A factor of much importance in determining the character and position of folds is the mode in which the strata were originally laid down. As we have already learned, the sheets of sediment which cover the sea-bottom are, on a large scale, nearly level, but they often show slight departures from such horizontality along certain lines. These initial dips often determine the place of flexures, because they divert the compression from its horizontal direction.

Fig. 190. - Model showing the slip of folded beds upon one another. (Willis).

The effects of lat-eral compression are shown in Figs. 190 and 191,taken from the models experimented on by Mr. Willis, which, when strongly compressed, imitate with remarkable accuracy the structures which may be observed in folded rocks. Fig. 190 shows that in folding, the beds must slip upon each other, as is proved by the lines perpendicular to the bedding-planes, which were continuous before folding, but in the anticline are broken by the differential motion of the layers, each bed rising farther up the slope than the one beneath it. The same thing must occur in folded rocks, which sometimes show polished bedding-planes, due to the slipping of the beds upon one another. The series A to D (in Fig. 191) represents a model before and in various stages of lateral compression, and exhibits the effect of the slight initial dip at x in determining the position of the anticlinal fold, which is developed by compression. The formation of one fold assists in the develop-' ment of another, for it both changes the direction of compression and redistributes the load of overlying strata.

The arch of the anticline lifts the load and diminishes the weight upon the beds that lie beneath the flexure, but increases the weight upon the lines from which the arch springs.

Fig. 191. - Model showing effects of lateral compression. A, before folding, with slight initial dip at x, B, C, D, in various stages of compression. (Willis).

There is much independent evidence to show that folding is a gradual process. The force exerted is enormous, but so is also the resistance to be overcome, and a steady or oft-renewed compression, acting upon strata under a great load of overlying masses, will produce regular flexures, where a sudden compression, however intense, could only shatter them.

Thrusts are likewise due to lateral compression, by which the rocks have been sheared and broken, and the beds on one side of the plane of fracture have been thrust up over those on the other. A plication or overturned fold may often be traced into a thrust, in a way that shows the direction of movement to have been the same in both fold and fracture. Numerous experiments also show that lateral compression will produce just such structures. A reduction of the overlying load, by diminishing the plasticity of the rocks, will occasion shearing and overthrusts, when, under a greater load, the same strata, exposed to an equal force of compression, will simply flex and bend. As we have seen (see Fig. 186), an anticlinal fold whose load has been reduced by erosion, will, on renewed compression, fracture and develop a thrust.

While thrusts are associated with violent folding, overturning and plication of strata, faults occur, as a rule, in regions where folding is absent, or very subordinate, or, if in areas of folded rocks, the faults were, generally at least, formed at a period more or less subsequent to the period of folding. The association of the different classes of faults shows that locally tension and compression may be generated in the same area and probably simultaneously. Reversed and horizontal faults are due to compression, the force in the latter case acting parallel to the fault-plane and in the former case across it, while normal faults are the result of a local tension. It is still an open question how these local compressions and tensions are generated.

Fig. 192. - Model illustrating the development of a fold-thrust.

One explanation is that such phenomena are developed in regions that have been raised by upwarping above a position of adequate support, whence results a system of fractures and the settling and readjusting of the fault-blocks. " If we endeavour to restore a system of normal fault-blocks to the relations which they may have had before faulting, we must commonly construct a dome-shaped figure of some sort, whose surface occupies more space than the displaced blocks occupy. That is to say, in any cross-section an elongated arc has in consequence of faulting been brought into a shorter chord, commonly by bringing the narrower parts of wedges into juxtaposition. . . . The doming may produce elongation or stretching in superficial sections at least, and thus tend to provide the opportunity for the development of planes whose attitude is that of the normal fault-plane. In so far as the inadequacy of support gives rise to vertical displacements pari passu with the stretching, the blocks will adjust themselves with reference to each other by relative displacement in the direction of maximum stress and least resistance. ... In this process elongation is the primary condition and a settling down of the blocks is a result. Through that settling a secondary effect of compression is set up.

The large masses become wedged against one another, and as their magnitude is such that their own weight is sufficient to deform them, they suffer more or less folding and even reversed faulting as an after effect. . . . We may reasonably expect to find some reversed faulting in connection with normal faulting wherever the latter is developed on a truly large scale The absence of folding or reversed faulting could only follow in case the blocks were free to move outward to the extent demanded by the elongation due to the attitudes of the normal fault-planes." (Willis).

In some cases, normal faults are due to pressure acting along and parallel to the fault-plane and causing the strata to arch gently upward on the upthrow side, downward on the downthrow side. Faults of this class have been observed in central Pennsylvania, Tennessee, and Alabama. Though due thus to pressure, a tension is developed across the fault-plane.

Quite a different type of explanation seeks to account for the phenomena of faulting by the transfers of molten magmas deep within the earth. In certain regions, as in the Tonopah district of Nevada, it has been made exceedingly probable that such transfers are the actual cause of the fracturing and dislocation of strata, and some observers would give this principle a widespread, if not a general, application. "Not only are the violent migrations of igneous material the cause of complex faulting, but also it is most reasonable to conceive that the deeper and more gradual movements of the subcrust are the cause of the larger fault systems. . . . Given this cause of faulting, the heretofore puzzling facts are satisfactorily and easily explained. Compression and tension still remain true causes of faulting, but mainly as local and proximate ones. The common expression, tilting of fault-blocks, attains a deeper significance, for this tilting may be more largely the result of subcrust migrations than of the mere force of gravity.

Cases of horizontal motion and pivotal motion become simple, for there is no necessary unchangeable relation between the direction of the force and the position of the fracture-plane." (J. A. Reid).

Even if it be granted that the effective forces which cause the folding and dislocation of rocks are, in the last analysis, a horizontal or tangential compression, it still remains to inquire how this great force was generated. There is no general agreement concerning the solution of this problem. For a long time it was supposed that a satisfactory solution was given by the contraction of the earth from cooling, and perhaps the majority of geologists still adhere to this view, which may be briefly expressed as follows: The earth's crust long ago reached a state of fairly constant temperature, but the highly heated interior is steadily cooling by radiation, and consequently contracting. As the crust cannot support itself, it must follow the shrinking interior, and is thereby crowded into a smaller space, thus setting up irresistible lateral stresses. If the earth were homogeneous, its surface would be wrinkled all over, as is the skin of a withered apple, of which the pulp contracts from loss of water, crowding the skin into a smaller space; but as the crust is heterogeneous, with special lines of weakness, the compression results in the formation of long, narrow belts of folded rocks, separated by broad areas of relatively little disturbed strata.

The contractional hypothesis has been attacked from many points of view, and very serious doubt has been thrown upon its adequacy to explain the facts, and even upon its reality. A modification of this hypothesis has been proposed by Professor Chamberlin, who regards the downward movement of segments of the earth's crust as primary and the horizontal movements as incidental to the former. The lithosphere is regarded as made up of a number of heavier and stronger segments, the surface of which forms the ocean basins, and of lighter and weaker segments which, on the surface, are the continental platforms. The general shrinkage of the earth causes the oceanic segments to descend, compressing the lighter continental segments and producing belts of folded rocks upon their borders.

The study of the radio-active substances and their distribution in the rocks of the earth's crust has led some observers to the conclusion that the earth's loss of heat is fully compensated by radioactivity and that since a very early period in the history of the globe, there has been no shrinkage at all, a standpoint which others have reached from entirely different lines of evidence and reasoning.

An elementary text-book is not the proper place for the discussion, or even the full statement, of all the different hypotheses which have been proposed in explanation of these most difficult problems. Suffice it to say that all the questions concerning the mechanics of the earth's interior are bound up together in an indivisible unity and that the full and satisfactory answer to any one question will involve the solution of all the cognate problems.