December 5, 1884.] 



SCIENCE, 



517 



the plane of the lever, save that due to the very 

 slight friction at b and c and at the marking-point ; 

 and the purpose of the heavy mass is chiefly to fur- 

 nish, by its inertia, the necessary fulcrum upon which 

 to overcome this slight friction. These principles, as 

 regards centre of percussion, and axis of instantane- 

 ous rotation relative to the axis of support, are espe- 

 cially insisted on by Professor Ewing as essential 

 features of a reliable seismometer. Two of these 

 horizontal pendulums at right angles record the two 



already obtained. Fig. 2 shows this record, made by a 

 pair of horizontal pendulums multiplying the motion 

 six times. As here reproduced, it is about twice the 

 actual motion of the ground. The inner circle gives 

 the N-S, and the outer the E-W components. At 

 d and d' are two cross marks, showing where the 

 pointers rested when the plate was stopped ; and 

 their angular distance is that to be used in connect- 

 ing the two circles to obtain the simultaneous motion 

 of the two points. The motion began on the outer, 



Fig. 2. 



rectangular components of the horizontal motion 

 upon a rotating smoked-glass plate. 



Many other ingenious and novel forms of appara- 

 tus for registering the horizontal motion of the earth 

 are described in chapters ii. and iii., and various de- 

 vices are described in chapter iv. for measuring the 

 vertical component of the motion. 



We reproduce the trace of the earthquake of 1881, 

 March 8. Fortunately, Professor Ewing was present 

 during its occurrence, and, as it drew to a close, lifted 

 the marking-levers from the plate, so that their sub- 

 sequent trace should not obliterate at all the record 



E-W, circle at a (corresponding to a' on the inner), 

 and thence can be easily traced nearly twice round 

 the circle to c, where the pointer was lifted (corre- 

 sponding to c on the inner circle). There was no 

 appreciable motion in a N-S direction till about ten 

 seconds after it began at a, E-W ; but it began quite 

 suddenly just before &', and can be traced twice round 

 to where the plate stopped at d'. The rotation time 

 was about eighty seconds, and the earthquake had 

 lasted about two minutes and a half when the plate 

 was stopped. Feeble movements were observed some 

 time longer. 



