T. GBAY ON A SEISMOGEAPHIC APPAKATTJS. 221 



B, is hung, the other end of the box being fixed to a horizontal axis. 

 This box is partly filled with mercury for the purpose of compen- 

 sating by its negative stability the positive stability of the spring. 

 The mode of carrying the mass on the end of a lever is adopted in 

 these instruments because the period of free vibration can by this 

 means be made moderately long with a short spring, thus render- 

 ing the instrument more compact and easily compensated. To 

 the front of the box B, a very light vertical index is attached. This 

 index is made of a very thin tube of aluminium supported by angle 

 ties of silk thread so as to give stiffness. The threads are kept 

 tight by a spring which presses against the top of the tube. 



The three indexes of the instruments just described rest with their 

 outer ends against the surface of the drum D, covered with smoked 

 paper, the points being arranged in the same straight line parallel 

 to the axis of the drum. When the earth, to which the heavy cast- 

 iron sole plate and supporting pillars are rigidly attached, moves in 

 a direction at right angles to the line joining the knife-edge and the 

 centre of inertia of the masses P and M of the respective instru- 

 ments, they remain behind in virtue of their inertia, and thus the 

 indices are made to move across the drum in a direction at right 

 angles to that in which it is kept moving by the train of clockwork 

 W. The distance moved by the points of the indices, combined with 

 the known multiplication of the instrument, is sufficient to indicate 

 the direction of the motion, while the shape of the curves traced 

 on the paper and the distances between them indicate the nature and 

 period of the motion, the rate of motion of the drum being known. 

 The drum D is driven by a clockwork train of such length that a fall 

 of the driving weight of one foot is sufficient to keep the drum going 

 at the rate of one turn in two minutes for twenty-four hours. This 

 train of clockwork is actuated by two separate driving wheels, one 

 on each side of the first pinion, so as to avoid at the same time ex- 

 cessive pressure on the bearings of the driving wheels and any 

 great pressure on the bearings of the pinion wheel. It is governed 

 by a continuous-motion governer, which will be understood from 

 fig. 4, in which b is a cylindrical brass box mounted on a vertical 

 axis of hard steel running at its lower and upper ends in jew- 

 elled bearings. The box is connected to the clockwork through the 

 pinion p and the crown wheel w, A small quantity of liquid, I, is 

 poured into the box ; and this, when the box is made to revolve 

 rapidly, forms into a paraboloid of revolution. When the liquid 

 reaches a certain height dependent on the speed required, it comes 

 into contact with a fixed vane v, which checks the motion of the box 

 and keeps it going at a uniform rate. It is intended to keep the 

 clockwork continuously in motion, so that when an earthquake takes 

 place it is ready to receive the record. The duration of the shock 

 is then to be reckoned from the length of the record. 



The time of occurrence of the shock is determined by a modified 

 form of the apparatus described in the ' Philosophical Magazine * for 

 Nov. 1881, p. 363, and called a "time-taker." The earthquake 

 causes a very delicate circuit- closing apparatus (see ' Philosophical 



