142 



REPORT — 1895. 



4. jRc'IationsMp of Tremors to Barometrical Conditions. 



(1) Tremors are apparently as mai'ked with a high barometer as with 

 a low barometer. 



(2) Tremors chiefly occur with steep barometric gradients. 



(3) They are marked when barometric changes are rapid, whether the 

 barometer is rising or whether it is falling. 



(4) Although it is not likely that the daily fluctuation of the barometer 

 should have any marked effect upon the production of tremors, it may be 

 noted that the maximum of tremors occurs when the barometer is at its 

 greatest height and about to fall rapidly to its daily mimimum. Mr. T. 

 Wada, of the Meteorological Observatory, has kindly given me the follow- 

 ing table showing the daily barometrical maxima and minima deduced from 

 eight years' observations : — 



For the year 760-00-758 30= 1-70 mm., at 9 a.m. 



5. Relationship of Tremors to Wind, Temperature and suh-Surface 



Condensation. 



Although tremors have occurred when a heavy wind was blowing in 

 Tokio, as, for example, on October 26, tremors have been marked when 

 wind was practically absent. Neither does there appear to have been 

 any marked connection between the occurrence of tremors in Tokio and 

 the wind at Choshi, which is situated on the coast about 50 miles west 

 from Tokio. 



Although the morning breeze is apparently stronger than that in the 

 evening, it does not seem to be connected with the morning frequency of 

 tremors. 



Tremors are most frequent at the hours when the temperature is 

 lowest, or during the time that sub-surface condensation is taking place. 

 Tremors are at a mimimum during the time that the ground is becoming 

 heated, and there is a free flow of moisture in the form of vapour from the 

 earth to the atmosphere. 



Tremors and Waves on the Coast. — At the various lighthouses round 

 the coast of Japan at 2, 6, and 10 A.M. and p.m., records are made of the 

 force of the waves. In these records = calm and 6 = waves, which are 

 unusually large. I have compared the records from Jogashima, 33 miles 

 south of Tokio, and Inuboye, 53 miles west of Tokio, with the records 

 of tremors, but I do not observe any connection between them (see 

 Tables). 



Occurrence of Tremors in Manila. — A set of diagrams which clearly 

 show the relationship between various atmospheric phenomena and the 

 occurrence of tremors are the monthly sheets of the Meteorological Obser- 

 vatory in Manila, where from some date prior to 1883 observations have 

 been made with Bertelli's tromometer. Tremor storms are apparently 



