392 Dr. C. Davison—Founders of Sersmology— 
the year 1000 the most destructive of all earthquakes can hardly 
have escaped record. They were events of historical importance. 
Grouping them in periods of fifty years, from a.p. 1000 to 1850, he 
shows that, from about the year 1650 there has been a rapid increase 
in frequency, the numbers in successive half-centuries from 1600 
being 3, 10, 15, 17, and 30. He suggests that these figures indicate 
a real increase in seismic activity from about 1650, accompanied, 
as he showed later, by a corresponding increase in volcanic activity. 
The rest of Milne’s work on the variations of earthquake-frequency 
must be noticed somewhat slightly, as, owing to the brevity of his 
record of world-shaking earthquakes, the results must be regarded 
as suggestive rather than as established. Like Mallet, he was 
attracted by the alternations of seismic activity and repose, and he 
showed that earthquakes have a tendency to occur in groups, the 
number in a group usually varying from two to fifteen or more, 
the groups lasting from one to three days, seldom more than six 
days. The interval between the centres of successive groups varies 
from fifteen to fifty days, being roughly proportional to the intensity 
of the earlier group. 
More remarkable still is the tendency of strong earthquakes to 
occur in pairs, and even in triplets, in districts so widely separated 
as, for instance, Guatemala and the Indian Ocean (19th April, 1902), 
the interval between them being in this case less than the time 
required for the first preliminary tremors to cross the distance 
between the foci. Milne records fifteen cases of great double earth- 
quakes, and two of triple earthquakes, during the years 1899-1906. 
Such inquiries lead naturally to the suggestion that the variations 
of frequency in widely separated districts may to a certain extent 
be synchronous. The materials are perhaps at present insufficient 
for such wide generalizations, but Milne gives reasons for supposing 
that, in two districts so far apart as the eastern and western margins 
of the Pacific, the frequency in one increases and decreases 
simultaneously with that in the other; and that during the last 
two centuries, Italy, Japan, China, and America have roughly 
agreed in their periods of activity and repose. 
The last point to which I can refer is Milne’s interesting suggestion 
that the frequency of great earthquakes may be connected with the 
migrations of the pole. The earthquakes of the years 1892-1904 
were grouped in successive intervals of 364 days each. Counting 
the earthquakes in those intervals in which changes of direction took 
place in the movements of the pole, and in the preceding and 
following intervals, Milne found the numbers to be. respectively 
287, 167, and 217, and he thus concludes that great earthquakes 
are frequent about the times when changes occur in the direction 
of the polar movements, and especially when those changes occur 
most rapidly.’ 
1 Rep., 1900, pp. 107-108 ; 1903, pp. 78-80; 1906, pp. 97-9; 1908, pp. 78-9; 
1909, pp. 56-8; 1910, pp. 54-5; 1911, pp. 32-5, 36-8; 1912, pp. 92-4 
