Experiment No. 21 (Plate XXV.).-This was practically No. 20 repeated, but with more pronounced radial arms, seven in number, having a greater initial slope towards the centre. Two layers of clay, in. thick each, were moulded over the radii as in No. 20, the surface being sanddusted.

The result was a very decided peripheral folding over the radial arms. Each fold was a more or less perfect ellipsoidal truncated dome, the longer axes being radial. The centre over the lenticular core rose in a small axial dome or boss, from which the radial folds were separated by synclinal hollows, excepting in one instance, where the central uprise became an irregular continuation of a radial fold.

There was a tendency to lift at the periphery on screwing up the apparatus, instead of a general doming up of the whole assemblage as in previous

cases.

This experiment is a purely artificial one, made with the object of analysing the movements of clay plates under certain exceptional conditions. They may never occur in nature.

Explanation of Plate XXV.-View showing peripheral folding in the form of truncated ellipsoidal domes radially arranged.

The axis of each fold, of which there are seven, is irregularly radial. The small domical uprise in the centre is seen.

Experiment No.22 (Plates XXVI. and XXVII.) : Spiral Shearing.-A single plate of clay, ğ in. thick,

was placed in the compressor. No bias was introduced in any form, excepting what might naturally have come about in rolling out the clay.

The result was most astonishing. On drawing in the band the clay began to shear and screw round on a spiral shear-plane of low inclination. As the screwing up proceeded the upper portion rotated rapidly to the right over the lower portion until the shear-plane measured 8 inches on the periphery. The upper surface assumed a truncated domical form, the pointed end of the screw being 3 inches high above the metal base-plate (see figs. 4, 5, and 6, Plate XXIV.).

The underside of the overriding plate is beautifully slickensided.

The circumference of the plate before compression measured 2 ft. 6 in. After compression,

1 ft. 10 in. = difference 8 in.

The surface of the top was in many places sheared into thin layers in spiral directions following the main shear.

The clay in the last three experiments was rather soft.

This was a most satisfactory and instructive experiment, demonstrating that massive beds responding to centripetal pressure under certain conditions shear in spirals. A combination of thin beds subjected to similar pressure would, as we have seen, and will see again in Experiment No. 23, develop peripheral folds before shearing.

This was technically a right-hand screw.