ON THE VOLCANIC PnENOMENA OF JAPAN. 



433 



been, recorded -when the temperature has been above the average as 

 60 : 40. 



The mimber of earthquakes which have occurred each month when 

 the temperature has been rising, falling, or steady are given in the 



following table : — 



Earthquakes have therefore beea more numerous when the tempera- 

 ture is falling. To connect these observations with those relating to 

 barometric pressure, it may be well to remember that generally a low 

 temperature accompanies a high barometric pressure, and a high tem- 

 jjeratnre a low barometric pressure. 



These latter results respecting earthquakes aud fluctuations of 

 atmospheric pressure and temperature are similar to results obtained by 

 the analyses of more meagre data which the Meteorological Department 

 very kindly placed at my disposal some years ago. As I did not consider 

 that the results were of sufficient importance for publication, until now 

 no special reference has been made to them. The belief that there is an 

 immediate connection between earthquakes and atmospheric conditions 

 is, however, very general, and it is for the purpose of showing how little 

 foundation there is for such a belief that I have been led to incorporate 

 the previous analyses in the present report. The facts which have been 

 recorded may also be of value to investigators who wish to carry out 

 analyses on other lines. 



Eaeth-tremoes. 



In the report to the British Association for 1887 T gave an account 

 of observations made with an automatic tromometer of tremors which 

 had been observed between January 13, 1885, and May 14, 1886. Details 

 of these observations may be found in ' Trans. Seis. Soc' vol. xi. p. 1-78. 

 The records referred to in the present note were obtained between 

 December 22, 1886, and February 1888. 



Results of Analysis. — From a general inspection of the tri-daily 



■ weather maps, it is quite clear that when little or no wind is indi- 



.' cated, or when the isobars are few, no tremors have been recorded ; 



I while, on the contrary, when the wind is strong at many stations 



I in Central Japan, and when the isobars occur in close proximity, 



; tremors are almost always recorded. On the Japanese maps the 



« isobars are drawn at intervals of five millimetres of pressure. On the 



Italian maps, when the intervals are only one millimetre, the relationship 



between tremors and the frequency of isobars, which, when they are 



numerous, indicates a steep gradient, is even more marked than it appears 



to be in Japan. On the Italian maps, which are published under the 



direction of Professor M. S. de Ro'ssi, the state of the wind is not 



indicated, but it may be inferred that when the gradients over the 



Italian peninsula are steep, wind is blowing somewhere in the peninsula, 



1888, p P 



