62 REPORTS ON THE STATE OF SCIENCE.—1913. 
shown that great megaseismic activity is followed by long periods 
of rest. In that case the intensity of a group was considered inde- 
pendently of the number of days over which it was spread. 
XI. Frequency of Earthquake Followers. 
In the British Association Reports, 1899, p. 227, and 1900, p. 71, 
under the title of ‘Earthquake Echoes,’ I discussed the vibrations 
which follow the main shock or shocks of an earthquake and which 
bring the same to a conclusion. These occur in groups, and as these 
rise and fall in amplitude it may be inferred that an earthquake does 
not become extinguished at a uniform rate, but it dies in surges. 
I sought for the origin of these surgings, particularly for those groups 
which resemble each other in form, in the hypothesis of repeated 
reflection. 
Another possible explanation is to assume that these repetitions 
are interference phenomena consequent on the difference in period 
between the free swing of the recording pendulum and that of the 
earth. 
Now steady-point seismographs have shown for earthquakes we can - 
feel that their period increases as they die out at a given station, 
and that it also increases as they radiate from an origin. Assuming 
this to be correct for megaseismic motion, then beats or recurrences 
at stations at different distances from an origin should show differences 
in their frequency. To test this I have compared the time frequency 
of pulsatory recurrences for disturbances recorded in the Isle of Wight 
which originated in different localities. 
The localities chosen were as follow :— 
1. East Coast of Japan, distant from Shide . ‘ . -1 80? tte’) 85°. 
2. West Coast Central America, distant from Shide . 80° to 885°. 
3. Central Asia, distant from Shide . F 3 {: ORE GO? 
4. Between East New Guinea and Fiji, distant from Shide 130° to 140°. 
The earthquakes originating in these four districts, and of which 
I have seismograms recorded in the Isle of Wight, are referred to 
by numbers corresponding to the numbers given in the Earthquake 
Catalogue published in the British Association Reports, 1911, p. 57, 
and 1912, p. 71. 
For District No.1 these numbers were 263, 397, 405, 425, 431, 
446, 448, 450, 457, 483, 493, 514, 884, 1031, 1266, 1427, and 1510. 
The average time interval between successive groups was found to be 
2°8 minutes. 
For District No. 2 the numbers were 248, 264, 407, 415, 417, 
432, 447, 536, 576, 606a, 642, 806, 924b, 1164, 1450. The average time 
interval for these disturbances was 3'4 minutes. 
For District No. 3 the numbers were 542, 558, 626, 644, 662, 
663, 832, 886, 982, 1064, 1070, 1293, 1468. The average time interval 
for these disturbances was 2°7 minutes. - 
For District No. 4 the numbers were 351, 352, 354, 377, 435,- 
515, 530, 581, 977, 1025, 1128, 1190, 1272, 1301, and 1460. The 
average time interval for these disturbances was 28 minutes. 
If we compare the time intervals for Districts 1, 8, and 4, it would 
