66 



REPORTS ON THE STATE OF SCIENCE. 



Fig. 1. 

 After-shocks of the Jamaica Earthquake, January 11, 1907. 



« 

 /3 

 iz 

 II 

 10 



I' 



14 I 5 JAN. 19 



e 



13 



18 



Z4 29 3 FEB. 



I = After-shocks in Jamaica. 

 I[ = After-shocks which reached Great Britain. 

 Ill = Mean time intervals between after-shocks in Jamaica. 



The Time hitervals between After-shocks. 



As we recede from the primary shock the intervals between the after- 

 shocks ought to become greater and greater, the increase in time interval 

 indicating the nearness of approach to complete settlement. But in 

 taking means of the intervals, as pointed out by Mr. O'Neill, a grave 

 error may creep in by reckoning the means over arbitrary periods. In 

 this way, to reckon the mean interval between the shocks for each day, 

 three dajs, or each week, may or may not give any satisfactory result. It is 

 mere chance in either case. But if we take tlie intervals over a sufficiently 

 long period, and take means at points where the intervals have appreciably 

 increased, we stand to obtain a result which more nearly approaches a true 

 statement of the case. Thus means were first taken from day to day 

 between January 14 and February 17, and the daily means worked out 

 as follows : Ih. 19m., 2h. 31m., 5h. 20m., 5h. Im., 3h. 11m., 2h. 41m., 

 4h. 48m., 6h. 21m., 3h. 38m., 5h. Om., llh. 51m., 19h. 30m., lOh. 52m., 

 18h. 40m., 14h. 10m., 6h. 38m., 18h. 18m., 28h. Om., 23h. 55m., 24h. 10m., 

 13h. 57m., 7h. 10m., 8h. 28m., 35h. 8m., Gh. 5m., 82h. 40m. Now it will 

 be noticed that there are appreciable increases on the ITtb, when the 

 mean becomes about 5 hours. Another occurs on the 21st, when the 

 interval has reached bh. 21m. On the 25th it has become 19h. 30m., on 

 the 31st, 28h. Om., on February 6, 35h. 8m., and on the 11th, 82h. 40m. 



If, then, we take means at these points we obtain results as follows, 

 which show a constant increase in the time intervals as the distance from 

 the primary shock increases. 



Between January 14 and January 17, for 24 intervals, the mean is 3h. lOra. 

 17 „ 21, „ 27 „ „ ah. 48m. 



7h. 21ni. 



12h. 3m. 



14h. 4fim. 



„ 4Cb. 52m. 



The right-hand curve in fig. 1 shows the rate at which the intervals 

 between after-shocks increased and seismic stability was approached. This 

 curve, which should be the inverse of the upper curve in the left-hand side 

 of the figure, is now shown, I believe, for the first time. 



