66 s. SEKIÏA. 



The later and leebler parts of the record are eut out from Plate XI. 

 in order to reduce the side of tlie plate. 



Horizontal Motion. — The records begin at and from the fifth to 

 sixth second there suddenly appears one large undulation with the 

 period of 1.7 seconds in the East- West circle with comparatively short- 

 period waves on the other, and hence when these two are compounded 

 together it appears that the ground moved approximately 1.1 mm. 55° 

 W. and then 1.7 mm. N. 30° W. within the interval of 1.7 seconds. 

 This is the largest horizontal motion in this shock. The record on 

 the smaller plate may be traced over one and one-quarter revolution of 

 the plate, corresponding to 106 seconds of time. 



Vertical-Motion. — This begins at 0, and up to the end of the 

 seventh second there only but minute tremours of from .03 mm. to .09 

 mm. with a period of nearly 0.25 seconds. At the sixth second 

 when the largest horizontal displacement appears the vertical motion 

 is almost nothing, or at this point the soil still moved purely in a 

 horizontal plane. From the tenth to the fourteenth second, a series of 

 8 distinct shocks of nearly equal amplitude is recorded, giving an 

 average period 0.5 seconds. The greatest motion is 0.22 mm. at the 

 eleventh second. The vertical motion may be traced up to the forty- 

 third second with occasional rests between. 



If we combine the three components exhibited on the diagram at 

 the tenth second we will discover that, at this point, the ground was 

 describing a loop or an elliptical figure whose major axis (pointing 

 East and West) is 1 mm. and whose minor axis 0.4 mm. if we look 

 from above downward, but at the same time the particles of tlie earth 

 were actually performing two double oscillations of 0.2 mm. in the ver- 

 tical plane. In the next second the motion was more nearly East and 

 West with the usual rapid vertical oscillation. The composition of the 

 three components may be carried out further in similar ways, but always 



