AND SEMI-ANNUAL SEISMIC PERIODS. 
1115 
in §§ 4 and 6 are given in Table II. ; and in Table III. the reduced six half-monthly 
means, or three-monthly means, obtained in investigating the semi-annual period. 
At the end of the section relating to each record examined, are given the maximum 
epoch and amplitude of each period. The epoch mentioned is the end.of the month 
or half-month indicated, the letters a and h denoting the first and second halves of 
the month. The date of the maximum and the amplitude have been taken as those 
of the greatest mean. It would not have been difficult to have given the result more 
- minutely, but it is doubtful whether the appearance of greater accuracy would have 
any value, considering the unavoidable imperfections of most seismic records. 
In the same place, I have also given the date of the maximum epoch of the annual 
barometric period over the greater part of the district considered, excluding high-level 
stations. For this purpose I have made use of Dr. Buchan’s great work on 
Atmospheric Circulation.* The method of investigation is the same as that employed 
for the annual seismic period, except that its application is much simpler in this case. 
An asterisk placed before the name of a record, either in the Tables or detailed , 
discussion, indicates that the results obtained are probably more reliable than in other 
cases. The series of six-monthly and six half-monthly means calculated from these 
records are illustrated by the curves in figs. 5-15, the continuous line representing 
the annual, and the dotted line the semi-annual, period. 
The Beality of the Annual Period. 
13. It is well known that many more earthquakes are recorded as felt during the 
night than during the day, and that this is due partly to the stillness of the night, 
but more, perhaps, to the position of the body during repose being more favourable 
for the detection of slight tremors. Now, the maximum of the annual period occurring 
during winter, it may seem possible that the periodicity may be apparent only, and 
due to the greater number of shocks observed during the longer nights of winter. 
The question is one that deserves examination, and the following facts seem to me 
conclusive in establishing the reality of the annual period. 
In the case of the earthquakes recorded at Tokio by the Gray-Milne seismograph, 
there can, of course, be no doubt. But this is the only case treated in this paper in 
which the shocks are registered entkely by instrumental means. 
The hours of rest may be taken roughly at between 10 p.m. and 8 a.m., and we may 
therefore divide the day into two parts, reckoning day-time from 9 a.m. to 9 p.m., and 
night-time from 9 p.m. to 9 a.m. Using Fuchs ’ valuable ‘ Statistik der Erdbeben,’ 
I have counted the number of shocks, whose time of occurrence is recorded, in Austria, 
* “Report on Atmospheric Circulation, based on the Observations made on board H.M.S. Challenger 
during the years 1873-1876, and other Meteorological Observations.” ‘Report of the Scientific 
Results of the Voyage of H.M.S. Challenger, &c.; Physics and Chemistry,’ vol. 2, 1889, pp, 1-263. 
7 C 2 
