FAULT (Mid. Eng. faute, through the French, from the popular Latin use of fallere, to fail; the original l of the Latin being replaced in English in the 15th century), a failing, mistake or defect.
In geology, the term is given to a plane of dislocation in a portion of the earth's crust; synonyms used in mining are "trouble," "throw" and "heave"; the German equivalent is Verwerfung, and the French faille. Faults on a small scale are sometimes sharply-defined planes,  as if the rocks had been sliced through and fitted together again after being shifted (fig. 1). In such cases, however, the harder portions of the dislocated rocks will usually be found "slickensided." More frequently some disturbance has occurred on one or both sides of the fault. Sometimes in a series of strata the beds on the side which has been pushed up are bent down against the fault, while those on the opposite side are bent up (fig. 2). Most commonly the rocks on both sides are considerably broken, jumbled and crumpled, so that the line of fracture is marked by a belt or wall-like mass of fragmentary rock, fault-rock, which may be several yards in breadth. Faults are to be distinguished from joints and fissures by the fact that there must have been a movement of the rock on one side of the fault-plane relatively to that on the other side. The trace of a fault-plane at the surface of the earth is a line (or belt of fault-rock), which in geological mapping is often spoken of as a "fault-line" or "line of fault." Fig. 3 represents the plan of a simple fault; quite frequently, however, the main fault subdivides at the extremities into a number of minor faults (fig. 4), or the main fault may be accompanied by lateral subordinate faults (fig. 5), some varieties of which have been termed flaws or Blatts.
Fig. 2. - Section of strata, bent at a line of fault.
Fig. 3. - Plan of simple fault.
Fig. 4. - Plan of a fault splitting into minor faults.
"Fault-planes" are sometimes perpendicular to the horizon, but more usually they are inclined at a greater or lesser angle. The angle made by the fault-plane with the vertical is the hade of the fault (if the angle of inclination were measured from the horizon, as in determining the "dip" of strata, this would be expressed as the "dip of the fault"). In figs. 1 and 2 the faults are hading towards the right of the reader. The amount of dislocation as measured along a fault-plane is the displacement of the fault (for an illustration of these terms see fig. 18, where they are applied to a thrust fault); the vertical displacement is the throw (Fr. rejet); the horizontal displacement, which even with vertical movement must arise in all cases where the faults are not perpendicular to the horizon and the strata are not horizontal, is known as the heave. In fig. 6 the displacement is equal to the throw in the fault A; in the fault B the displacement is more than twice as great as in A, while the throw is the same in both; the fault A has no heave, in B it is considerable. The rock on that side of a fault which has dropped relatively to the rock on the other is said to be upon the downthrow side of the fault; conversely, the relatively uplifted portion is the upthrow side. The two fault faces are known as the "hanging-wall" and the "foot-wall."
Fig. 5. - Plan of main fault, with branches.
Fig. 6. - Section of a vertical and inclined fault.
Fig. 7. - Reversed fault, Liddesdale.
The relationship that exists between the hade and the direction of throw has led to the classification of faults into "normal faults," which hade under the downthrow side, or in other words, those in which the hanging-wall has dropped; and "reversed faults," which hade beneath the upthrow side, that is to say, the foot-wall exhibits a relative sinking. Normal faults are exemplified in figs. 1, 2, and 6; in the latter the masses A and B are on the downthrow sides, C is upthrown. Fig. 7 represents a small reversed fault. Normal faults are so called because they are more generally prevalent than the other type; they are sometimes designated "drop" or "gravity" faults, but these are misleading expressions and should be discountenanced. Normal faults are regarded as the result of stretching of the crust, hence they have been called "tension" faults as distinguished from reversed faults, which are assumed to be due to pressure. It is needful, however, to exercise great caution in accepting this view except in a restricted and localized sense, for there are many instances in which the two forms are intimately associated (see fig. 8), and a whole complex system of faults may be the result of horizontal (tangential) pressure alone or even of direct vertical uplift. It is often tacitly assumed that most normal and reversed faults are due to simple vertical movements of the fractured crust-blocks; but this is by no means the case. What is actually observed in examining a fault is the apparent direction of motion; but the present position of the dislocated masses is the result of real motion or series of motions, which have taken place along the fault-plane at various angles from horizontal to vertical; frequently it can be shown that these movements have been extremely complicated. The striations and "slickensides" on the faces of a fault indicate only the direction of the last movement.
Fig. 8. - Diagram of gently undulating strata cut by a fault, with alternate throw in opposite directions.
Fig. 9. - Section of strata cut by step faults.
Fig. 10. - Trough faults.
Fig. 11. - Plan of a strike fault.
A broad monoclinal fold is sometimes observed to pass into a fault of gradually increasing throw; such a fault is occasionally regarded as pivoted at one end. Again, a faulted mass may be on the downthrow side towards one end, and on the upthrow side towards the other, the movement having taken place about an axis approximately normal to the fault-plane, the "pivot" in this case being near the centre. From an example of this kind it is evident that the same fault may at the same time be both "normal" and "reversed" (see fig. 8). When the principal movement along a highly inclined fault-plane has been approximately horizontal, the fault has been variously styled a lateral-shift, transcurrent fault, transverse thrust or a heave fault. The horizontal component in faulting movements is more common than is often supposed.
Fig. 12. - Section across the plan, fig. 11.
A single normal fault of large throw is sometimes replaced by a series of close parallel faults, each throwing a small amount in the same direction; if these subordinate faults occur within a narrow width of ground they are known as distribution faults; if they are more widely separated they are called step faults (fig. 9). Occasionally two normal faults hade towards one another and intersect, and the rock mass between them has been let down; this is described as a trough fault (fig. 10). A fault running parallel to the strike of bedded rocks is a strike fault; one which runs along the direction of the dip is a dip fault; a so-called diagonal fault takes a direction intermediate between these two directions. Although the effects of these types of fault upon the outcrops of strata differ, there are no intrinsic differences between the faults themselves.
Fig. 13. - Plan of strata cut by a dip fault.
Fig. 14. - Plan of strata traversed by a diminishing strike fault.
Fig. 15. - Plan of an anticline (A) and syncline (S), dislocated by a fault.
Fig. 16. - Section along the upcast side of the fault in fig. 15.
Fig. 17. - Section along the downcast side of same fault.
The effect of normal faults upon the outcrop may be thus briefly summarized: - a strike fault that hades with the direction of the dip may cause beds to be cut out at the surface on the upthrow side; if it hades against the dip direction it may repeat some of the beds on the upthrow side (figs. 11 and 12). With dip faults the crop is carried forward (down the dip) on the upthrow side. The perpendicular distance between the crop of the bed (dike or vein) on opposite sides of the fault is the "offset." The offset decreases with increasing angle of dip and increases with increase in the throw of the fault (fig. 13). Faults which run obliquely across the direction of dip, if they hade with the dip of the strata, will produce offset with "gap" between the outcrops; if they hade in the opposite direction to the dip, offset with "overlap" is caused: in the latter case the crop moves forward (down dip) on the denuded upthrow side, in the former it moves backward. The effect of a strike fault of diminishing throw is seen in fig. 14. Faults crossing folded strata cause the outcrops to approach on the upthrow side of a syncline and tend to separate the outcrops of an anticline (figs. 15, 16, 17).
In the majority of cases the upthrown side of a fault has been so reduced by denudation as to leave no sharp upstanding ridge; but examples are known where the upthrown side still exists as a prominent cliff-like face of rock, a "fault-scarp"; familiar instances occur in the Basin ranges of Utah, Nevada, etc., and many smaller examples have been observed in the areas affected by recent earthquakes in Japan, San Francisco and other places. But although there may be no sharp cliff, the effect of faulting upon topographic forms is abundantly evident wherever a harder series of strata has been brought in juxtaposition to softer rocks. By certain French writers, the upstanding side of a faulted piece of ground is said to have a regard, thus the faults of the Jura Mountains have a "regard français," and in the same region it has been observed that in curved faults the convexity is directed the same way as the regard. Occasionally one or more parallel faults have let down an intervening strip of rock, thereby forming "fault valleys" or Graben (Grabensenken); the Great Rift Valley is a striking example. On the other hand, a large area of rock is sometimes lifted up, or surrounded by a system of faults, which have let down the encircling ground; such a fault-block is known also as a horst; a considerable area of Greenland stands up in this manner.
Faults have often an important influence upon water-supply by bringing impervious beds up against pervious ones or vice versa, thus forming underground dams or reservoirs, or allowing water to flow away that would otherwise be conserved. Springs often rise along the outcrop of a fault. In coal and metal mining it is evident from what has already been said that faults must act sometimes beneficially, sometimes the reverse. It is a common occurrence for fault-fissures and fault-rock to appear as valuable mineral lodes through the infilling or impregnation of the spaces and broken ground with mineral ores.
In certain regions which have been subjected to very great crustal disturbance a type of fault is found which possesses a very low hade - sometimes only a few degrees from the horizontal - and, like a reversed fault, hades beneath the upthrown mass; these are termed thrusts, overthrusts, or overthrust faults (Fr. recouvrements, failles de chevauchement, charriages; Ger. Uberschiebungen, Ubersprünge, Wechsel, Fallenverwerfungen). Thrusts should not be confused with reversed faults, which have a strong hade. Thrusts play a very important part in the N.W. highlands of Scotland, the Scandinavian highlands, the western Alps, the Appalachians, the Belgian coal region, etc. By the action of thrusts enormous masses of rock have been pushed almost horizontally over underlying rocks, in some cases for several miles. One of the largest of the Scandinavian thrust masses is 1120 m. long, 80 m. broad, and 5000 ft. thick. In Scotland three grades of thrusts are recognized, maximum, major, and minor thrusts; the last have very generally been truncated by those of greater magnitude. Some of these great thrusts have received distinguishing names, e.g. the Moine thrust (fig. 19) and the Ben More thrust; similarly in the coal basin of Mons and Valenciennes we find the faille de Boussu and the Grande faille du midi. Overturned folds are frequently seen passing into thrusts. Bayley Willis has classified thrusts as (1) Shear thrusts, (2) Break thrusts, (3) Stretch thrusts, and (4) Erosion thrusts.
Fig. 18. - Diagram to illustrate the terminology of faults and thrusts.
Fig. 19. - Section of a very large thrust in the Durness Eriboll district, Scotland.
Dr J.E. Marr ("Notes on the Geology of the English Lake District," Proc. Geol. Assoc., 1900) has described a type of fault which may be regarded as the converse of a thrust fault. If we consider a series of rock masses A, B, C - of which A is the oldest and undermost - undergoing thrusting, say from south to north, should the mass C be prevented from moving forward as rapidly as B, a low-hading fault may form between C and B and the mass C may lag behind; similarly the mass B may lag behind A. Such faults Dr Marr calls "lag faults." A mass of rock suffering thrusting or lagging may yield unequally in its several parts, and those portions tending to travel more rapidly than the adjoining masses in the same sheet may be cut off by fractures. Thus the faster-moving blocks will be separated from the slower ones by faults approximately normal to the plane of movement: these are described as "tear faults."
Faults may occur in rocks of all ages; small local dislocations are observable even in glacial deposits, alluvium and loess. A region of faulting may continue to be so through more than one geological period. Little is known of the mechanism of faulting or of the causes that produce it; the majority of the text-book explanations will not bear scrutiny, and there is room for extended observation and research. The sudden yielding of the strata along a plane of faulting is a familiar cause of earthquakes.
See E. de Margerie and A. Heim, Les Dislocations de l'écorce terrestre (Zürich, 1888); A. Rothpletz, Geotektonische Probleme (Stuttgart, 1894); B. Willis, "The Mechanics of Appalachian Structure," 13th Ann. Rep. U.S. Geol. Survey (1891-1892, pub. 1893). A prolonged discussion of the subject is given in Economic Geology, Lancaster, Pa., U.S.A., vols. i. and ii. (1906, 1907).
(A. Ge.; J. A. H.)
 The fault-plane is not a plane surface in the mathematical sense; it may curve irregularly in more than one direction.
Note - this article incorporates content from Encyclopaedia Britannica, Eleventh Edition, (1910-1911)