﻿Vol. 6 1.] TWIN-EARTHQUAKES. 29 



valuable memoir on the Charleston earthquake, publishes two maps 

 of the isoseismal lines which traverse the meizoseismal area, one 

 representing the impressions of Mr. Earle Sloan, who surveyed the 

 district with great care, the other giving his own interpretation of 

 the same evidence. In neither case do the isoseismal lines cor- 

 respond to the degrees of any definite scale of seismic intensity ; but, 

 although differing widely in detail, they agree in their expansion 

 around two small districts which probably represent the epicentres 

 corresponding to the twin-foci. In the Neapolitan earthquake, the 

 distribution of the seismic death-rate gives a fair approximation to 

 the positions of the epicentres. The south-eastern epicentre must 

 be close to Montemurro and Saponaro, where the death-rates were 

 respectively 71 and 50 per cent. : while the north-western epicentre 

 must be near Polla, where the death-rate was about 30 per cent., 

 Polla being 4 miles from the single epicentre ascertained by Mallet 

 from observations on the direction of the shock. 



More satisfactory materials for the purpose are provided by 

 the Colchester earthquake of 1884. In their well-known report, 

 Prof. Meldola & Mr. White note the existence of two areas of 

 maximum destruction, one near Peldon and Abberton, the other 

 near Wivenhoe and Rowhedge. They also give a table of the places 

 where damage occurred, with the population of each in 1881 and 

 the number of buildings repaired. Assuming that there are five 

 persons on an average in every house, the percentage of houses 

 needing repair may be calculated for each place, and from these 

 may be drawn curves of equal percentages of such houses. In 

 fig. 7 (p. 30) the continuous line represents the isoseismal 8, while the 

 dotted lines are the curves corresponding to percentages of 60 and 

 30 respectively. As the Colchester earthquake is known from other 

 evidence to have been a twin-earthquake, it is probable that the 

 epicentres are closely surrounded, if not outlined, by the two inner 

 curves of percentage 60. 



If the innermost isoseismal is excentric, the epicentre correspond- 

 ing to the stronger impulse must be situated close to its centre, 

 while the other epicentre probably occupies a similar position witli 

 regard to the centre of the next isoseismal, though perhaps slightly 

 farther removed from it. In this way are determined approximately 

 the positions of the epicentres of the Leicester and Pembroke earth- 

 quakes of 1893, the Hereford earthquake of 1896, the Carlisle 

 earthquake of 1901, and the Derby earthquake of 1904. In the 

 Pembroke earthquake of 1892, the epicentre corresponding to the 

 weaker impulse was probably submarine ; while, for the Derby 

 earthquake of 1903, we have to rely upon the form of the inner 

 isoseismals. 



The following estimates of the distance between the twin- 

 epicentres are therefore only approximate ; those for British earth- 

 quakes may err by as much as a mile, the others by even more. 

 The highest estimate is that of about 35 miles for the Bivieran 

 earthquake of 1887 ; but, in this case, there is some uncertainty 

 as to the position of the secondary epicentre near Nice. In the 



