8 Art. 6.— T. Terada 



intruding between tlie areas of Class I, as if it were the branches 

 of the area for IL It may be noticed that the latter branched 

 areas coincide nearly with the tectonic depression of Bungo 

 Channel and a chain of depression formed by the Bay of Ise, Lake 

 Biwa and the Bay of Wakasa. 



Prof. Omori^^ has investigated, early in 1902, the seasonal 

 variation of seismic frequency in different stations of Japan and 

 distinguished two groups of districts according to the season in 

 which the frequency is maximum. Those districts in which it is 

 maximum in winter or in summer were called respectively Group 

 A and B. Comparing his map reproduced in fig. 8, with fig. 7, 

 a certain similarity of the distribution ma}^ easily be discerned. If 

 in fig. 7, the areas for Class II and II' be allowed to encroach upon 

 that for Class I in the western and central part of Japan, while it 

 be made to shrink back in northern Japan, the result may become 

 quite similar to fig. 8. If this comparison is legitimate, the con- 

 clusions follow: The districts for which the seismic frequency is 

 greater in summer correspond to those for which it is maximum 

 when the ' hemispherical difference ' of the solar activity |iV^—/S'| 

 is large, while those with the greater frequency in winter are those 

 for which it is maximum when |iV— aSI is small. If the yearly 

 seismic frequency be chiefly influenced by the distribution of 

 barometric pressure as it seems probable, we may therefore be 

 justified to conjecture that the years with small hemispherical 

 difference would probably correspond to those in which the con- 

 tinental pressure is relatively high and the oceanic one is relatively 

 low, while the years with large difference would be characterized 

 by relatively higher oceanic and lower continental pressure. 



Referring to Table IV, B of the previous paper, it may be 

 observed that the small |/z'— s'| is generally associated with the 

 gradients with the western side higher. This may be regarded as 

 a partial verification of the above conjecture. 



Next, the yearly mean values of barometric pressure for dif- 

 ferent oriental stations were taken from Walker's memoir^ and 



1) F. Omori, Publications of the Earthq. luv. Coium. in Foreign Language, Xo. 8. 



2) Gr. T. Walker : " Suuspots and Pressure," Memoirs of the Indian Met. Dep.,21, Part 12. 



