SEISMOLOGICAL INVESTIGATIONS. 



217 



completeness of the later records, or rather imperfection in the earlier years. 

 The following table gives the analysis in periods of 284 years : — 



Table I. 

 Numbers of C'htnese Destructive Earthquakes [Drake). 



The totals in the last column but one are governed chiefly by the later 

 cycles. To minimise this effect the eight columns were all reduced to the 

 same total 66, using one place of decimals until the sums were formed. The 

 results are given under the heading ' Revised,' and it will be seen that they 

 give substantially the same curve, with pronounced minimum extending from 

 the 48th year to the 118th, and a pronounced maximum at the end. The 48th 

 year of the present cycle will be 1942, so that we are approaching the time of 

 minimum quakes and have passed the maximimi. But it is not yet clear whether 

 these figures for China apply unmodified to the whole earth. It may be possible 

 to observe this decline in the near future, but up to the present the records are 

 affected by so many uncertainties, owing partly to the novelty of the science, 

 partly to the war, and to other causes, that it is very difficult to compare one 

 year with another. Thus the Eskdalemuir records show the foDowing total 

 numbers of earthquakes : — 



which at first sight might be interpreted as a notable falling-off in earthquake 

 activity, but is probably chiefly due to a change of method in 1915. The point 

 will, however, be further examined. Analysing the last two columns of Table 1. 

 harmonically we get from the simple totals 



20 cos [6 - 301°) -f 11 cos (2fl - 348°) + 1 cos (39 - 304°) -f 5 cos (4« - 98°) 



from the revised 



14 cos (9 - 304°) + 8 cos (29 - 340°) + 3 cos (39 - 211°) -f 8 cos (49 - 138°). 



The third harmonic is small — smaller than the fourth, for instance. But on 

 analysing the results in 101 years a larger term is obtained. The totals are 

 (for twelve groups to the cycle, which gives nearly the same mean as above) 



48 41 45 41 31 53 38 37 33 66 43 54 

 which gives a term 5 cos (9 - 331°). 



This is in accordance with the results found from trees — that the lOlyear term 

 should exceed the 94 year. 



