each of these a smooth round rod of wood (an uncut. pencil will do well), and make them, by greasing, &c., slide freely, but with slight friction through the holes.

Secure the hoop horizontally at the level of the centre of the ball by struts from the block, and the ball being in the middle of the hoop, slide in the four rods through the hoop until just in contact with the ball.

It is now obvious that a shock, causing the ball to oscillate in any direction, will move one or more of the rods through the holes in the hoop, and that they will remain to mark the amount of oscillation.

A similar apparatus, with the pendulum-rod secured horizontally (wedged into the face of a stout low wall, for example), will give the vertical element of the wave. Two of these should be arranged, one north and south: the other east and west. The objection to this and all apparatus upon the same principle is, that as the centre of elastic effort of the pendulum rod never can be had perfectly in the plane passing through the centre of gravity of the ball, for every possible plane of vibration, so an impulse in a single plane produces a conical vibration of the pendulum, and hence the ball deranges the position, more or less, of the index rods out of the true direction of shock. Moving the apparatus by hand, and a little practice in observation of its action, will, however, soon enable a pretty accurate conclusion as to the true line of shock to be deduced from it.

It will be manifest that the observer must record minutely the dimensions and other conditions of such apparatus, where it is not permanently kept, to enable calculations of scientific value as to the wave to be made from his observations of the range of either fluid or solid pendulums.

A common bowl partly filled with a viscid fluid, such as molasses, which, on being thrown by oscillation up the side of the bowl, shall leave a trace of the outline of its surface, has been often proposed as a Seismometer. This method has many objections: it can only give a rude approximation to the direction of the horizontal element; but as it is easily used, should never be neglected as a check on other instruments. A common wooden tub, with the sides rubbed with dry chalk and then carefully half filled with water or dye stuff, would probably be the best modification.

Another extemporaneous instrument for measurement of vertical motion in the wave may be sometimes useful. Make a spiral spring of eighteen inches or so in length by twisting an iron wire of one-eighth of an nch diameter round a rod of about 1 inch diameter (the staff of a boarding-pike); suspend it by one end vertically from a fixed point, and fix a weight (a twelvepound shot will do) to the lower end, and below and in a line passing vertically through the centre of gravity of the weight fix the stem of a common tobacco-pipe; let the lower end of this stem just dip into a deep cup filled with pretty thick common ink or other coloured fluid: the action of this needs no description.

The preceding instruments suffice at once to giv the direction of transit of the earth-wave and its dimensions; its rate of progress or transit over the shaken country remains to be observed.

Several distant observers, with chronometers, will of course best observe this, but such observations cannot be very numerous or extend over a large tract of country; yet it is most desirable that a network of such observing points should be stretched over the

shaken country. For this purpose common houseclocks, situated at several distant points, may be easily arranged, so that the pendulum shall be brought to rest and the clock stopped at the moment that the shock passes.

Fig. 5.

Fig. 5 shows part of the case and pendulum of a common clock. To fit it for this purpose bore two holes of a quarter of an inch diameter, one through either side of the clock-case, at a b, at the level of the lowest point of the pendulum-bob and in the plane of its vibration; round off the edges of these holes, and grease them.

In the centre of a piece of fishing-line or stretched whipcord, make a loop and pass it round the screw or other lower projection of the pendulum-bob; pass the two free ends of the cord out, one through each of the holes in the sides of the clock-case; provide a squared log of heavy wood of about five or six inches thick each way, and from four to five feet in height; cut

both ends off square, and stand the log upright on one end directly opposite the dial of the clock.

Measure off equal lengths of the cord at each side of the pendulum, and make fast their extremities to the two opposite sides of the upright log, c d, close to the top; bring the log backwards from the clock now, until the pendulum being at rest, both cords are drawn tight; and then advance the log two or three inches towards the clock, so that the cords may be slacked down into a festoon or bend at each side of the pendulum, and within the clock-case, so that the pendulum may have room to swing freely; and very slightly wedge the cord to keep it so, through the holes in the clock-case, and from the outside; see that the log rests firmly and upright upon a firm floor; and now set the clock a-going. The length of the cords, or the distance of the log from the clock in relation to its height, must be such that if it fall towards the clock it shall bring the cords up tight before the upper part of the log touches the ground. It is now obvious that in whatever direction the log may fall, it will arrest the motion of the pendulum and stop the clock within less than a second of the true time of transit of the wave at the spot.

If the adjustments are similar for all the clocks this error will be constant for them all; and if the true time be noted at the principal station it can be got for the rest.

Clocks with seconds pendulums only should be chosen for this use. They should be all set by one chronometer, and their errors afterwards taken.

Where convenient, the pendulums should be all placed to swing north and south, or east and west; and

in this case the sides of the logs will face the cardinal points, and the directions of their fall (where not entangled) be a rude index of that of the wave. It will be also desirable to place a bowl of fluid to mark direction with each clock.

The positions chosen for the clocks must vary with circumstances, but they should, as far as possible, surround the principal station; their distances apart must be considerable, as the speed of the wave or shock is immense probably five miles is the ordinary minimum, and thirty to fifty miles a convenient maximum distance. Such arrangements should be made as rapidly as possible after the first shock has given the expectation of others in succession.

When practicable, the following method of fitting common clocks may be adopted advantageously.-Let

a, Fig. 6, be the pendulum-bob; fix a pin of stou wire into a hole in the centre of it, b, at right angles to the plane of vibration; cut two small mortices through the sides of the clock-case, so that a lath of deal or other light wood, of about an inch and a half