320 EARTHQUAKES. 



at its end there is usually an earthquake. Usually these 

 storms are closely related to barometric depressions. To 

 distinguish these movements from those which occur 

 under high pressure, the latter are called haro seismic 

 movements, and the former vulcano seismic movements. 

 The relation of these storms to barometric fluctuation has 

 been observed to have been very marked during the time 

 of a volcanic eruption. 



At the commencement of a storm the motions are 

 usually small, and one storm, lasting two or three days, 

 may be joined to another storm. In such a case the 

 action may be a local one. It has been observed that a 

 barometrical depression tended to bring a storm to a 

 maximum, whilst an increase of pressure would cause it 

 to disappear. Sometimes these actions are purely local, 

 but at other times they may affect a considerable tract of 

 land. 



If a number of pendulums of different length are 

 observed at the same place, there is a general similarity 

 in their movements, but it is also evident that the free 

 period of the pendulum more or less disturbs the character 

 of the record. The greatest amplitude of motion in a set 

 of pendulums is not reached simultaneously by all the 

 pendulums, and at every disturbance the movement of 

 one will predominate. From this Eossi argues that the 

 character of the microseismical motions is not constant. 

 Bertelli observed that the direction of oscillatioivof the 

 pendulums is different at different places, but each place 

 will have its particular direction dependent upon the 

 direction of valleys and chains of mountains in the neigh- 

 bourhood. Rossi shows that the directions of movement 

 are perpendicular to the direction of lines of faults, the 

 lips of these fractures rising and falling, and producing 

 two sets of waves, one set parallel to the line of fracture, 



