264 T. Terada and T. ]Matuzaa\'A 



the change of sea-level at different mareographical stations and came 

 to the conclusion that the combined effect of the barometric and tidal 

 pressure is to be considered as an important secondary cause of earth- 

 quake. T. Terada^'^ investigated the relation between the annual 

 variation of barometric gradient and that of the frequency of earth- 

 quakes in different districts and found that the latter shows a paral- 

 lelism with the component of the gradient in a certain direction 

 proper to each district. K. Hasegawa^-^ studied the same relation with 

 respect to fifty local earthquakes in the Giliu district, of which the 

 locations of the oi'igiiis were ascertained. He confirmed that the 

 gradient 30™ before the occurrence of the individual shock is most 

 frequent in the direction NLO. or S. which differs distinctly from the 

 gradient most prevalent in that district. These directions were 

 respectively perpendicular to the conspicuous fault lines previously 

 located in this district. Neither the barometric height relative to the 

 surrounding regions nor the tinu^-rate of variation of the pressure 

 showed any sensible influence. The subject has since been investigated 

 by Saem. Nakamura^^^, S. Masuzawa^^^, D. Nukiyaraa^^^ etc. Saeni. 

 Nakamura showed, with respect to earthquakes in Wakayama Pre- 

 iecture, that the effect of the barometric gradient reveals somewhat 

 different aspects for annual and diurnal variations. Masuzawa found 

 that the gradient most probable on the occasion of earthquakes 

 originating in the Pacific submarine zone, between Idu and Tyosi, 

 is in opposite directions ibr both sides of a line crossing the land 

 transversely. D. Nukiyama and M. Mukai verified Masuzawa's result 

 and, moreover, showed that the time variation of the gradient vector 

 on the occasion of an earthquake has also a predominant direction 

 transverse to the axis of the land. Saem. Nakamura^"^ studied again 

 the effects of barometric and tidal i)ressure on the frequency of after- 

 shocks of the recent Kwanto Earthquake. His result seems to confirm 

 Masuzawa's, as far as the effect of the barometric gradient is con- 

 cerned. Besides, he noticed peculiar relations regarding the anomaly 



82. 



(1) Journ, Met. Soc, 28 (1909), 1 ; T.S.B.K., 4 (1908), 454. 



(2) J.M.S., 32 [inS; 397; T.S.I5.K., 7 (1913), 181. 



(3) J. M.S., 34 (1915), 71; 41 (1922), 420; T.S.BK., 8 (1915), 69; [iii] 2 (1920), 



(4) T.S.B.K., [iii] 1 (1919J, 343. 



(5) Jap. Journ. Astr. Geophys., 1 (1922). 49. 



(6) Jap. Journ. Astr. Geophys., 3 (1925), 115. 



