Art. 6.— T. Terada : 



The latter data were taken from Maunder' s paper in the Monthly 

 Notices of the Royal Astronomical Society, Vol. 80, 1920, in which 

 the areas for the both hemispheres are separately tabulated. The 

 seismological data were kindly placed at my disposal by Prof. 

 Okada and Prof. Saem. Nakamura of the Central Meteorological 

 Observatory, for which my best thanks are due. The materials 

 consist of the records of the number of earthquakes observed in 

 the chief meteorological observatories of Japan'\ The shocks are 

 generally classified into two kinds, i. e. those ' felt ' and instrument- 

 ally recorded. The ' total ' number, therefore, shows often a sud- 

 den increase after the introduction of the seismograph. Beside 

 this apparent circumstance, it must be assumed that the said num- 

 ber is afïected by many uncontrollable causes apart from the actual 

 variation of the seismic activity. Hence, it was considered most 

 plausible to take the following procedure for studying the actual 

 influences of the secondary cause upon earthquakes. 



The successive yearly numbers were plotted as the ordinates 

 on a diagram with the years as abscissse and the successive points 

 were connected into a broken line or curve. Among the other ir- 

 regularities presented by the diagram, one may worth special men- 

 tion, i.e. when a destructive earthquake occurs in a district, the 

 aftershocks are felt often duiing several years after it. In such 

 case, the small variation of frequency due to the secondary cause 

 may be entirely overlooked, if we pay our attention onl}^ to the 

 absolute magnitude of the number. Hence, for the present pur- 

 pose, the maxima and minima of the curve were determined by 

 comparing each year only with the few neighbouring years. Any 

 year at the top of a convex part, or at the bottom of a concave 

 part of the graph was respectively called a ' maximum ' or a 

 'minimum'. On the slope of frequency curve following a de- 

 structive earthquake, it may therefore happen that a ' minimum ' 

 is followed by a ' maximum ' which is decidedly less in actual 

 magnitude than the former. In short, the maxima and minima 

 were judged by the second difference of the frequencies. 



1) Number of the years available varies from 18 to 30 for different stations. For sixteen 

 stations, the observations of thirty years are at hand. 



