Vol. 52.] SEISMIC PHENOMENA IN THE BRITISH EMPIRE. 653 



Thus is one enabled to define the seismicity of a region by the 

 inverse of the square root of an area in which on an average one 

 shock occurs every year. Let us suppose that during a period of n 

 years p shocks have been recorded in a region of supe rficie s A, then its 



eismicity S will be expressed by the inverse of a /■£ — In practice, 



we make use of this numerical expression, so that seismicity or insta- 

 bility is the greater the less the number of square miles. The map 

 of the region just dealt with may then be covered with equidistant 

 lines arranged according to the scale thus found. The surface, 

 neglecting a slight error due to irregularity of contours, is then 

 divided into small primary squares wherein one earthquake occurs 

 in each year. As a rule, the scale of maps is not such as to allow 

 of this extremely rational equidistance of divisions. 



The data of seismicity thus calculated will be the more exact the 

 greater the number of years over which observations have extended ; 

 the data will vary also according to the nature of the observations upon 

 which they are based — whether historical, seismological, or seismo- 

 graphical, as defined above. In order to base any reasoning upon 

 the results it is evidently necessary that the calculated ' seismicities ' 

 should all be derived from observations of the same order. Now, 

 any given country can only furnish observations of one order. But 

 the defect, if defect there be, may be made good. 



Let us represent the seismicities by the expressions S x , S 2 , S 3 , 

 according as they have been deduced from historical, seismological, 

 or seismographical observations. In 94 cases per cent, the values 

 of the relations of the three ' seismicities ' one to the other are 

 known, and one may therefore obtain from these certain empirical 

 formulaB which will serve to calculate any given seismicity by 

 means of the two others. Seismographical observations being few 

 and far between, the seismological seismicities, observed or calculated 

 as above mentioned, have been taken as the standard. We have 

 here the sole means of making rational comparisons. 



IV. Study of Seismic Phenomena in the United Kingdom and 

 the British Colonies. 1 



§ 1. The British Isles. (PL XXX. & fig. 1, p. 654.) 



Historical records regarding earthquakes are, so far as regards 

 the British Isles, numerous and valuable. On the other hand, the 

 earth's crust here is very stable, and earthquakes in Britain have 

 never caused any serious damage. Seismological documents, and 

 a fortiori seismographical documents, are conspicuous by their 

 absence. 



The number of shocks recorded is 1023, felt in 221 localities 

 scattered over 10 definite areas. 



1 Calculated seismological seismicities are distinguished in the following 

 pages by an asterisk. 



2y2 



