Experiment No. 8.-Fig. 1, Plate XIV., is a photograph of an anticlinal built up of seven layers of clay with calico partings (as in preceding experiments), compressed from the ends, which were square and vertical, giving thereby parallel or nonconvergent pressure, and consequently no domed development ensued. It will be observed that the building of the anticlinal took place near to the moving end of the compressor.

Experiment No. 9.-Fig. 2, Plate XIV., is a photograph of the anticlinal represented in diagram, fig. 5, Plate XLII., 'Origin of Mountain Ranges,' but showing the opposite side.

In this case the ends of the compressor were formed in step fashion, so that the pressure should first affect the basal layers and successively the superincumbent layers.

This model is to represent what I maintain actually takes place in the earth where the movements of the earth's crust are greatest at considerable depths, tailing off to nothing at the surface. It will be observed that the strata of the anticlinal are more compressed than in the preceding experiment and the tension at the apex greater.1

filled with the material to be acted upon, which in figs. 4, 5, 6 consisted of layers of a fine tenacious clay. The pressure was applied with a cramp against blocks of wood sliding in the wooden trough and acting on the zinc bands. The layers of clay were free to move vertically. In correction I may add that the zinc bands were only used in model fig. 4; in 5 and 6 stepped blocks of wood were used, as is, indeed, described in the details of each experiment.

The reader who wishes for further particulars should refer to the original work.

1

A paper by F. A. Steart, Esq., upon Overthrusts and other

Fig. 3, Plate XIV., is a photograph of the contorted beds figured in 'The Origin of Mountain Ranges,' fig. 6, Plate XLII., but exhibiting the opposite side. The object of this model is to illustrate what would take place if the surface layers of the earth were compressed the most, and the pressure decreased to zero at a certain depth below the surface, which would happen on the contraction theory. The result has been attained by reversing the stepped end of the compressor, so that the pressure came first upon the top layers. In neither of the last two examples were there any calico partings between the layers.

Disturbances in the Braysdown Colliery,' Somerset (Q. J. G. S., vol. lviii. pp. 609-17) describes phenomena bearing very closely upon this experiment. Considerable overthrust faults were disclosed by the workings, similar in character to those at Radstock, but of less overlap and throw. Mr. Steart considers both sets to be part of the same movements. The interesting fact recorded is that the overlap increases rapidly with the depth. The bottom beds are evidently longer than the top beds. This is a feature to be expected on the principles enunciated in this work and 'The Origin of Mountain Ranges' and illustrated in Experiment No. 66 (p. 153). A rise of temperature proceeding from below upwards would expand the bottom beds first and to the greatest extent, other things being the same. On the other hand, the beds of different kinds of rock have varying coefficients of expansion, and this, again, on a change of temperature will cause differential movements. Mr. Steart also proves that the soft black shales have got stripped off in one place and piled up in another, making up what the miners call 'dead ground.' Careful detailed observations such as these are invaluable.

CHAPTER XIV

EXPERIMENTS IN THE COMPRESSION OF SHEETS OF DIFFERENT SUBSTANCES BY CONVERGING PRESSURE APPLIED AT THE EDGES

THE FORMATION OF DOMED ANTICLINALS

EXPERIMENT No. 10.-Preliminary to the

actual compression of sheets into domed structures, with the object of elucidating the movements and flow of material that must take place in their formation, I took a piece of apron linen and cut out a circle 9 inches in diameter. The problem I set myself was to reduce the periphery to 8 inches diameter. With material such as cloth the only possible mode of reducing the circle is by folding at the edges, the tendency being to force up a quasi-dome formed of radial anticlines. An examination of Plate VIII., fig. 1, p. 150, will show that in this case I reduced the circular area covered by the linen cloth from 9 inches to 8 inches diameter by folding at the periphery in four places, the folds being pinned down to the board. This reduced the area covered by the cloth from 81 circular inches to 64 circular inches, which formed the area of the base of the intended dome. These folds, being all in one direction, created a