CHAPTER XVIII NORMAL OR CONTRACTION FAULTS E have now experimentally analysed the com WE ponent movements imparted to strata by lateral compression applied in several ways and under varied conditions. Furthermore, it has been shown that the internal stresses set up in the lithosphere by alternate expansions and contractions would result in movements of the rocks comparable to what we see in mountain chains and folded regions of the earth's crust. It has also been pointed out, both here and in 'The Origin of Mountain Ranges,' that a secular contraction of the earth's nucleus, on the contractional hypothesis, provides only continued or successive compressive stresses in the lithosphere, but no alternations with tensile stresses, such as are registered by normal faults in almost every accessible section of the crust of the earth. This, I conceive, is the inherent weakness of the contractional hypothesis. Suess, in his important work, 'The Face of the Earth,' clearly recognises the existence of two types of structure in the earth's crust. As admirably summarised by Mr. Teall, he holds that a study of the structure of the earth's crust proves that it has been affected by two types of movement-the one characterised by a compression of the stratified rocks along certain zones; the other, by a separation of the crust into blocks, some of which have sunk down relatively to others; and he goes on to say: 'One and the same area cannot be simultaneously affected by these two types of movement.' 1 Though these two types of structure are recognised, Suess apparently does not see their true relations; nor is it explained how this striking difference of effect should be produced by the same cause, for Suess seems only to recognise radial contraction in his theory of the earth. In 'The Origin of Mountain Ranges' I have attempted to show that the folding is due to lateral expansion, caused by a rise of temperature acting either continuously or successively on a compound aggregate of strata, termed by me the strata-plate.' The best conception of this action is to be gained by a study of Plates III., IV., V., and VI. of the present work. Normal faulting, by which blocks of the crust are let down, relatively to others, is the result of contraction following a fall of temperature. The folding is due to compression, the faulting to tension. Since The Origin of Mountain Ranges was published normal faults have been pretty generally recognised as due to tension. This necessary complementary and opposite stress to 1 Nature, December 19, 1901, p. 145. that of compression is, so far as I am able to discover, unprovided for by any theory of mountain building other than my own. The contraction hypothesis, as I have reiterated -but apparently none too often-only provides a continuously compressive force. Normal faulting results from the shortening of the strata, together with the inadequacy of the crust to maintain voids. If the materials of the crust were of sufficient rigidity and strength, the cubical contraction due to a fall of temperature would simply cause great rifts, which would remain open until filled up with débris, and the surface of the blocks, though sinking actually, would not be differentially lowered with respect to each other. The rigidity and tenacity of the materials of the crust are, however, but slight compared with the enormous forces simple gravitation lets loose. The yielding crust sinks differentially along huge shearing planes called faults, which it is well known 'hade to the downthrow.' The horizontal contraction is thus provided for by the keying up of the strata-plate in sections, which enables it to occupy a larger horizontal area than it otherwise would, and at the same time keeps the crust solid and continuous. Nor must it be lost sight of that a fall of temperature, within certain limits, increases from zero at the surface to a maximum at a certain depth, which depth is determined by several factors. The combined result is, that below a certain depth the normal faults may shade off and the vacuities-or rather what would otherwise be vacuities-in the crust become filled solid by a flow of rock rather than by keying up. The effects of all these factors in filling up tri-dimensional space, and maintaining a solid crust, are to be seen in the mapping of any well tension-faulted area. A secondary effect is often the turning up of the edges of the strata against a boundary fault, through the compression produced by the blocks keying up at the upper part before the mass comes to rest on a solid foundation. The fact, well known to engineers, that faults are often the most watertight parts of the strata, is another proof, if one were needed, that contractional faulting does not produce an open fissured crust, but that the keying by the sheared and sinking blocks goes on simultaneously with the sinking, and that enough lateral pressure is so produced by the gravitating mass to maintain a solid continuous lithosphere. There is every reason to infer, as geologists usually do, that the whole of these adaptations take place very slowly, whether accumulating stresses be relieved by small or large successive slips. That the forces to which these two broad types of structure-the folded and the normal faulted-are due cannot be simultaneously operative in the same area is very obvious, the one being the antithesis of the other. I have shown that no theory which does not provide for both movements can be a true explanation of the |