, in general, parallel or almost parallel to \((\bar{1}01)\), and with (010) makes an angle of 90–85°. The surface itself is characterized, as it were, by a fine-fibrous structure and by a clearly expressed silky luster, usually characteristic of parallel-fibrous aggregates of gypsum.

When viewed through a loupe, it is clearly seen that such a fracture surface consists of a series of parallel faces (010), ridges, and troughs; in a number of cases the ridges are truncated by a narrow plane \((\bar{1}01)\), and such narrow planes can also be seen at the bottoms of the troughs; the slopes of the ridges throughout prove to be stepped, or to carry ridges of the second order.

If reflections are obtained from the fibrous fracture surface of gypsum, situated perpendicular to the transport plane, then the angles formed by the reflecting surfaces with the plane (010) will have the following values: 51—53°, 58—62°, 66—73°, 78—81°, 85—90—86°, 68—75°, 62—53°. The reflection signals within 85—90—86° are obtained from numerous narrow strips, separated as if by dark strips. Signals from the planes forming angles of 66—73° are brighter and broader, separated by narrow dark strips. The reflecting planes with angles of 58—53° are likewise very narrow and are separated by comparatively broader dark strips. The intensity and number of reflections from 58 to 51° decrease.

When the surface with fibrous fracture is examined under a microscope, it proves to consist of a system of parallel and alternating white and gray strips that sparkle in reflected light. The width of the light “shining” strips ranges from 0.0010 to 0.0005 mm and less; rarely the width of the strips reaches 0.0025 mm; the width of the gray strips varies within the same limits, but often their width is two to three times greater than that of the light ones. However, when a magnification of 250—400 times is used, the relatively broad gray strips in turn prove to consist of still finer alternating strips, gray strips 0.0005—0.0001 mm wide and light strips 0.0001 mm and thinner.

The regularity and relief in the alternation of gray and light strips appear especially clearly if the gypsum specimens being studied under the microscope are illuminated, with the aid of a long-focus loupe, by a bright beam of light grazing along the fracture surface. In this latter case, the inclined fracture surfaces of gypsum turned toward the light, reflecting and refracting the light, appear shining—white, while surfaces parallel to \((\bar{1}01)\) in the rays of grazing light appear gray; planes that are usually inclined in the other direction appear dark and almost black. When the microscope tube is lowered or raised, the shining-white and dark strips shift to one side or the other, and the gray strips shift together with them.

Measurements on a goniometer of cleavage pieces up to 2 mm thick, with even planes (010) and with a well-pronounced “fibrous fracture” on the surface parallel to \((\bar{1}01)\), gave the following results. The signals from the faces (010) are sharp; the scatter does not exceed two degrees. The signals from the fibrous fracture surface in the same zone give a clearly visible continuous light strip, in which only with great difficulty can one distinguish certain signals corresponding to \((010):(\bar{1}01)=86—90^\circ\); the ratios \((010):(hkl)\) correspond respectively to angles of 87—90°, 82—78°, 72—70—66°, 58—50°, i.e., the scatter of the signals corresponds to an angle of 40° on either side of the signal corresponding to \((\bar{1}01)\).

The data obtained from the goniometric measurements and the study of the fibrous striation indicate that the very fine fibrousness is nothing other than very thin plates protruding relative to their neighbors to different heights. In addition, during fracture the free ends of the plates are deformed, and in a number of plates the end part of the “fibrous fracture” proves to be inclined and bent, forming oblique angles with (010).

An explanation of what has just been said can be found in the following experiment. From a gypsum plate \(0.9\) cm thick, it was necessary, with the aid of a fretsaw blade, to separate a cleavage piece along the plane \((10\bar{1})\). When sawing with a fine-toothed saw in the prescribed direction, the surface of the cleavage piece of gypsum near the cut acquired a cylindrical bending. As the saw cut into the gypsum, the latter, along the cleavage plane, broke up into separate plates, each of which likewise underwent cylindrical bending (Fig. 3). After the saw emerged from the gypsum, the lower surface \((0\bar{1}0)\) of the specimen also proved to be plastically deformed, with the formation of a cylindrical bend of greater curvature than the bend of the upper surface \(010\).

Let us next turn to consideration of the cleavage parallel to \(\{010\}\), \(\{100\}\), and \(\{\bar{1}01\}\), which arises inside gypsum crystals during deformation. Cleavage along \(\{010\}\) is regarded as very perfect. By means of this cleavage, gypsum crystals can split into very thin leaves, and under appropriate conditions its crystals break up into plates of various thicknesses. The same cleavage plane within crystals manifests itself not only as a result of sliding, but also as a result of any kind of bending. In this process, within the crystal there occurs a lamination of gypsum along the plane \((010)\), with the formation of irregular, often very complex outlines of void-lenses (Fig. 4a). In their peripheral part there are two, three, four, or more colored rings analogous to Newton’s rings. In the central part of such lenses, in a number of cases, the colored rings are absent. In the regions of lamination there is a clearly expressed pearly luster, which, together with the colored rings, indicates that in the places of lamination and bending the lenses that arise are hollow (Fig. 4).

Fig. 3. Splitting along the cleavage \(\{010\}\) and spherical curvature of gypsum plates as a result of deformation. The thin line indicates the position of the cleavage plane \((10\bar{1})\).

The void-lenses described, as observations show, appear in groups of two, three, four, and more, arranged one above another (Fig. 4b). The thickness of such lenses, from the formula \(d_m=\frac{r_m^2}{2R}\), is measured from fractions of a micron to several microns (\(d_m\) is the thickness of the lens, whose radius is \(r_m\), and the magnitude of the curvature of the lens itself is equal to \(2R\)) [9].

Further, within gypsum crystals there quite often develop cracks (Fig. 5) of clear cleavage along \(\{\bar{1}01\}\), which either cut through the cleavage piece of gypsum (Fig. 6a) from one of its surfaces \((010)\) to the other \((010)\), or, in the form of isolated narrow, long, wedge-shaped planes, penetrate the crystal at different levels (Fig. 6c). Such cleavage cracks more often develop from the periphery in regions of greatest deformation (Fig. 5), but they also appear inside the crystal. The cleavage planes along \(\{\bar{1}01\}\) within the crystal, wedging in, often split—

Fig. 4. Delamination and splitting of gypsum crystals along the \(\{010\}\) cleavage.

Fig. 5. Cleavage cracks in gypsum, parallel to \(\{101\}\) and \(\{\bar{1}00\}\); b and c—relief of the \((101)\) surface.

The split portions deviate to one degree or another from the original direction and splay out, as is shown in the photograph (Fig. 6, б).

The cleavage planes \(\{\overline{1}01\}\), relative to \((010)\), are inclined at different angles from 90 to \(60^\circ\). As a result of slip, each such plane, as is clearly visible under the microscope, breaks up and splits into a large number of narrower planes bounding steps parallel to \((010)\). Under the microscope, at magnifications of \(45\times\) and \(200\times\), the width of the steps ranges from 0.0005 to 0.036 mm. Along their length the steps alternately widen and narrow. The height of the steps ranges from 0.005 to 0.050 mm; however, the surface bounding the steps usually appears inclined and, in transmitted light, breaks up into smaller steps with heights of 0.002–0.008 mm. The number of such steps, for example, in a plate 0.8 cm thick is not less than 100–300. With the same plate thickness and a step height of 0.0005–0.001 mm, their number may exceed 800.

Fig. 6. а — a cleavage fragment of gypsum measuring \(80 \times 70 \times 10\) mm. As a result of deformation, a broad crack develops, parallel to the cleavage plane \((\overline{1}01)\); б — the same crack plane at enlarged size (\(9\times\)); it is clearly visible that this plane inside the crystal splits into separate parts—tails—which are partly cylindrically bent; в — planes analogous to the preceding one, in the direction toward the center, wedge out and break up into systems of thinner cleavage plates, appreciably displaced relative to one another; г — a complex cleavage plane along \((\overline{1}01)\) inside the crystal is cut by a thin cleavage plane along \((100)\), vertical to the figure, and in the form of thin threads splays out in a complex manner.

Cleavage parallel to \(\{100\}\) (Fig. 5) is usually characterized as having a conchoidal fracture [3, 7, 5]. Study of a large body of gypsum-crystal material shows that the surface of this fracture is not always described correctly. Indeed, in a number of cases the fracture surface bears clear signs of small conchoids (Fig. 7, а). However, one can often observe on the surface parallel to \((100)\) a rougher, more fibrous fracture than on \((\overline{1}01)\), or this fracture surface breaks up into a series of parallel \((010)\) plates 0.2–0.5 mm thick, and within each such plate there appear small narrow conchoids with elongated wings, resembling certain kinds of spirifers. The layers of conchoids are separated from one another by several thin...

plates, clearly distinguishable over the entire extent of the fracture surface. This latter surface, the surface of the cavities, and the spaces between them are likewise affected by very fine striation parallel to $(010)$, caused by translational slip in the direction $[100]$.

No less often, on the fracture surface $(100)$ one can observe fine and relatively coarse striation and ribbing, caused by protruding individual plates, which at their free ends prove to be somewhat deformed, while the spaces between them are grooved. The platy character arising during plastic deformation, visible on the face $(100)$, generally has the same dimensions as on the face $(\overline{1}01)$. Cleavage cracks parallel to $(100)$ appear inside the crystal as relatively broad and more often narrow planes, which, before wedging out, often split into very fine fibrous formations and, branching and bending, deviate from their rectilinear direction and in a number of cases pass, in the form of hairlike or fibrous cracks, into a direction parallel to $(\overline{1}01)$ (Fig. 7, b).

Fig. 7. a — relief of the cleavage surface along $(100)$ (shell-like fracture); b — in a section parallel to $(010)$, in crystals of gypsum it is seen how cracks that branch off from cleavage cracks along $(100)$ then pass into cracks parallel to $(\overline{1}01)$.

With respect to $\{010\}$, the described cleavage $\{100\}$ is often situated perpendicularly, or in a number of cases it forms oblique angles of up to $80^\circ$ with the plane of very perfect cleavage. On the surface of the cleavage $(100)$ developing inside the crystal, a pearly luster is also clearly noticeable, and in transmitted light under the microscope it is quite clearly seen that this cleavage plane too is complex and bears distinct traces of a stepped structure. This latter circumstance indicates that translational slip also develops in the direction $[100]$ and causes the visible inclination of $(100)$ to $(010)$, and, on the plane $\{110\}$, a clearly distinguishable longitudinal striation.

As for cleavage along the plane $\{\overline{1}11\}$, it is absent as such in gypsum crystals; cleavage along the plane $\{\overline{1}11\}$ was erroneously taken to be cleavage along the plane $\{\overline{1}01\}$, which, as a result of translational slip—T and plastic deformation of individual plate-blocks, forms oblique angles of intersection with the plane $(010)$, as was shown above.

From all that has been said above it follows that deformations of gypsum lead to the appearance of two kinds of plastic dislocations. Plastic deformations of the first kind appear in gypsum crystals in the form of slips in the plane \((010)\), predominantly in the direction of the edge between the faces \((100):(010)\), and to a lesser degree in the direction \([100]\). This type of deformation is closely connected with translational slip in the plane \((010)\) and is essentially a shear deformation which, developing in gypsum, as observations show, causes the appearance of a large number of elementary and mutually parallel, very thin lamellae-blocks. The width of the lamellae varies within broad limits from 0.001 to 0.0001 mm and less, while the magnitude of the shear, equal to (or a multiple of) the value \(C_0 = 6.28 \text{ Å}\), varies from 0.1 to 0.005 mm. During the slipping of one block over the surface of another, the primary bonds between the blocks are, naturally, ruptured along the cleavage \((010)\), and from the moment the motion of the blocks in the process of slipping stops, new, secondary bonds arise. Whether such differential slip is accompanied by hardening (apparently, it is accompanied) of the body remains unclear for the time being, but what is most important in this motion in the plane 010 is that ruptures between the blocks are not physically detected, although they are quite probable.

Slip-shears in gypsum are most readily realized in the plane \((010)\) in the direction of the third crystallographic axis, which, as we have seen, leads to a change in the position of the plane \((\bar{1}01)\) and causes the appearance of a fibrous fracture. The slip front in this direction develops at different speeds and by the system of thin lamellae. This latter circumstance leads to the occurrence of a flexure (bend) of small amplitude, the axis of which is parallel to \((100)\).

Along with translation along the \(Z\) axis, slips also occur in the plane \((010)\) in the direction \([100]\), which lead to the fact that the cleavage plane \((100)\) becomes inclined and forms angles with \((010)\) equal to \(90—80^\circ\). This slip may be independent or may be the result of uneven development of the translation front along the \(Z\) axis. In any case, as a result of slip on \((100)\), a fibrous fracture also develops, though more rarely; more often relatively narrow zones of fibrous fracture alternate with somewhat broader zones of conchoidal fracture, as is shown schematically in Fig. 7, a.

The second kind of plastic deformation in gypsum has the character of cylindrical or complex-spherical bending of lamellae. Most often the axis of such bending of various radius, lying in \((010)\), is, as O. Mügge [5] indicated in his time, the perpendicular to \((100)\); another direction of curvature is less sharply expressed, its axis lying in the plane \((010)\) and passing parallel to \((100)\). It is often necessary to note cases when the plane of gypsum lamellae proves to be complex-spherically deformed.

During cylindrical bending and complex warping in gypsum lamellae, their delamination occurs, developing on several parallel \((010)\) levels in separate segments (Fig. 4, b). The formation of cavities inside gypsum as a result of deformation also occurs along cleavage cracks, predominantly along \((\bar{1}01)\) and rarely \((100)\), whose surface, as a rule, bears a pearly luster.

O. Mügge [5, 6] explained the nature of bending of gypsum lamellae by the displacement of individual particles relative to neighboring ones along the plane \((\bar{3}01)\). As was indicated in the proper place (p. 56), a gypsum crystal under the corresponding direction and deformation breaks up into separate lamellae parallel to \((010)\), each of which thereby undergoes

to varying degrees, cylindrical bending (Fig. 3). Therefore such bending, as shown schematically in Fig. 3, can be explained by slipping (translation) along the \((\bar{1}01)\) plane of one block relative to another adjacent to it. When gypsum plates are bent, on the inner surface at the bend angles one can observe wrinkling, as a system of very small ridges; on the outer surface, ruptures in the form of a dense network of parallel \((\bar{1}01)\) cracks wedging inward into the plate.

Therefore there are grounds to believe that, upon bending, each plate parallel to \((010)\) breaks up into blocks (Fig. 8, a) bounded by cleavage surfaces \((\bar{1}01)\). At the same time, on the inner surface of the bent plate, owing to compression, parts of the blocks are tightly pressed against one another. Toward the outer surface the cracks of the blocks open and become wedge-shaped. Within each block there is sliding of very thin plates (Fig. 8, в, г), which correspondingly explains to us the behavior and structure of the inner and outer surface of each block of the plate.

Complex-spherical bendings of gypsum crystals naturally also give rise to more complex block slidings along all three \((010)\), \((\bar{1}01)\), and \((100)\) cleavage planes.

Fig. 8, б clearly and simply explains to us that the slidings of a large number of very thin plate-blocks (Fig. 8, в, г) parallel to \((010)\) along the \(Z\) axis make it possible to understand why the cleavage plane \((\bar{1}01)\), in a number of observed cases, proves to be convex or concave, and the angle of curvature increases toward the outer edge of the cleavage piece. The unevenness

Fig. 8. Schematic illustration making it possible to understand the spherical deformations of plate-blocks (a) and the deformation of the \((\bar{1}01)\) surface in cleavage pieces of gypsum (б).

of the sliding of individual packets of plates accounts for the appearance of the pattern (Fig. 2) that was described on p. 4 and makes it possible to understand correctly the nature of the striation on the faces \(\{\bar{1}11\}\) and \(\{110\}\).

As for the cleavage of gypsum along \(\{\bar{1}11\}\), it seems doubtful to us, as follows from the data presented above. Cleavage of one or another crystal can be obtained by splitting or breaking it. The application of such methods to gypsum, in order to obtain a cleavage plane parallel to \((\bar{1}11)\), as was indicated above, leads at the fracture site to a sharp deformation: the crystal breaks up into individual plates, each of which undergoes spherical deformation. The cleavage plane \((\bar{1}01)\) of each plate, as shown in Fig. 3, inclines, and with the former direction of the plane

(010) forms an oblique angle, the magnitude of which increases toward the outer edge of the crystal.

In conclusion it should be said that the study of a large number of gypsum crystals and experimentation with them convinces us that plastic deformations accompanied by gliding develop only in the direction of the cleavages existing in the crystals, as a special property of the crystal lattice strictly corresponding to a crystallographically defined direction, in particular in gypsum \((010)\), \((100)\), and \((\overline{1}01)\). Glides developing along the cleavage lead to rupture of the primary bonds and to a more or less pronounced splitting in the cleavage plane. What has been said fully confirms the considerations we expressed in the article “Cleavage and Glide Planes in Rock-Salt Crystals” [8].

LITERATURE

1. A. K. Boldyrev. Course of Descriptive Mineralogy. Issue III, Moscow, 1935.
2. J. D. Dana and E. S. The System of Mineralogy. N. Y. 1900.
3. J. D. Dana, S. S. Dana, W. E. Ford, G. Herman, C. Frondel. The System of Mineralogy. Vol. II, Pt. 1, Izd. IL, Moscow, p. 569, 1953.
4. Johnsen. Centralblatt. 1918, p. 234.
5. O. Mügge. Ueber Gleitflächen an Gyps, Antimon—glanz, Wismutglanz, Auripigment und Cynit. N. J. f. Miner., Geol. u. Paleont. II Bd. 1883.
6. O. Mügge. Ueber Translationen und verwandte Erscheinungen im Krystallen. N. J. f. Miner. Geol. und Paleont. I. Bd. 1898.
7. H. Rosenbusch. Mikroskopische Physiographie der petrographisch wichtigen Mineralien. Bd. 2. Hf. ST. 1927, p. 424.
8. A. M. Kuzmin. Cleavage and glide planes of rock salt. Geologiya i geofizika, No. 6, 1960, p. 60.
9. G. S. Landsberg. Optics, OGIZ, Moscow–Leningrad, 1947.

NOTED MISPRINTS