ON THE AFTER-SHOCKS OF EARTHQUAKES. 137 



district, Sicily, and in the Balkan Peninsula and neighl)ouring 

 Islands.^ The curves for the Vesuvian district and Sicily shew each 

 a series of periods of about 5 years; and tliat for Switzerland shews one 

 well-defined of 12 years besides some ill-defined fluctuations of shorter 

 averaofe lens'th. It will be observed that the maxima and minima of 

 seismic activity for the two Italian districts occurred simultaneously, 

 but at quite different epochs from those for Switzerland. 



§ oO. To investigate the relations, if any, between earthquakes 

 and the phases of the moon, sun-sp(jts, tem[)erature of the atmosphere, 

 etc., seems not likely to lead to valuable results and w^ould be, as 

 Mallet remarked, a ^vaste of scientific time and labour. Willi atmos- 

 pheric changes of pressure it may be different and I sliall here, there- 

 fore, treat shortly of the possible connection of the barometric height 

 -with the frequency of after-shocks. 



In Table IV are li'iven the mean barometric heiijhts and the fluctua- 

 tions during successive days from October 28th, 1891, to A[)ril 3üth, 

 1892. It seems that earthquakes happen equally often with low as 

 with high pressiu'es. Thus from an examination of the record of tlic 

 daily seismic frequency at Gifu, I can count, between the above dates, 

 55 maxima and 55 minima in the frequency, and the means of the 

 barometric heights corresponding to these two sets are respectively 

 7()2.6J: and 763.3-1 mm., which are »practically identical. 



Ao'ain, bio* barometric falls of 10 or 20 mm. or rises of 5 or 10 mm. 

 were not accompanied by any marked change in the seismic frequency. 



A single abrupt change in the atmospheric pressure is not likely 

 to be accompanied by any fluctuation in the frequency of earthquakes. 

 If, however, barometric changes, whether small or great, (jccur at 

 regular intervals, then the earth's crust may finally assume certain 

 corresponding oscillations. Thus the dail}'' and annual fluctuations in 



1 The data are taken from Fuchs' " Statistik der Erdbeben von 1865-1885 " ; see Table XXI. 



