Oct. 23, 1884] 



NA TURE 



611 



points near the revolving register cylinder k. A pile of 

 electro-magnets, /, each of which is in connection with 

 the two corresponding segments of the two contact circles 

 and the battery, when the current passes, draws one soft- 

 iron spring style towards its poles, and brings the point 

 in contact with the cylinder k. The pendulum might be 

 made to act directly on the contact circles without the 

 magnifying lever c /, if thought necessary, and the breadth 

 of the outer circle might be greater and give a correspond- 

 ingly longer dash so as to distinguish those derived from 

 the two circles. 



It will be seen that the most complex movements of the 

 pendulum or rolling disk can be accurately registered, 

 since four contacts will be marked in each semiphase of 

 the oscillation of the pendulum, and a fifth point can be 

 obtained by calculation from the instrument for register- 

 ing amplitude, which, together with the position of rest, 

 will give sufficient data to obtain the loops for each semi- 

 phase throughout the disturbance. 



In large earthquakes where the wave-amplitude is very 

 great, the rolling disk and balls would require to be of 

 very great size, which in many cases it might be imprac- 

 ticable to carry into execution, although the results might 

 be of great perfection in so doing. The present instru- 

 ment (Fig. 4) is intended to replace the rolling disk where 

 that cannot be used of sufficient size. 



A strong rectangular frame, c, carrying two strong up- 

 rights at its two extremities, is made so as to rotate with 

 great facility around the vertical axis /. It supports the 

 registering drum h with the clockwork arrangement/'and 

 a counterpoise,^. In practice it might be found advisable 

 to attach these beneath the frame and so lower the centre 

 of gravity of the whole apparatus. At a suitable distance 

 above the drum a crossbar, p, is attached, which should 

 be highly polished and square in section. The weight j 

 should be made very heavy, and be allowed a very easy 

 motion by means of four pairs of wheels, /', which are in 

 contact with the crossbar^ ; these might be mounted on 

 friction wheels (not shown in diagram). Attached on 

 both sides of the weight are the silk threads z'/,.which 

 traverse the upright, run over the pulley /", and are 

 attached to the weights o, which are only heavy enough 

 to draw the weight j back to its place when it has been 

 disturbed, that is to say, only just sufficient to overcome 

 the friction of the wheels k. These weights, 0, are tra- 

 versed by the guide wire n to prevent them dangling about 

 during the swinging round of the frame. 



The upper part of the frame carries two parallel bars, 

 d, between which is a narrow groove to allow of the sliding 

 of the plate c. They form the segment of a circle whose 

 radius is equal to the length of the entire pendulum and 

 its suspension. The pendulum a has fisced rigidly to its 

 inferior extremity the steel axle b, which passes through 

 the rectangular flat block c, which is prevented from 

 slipping off by a bolt-head below, so that the flat block 

 can rotate around the axle without falling off. 



The action of the apparatus is as follows : — When an 

 earth movement takes place the whole apparatus is brought 

 into the azimuth of the wave-path by the oscillations of 

 the pendulum a in that direction, which is affected by the 

 block c sliding in the groove between the bars d. The 

 pendulum should, in preference, have a short suspension, 

 so that the period of its oscillations should be less than 

 the wave intended to be registered by the apparatus, and 

 should possess sufficient weight to have complete com- 

 mand over the frame, keeping it always in the wave-path 

 azimuth. The weight will now appear to slide backwards 

 and forwards on the bar p, registering its movements by 

 the writing stylus, attached beneath it, on the drum //. 

 The moment that j is moved from its central position one 

 of the weights is raised from its position of rest (these 

 weights should preferably be hollow brass boxes into 

 which only sufficient fine shot could be poured to over- 

 come the friction of/), and rises as long as/ continues to 



roll along the bar ; if then the second half of the semiphase 

 is not sufficient to bring it back to its normal position this 

 will be done by o. When/ has reached its central position, 

 t> will have come to rest at the base of the guide wire, and 

 so no longer has any action, but is replaced in the second 

 semiphase by its fellow of the opposite side. Of course 

 the influence of the counterweights in retarding the 

 rolling mass must be experimentally tried and taken into 

 consideration in the calculations made from the tracings. 



The principle of this instrument is the acceleration and 

 retardation of a falling body during each semiphase of an 

 earth-wave. Fig. 5 illustrates a means of registering such 

 changes in the rate of a falling body so acted on, al- 

 though some other person better acquainted with mecha- 

 nical movements might possibly suggest some improve- 



Fig. 5. — Apparatus for registering the vertical component of earth-waves 

 during the whole of the disturbance. 



ment. A rigid vertical support of considerable length, a, is 

 attached to the side of a well and is connected at its upper 

 end to the table c ; and by a small bracket b two vertical 

 guide wires, n, pass through rings in the sides of the weight 

 g, so far resembling Morin's apparatus illustrating the 

 falling of bodies. Attached to the weight g is the silk 

 thread/, which turns once or twice around the wheel d, 

 and is supplied from the drum /-. The wheel d is con- 

 nected by the axle e and the continuous screw h to the 

 apparatus i, which is a skeleton of flat steel springs, 

 generally used to illustrate the distortion of a sphere by 

 centrifugal motion. (I have proposed this in preference to 

 the ball governors as retaining less impressed energy, 

 which would unnecessarily complicate and even modify 

 the results.) Sliding on the central axis of the spring 

 sphere is a small cylinder, /, which is prevented from 



