shows the lead plates of which the bar is composed.

Experiment No. 2. (Plate XII., fig. 2.)—A compound series was prepared, 1 ft. 5 in. long, consisting of three sheets of lead, three sheets of millboard, and five layers of paper, forming a bar 2 in. by

in. in section. This combination was put into the apparatus just described, excepting that the bridging piece was sawn through the middle and the clamps were nearer together.

On applying pressure to the ends of the bar as before the upper layers rose rapidly in anticlinal form, the bottom sheet of lead at first remaining flat; it then suddenly sprang up. The pressure on the anticlinal could be regulated by the clamp screws and the elasticity of the boards they acted upon. Continuing the screwing up, the left limb of the anticlinal became vertical, the right being at an angle of over 40°. After this, by the more rapid approach of the bottom part of the left limb towards the right limb, the anticlinal closed up and became an overturn to the left.

The clamp screws were then removed and the screwing up continued, when a second anticlinal was initiated to the left, the various layers of the bars separating, and rising one inside the other in anticlinal form.1 The right-hand end of the bar was unaffected.

Experiment No. 3.-The whole series. straightened out and placed anew in the machine. The photo does not show this.

M

Again the left-hand end moved most on screwing up, but this time the anticlinal became an overturn to the right.

These experiments seem to show that very minor causes may influence the direction of the movement of rocks subjected to great pressure.

Experiment No. 4.-A bar of soap, 1 ft. 23 in. × 2 in. x 21 in., was cut into two layers as shown in section fig. 1, Plate XIII., and then placed in the machine. The effect was to shear the ends sideways with a vertical shear (fig. 2). By no amount of coaxing, clamping, or regulation of the screw could I make this curious material act otherwise. The pressure and movement produced slickensided surfaces on the planes of shearing.

Experiment No. 5.-In this experiment a wooden trough was filled with alternate layers of moistened quartzose sand, and sand and coal dust mixed. This formed a bar 2 ft. 0 in. × 2 in. × 2 in. before compression. On applying compression the ends commenced to lift, so I placed weights upon them to keep the sand down. The effect was curious. The sand flowed from under and rose round the weights, and the bar of sand sheared beyond as at a, fig. 3, Plate XIII.1

After removing the distorted ends the bar measured 12 in. long; this was compressed to 11 in., but again the ends sheared in the way shown in fig. 4, Plate XIII. When the sheared ends were removed, the shear in cross-section

1 Only one end of the bar is shown.