ox THE EARTHQUAKE AND VOLCANIC PHENOMENA OK JAPAN. ] il 



(2) Should there be a tremor storm extending over one or two days 

 tliere is a maximum motion at aljout 6 or 9 A.M. (figs. 9 and 13). 



(3) Large daily waves nearly always correspond -with pronounced 

 tremors (47 cases). 



(4) When no daily wave appears, or when it is feeble, which usually 

 happens when the weather is dull or wet, there have been eight cases of 

 feeble tremors, and nine cases where tremors have loeen practically 

 absent. 



(5) The greatest motion is experienced, or motion is most frequent, while 

 the pendulum is moving eastwai-ds — an observation which is connected 

 with the remarks in the next section. 



3. Relationship of Tremors to the Hours of Day and Night. 



(1) It has been shown that tremors are most frequent or at their 

 maximum at about 6 or 9 a.m. (figs. 9 and 13). 



(2) The hours during which storms are the most frequent are from 

 9 P.M. or midnight until mid day. During the afternoon and evening, 

 therefore, tremors are not so frequent. 



(3) The tremors at 6 or 9 a.m. may be attributed to irregular move- 

 ments accompanying the reversal in the inclination of the ground which 



Fig. 13. 



takes place at these hours, but if the tremors which occur at night are to 

 be attributed to an intermittent change in level, it then becomes neces- 

 sary to explain why the smallest changes in level are most intermittent in 

 their character, which supposition is improbable. The accompanyino- 

 figure shows how a tremor storm which may continue over several days 

 has a maximum at 6 or 9 a.m., and also, as it dies out, that it terminates 

 with slight tremoi's at these particular hours. It is taken from instru- 

 mejit A. 



