91 



Earthquakes which are bracketed occurred within a few hours of each other, and, therefore, may possibly 

 refer to the same relief of seismic strain. In the lower line of totals each of the groups has been regarded 

 as a single disturbance. Whichever line we take, it seems that the greatest frequency has been in April, 

 May and June, or the first part of the winter months. The seasonal distribution of Antarctic earthquakes 

 is, therefore, similar to the distribution noticed in many other countries. Dr. OMORI, however, has shown 

 that earthquakes with a sub-oceanic origin off the coast of Japan have their greatest frequency in the 

 summer, during which season a higher average sea level more than counterbalances a diminution of load 

 on the sea bed, due to a lower barometric pressure. The seasonal difference in load amounts to 

 18'3 millims. of mercury. Whether similar conditions prevail in the Antarctic regions remains to be 

 determined. 



3. On the Form of Areas Disturbed by Large Earthquakes. (See Plate 3.) 



For local earthquakes, such, for example, as are from time to time noted in Great Britain, we are 

 prepared to see isoseists occasionally in the form of circles, but more frequently in the form of ellipses. 

 The major axis of any one of these ellipses is usually parallel to the strike of a' fault, the sudden yielding 

 on the face of which gave rise to the shaking. If the movement originates at no great depth, the epifocal 

 area where motion is most pronounced has been shown by Dr. CHARLES DAVIDSON to lie on the side of 

 the fault towards which it hades. 



With very large earthquakes, which are not sufficiently strong to be recorded over the whole surface of 

 the world, but which may reach stations near to their antipodes, the idea of elliptical isoseists requires 

 modification. 



For example, earthquakes originating in District M to the S.W. of New Zealand have been recorded 

 to the >S.E. by the " Discovery," and along a band at least 20 in width, extending in a N.W. direction aa 

 far as Britain. They may or may not be recorded in India, whilst at comparatively near places like 

 Batavia, Manila, and Japan, lying northwards from the origin, they have been seldom noted. Also it 

 may be added that they have not been noted at Cape Town, or at Cordova in Argentina, each about 80 

 distant, nor anywhere on the American continents. It would appear, therefore, that recordable earth- 

 quake motion originating in District M may be propagated as a band running in a N.W. direction as far 

 as its antipodes. When more stations have been established in South America, it may be found that the 

 motion proceeds to great distances in two directions round the world. This, however, is doubtful. 



Earthquakes originating off the West Coast of South America have been recorded by the "Discovery" 

 to the S.W., but the greatest length of recognisable wave-path is found towards the N.E., in which 

 direction they have been recorded in Western Europe and also near to their antipodes in Siberia. They 

 have not been recorded at stations we should expect them to affect were they propagated with equal 

 intensity in an opposite direction round the world. 



Disturbances with origins in Japan, the Philippines, and the East Indies have been recorded as far S. 



N 2 



