ITI. John Milne. 395 
To locate the epicentre of a distant earthquake Milne employed 
chiefly two methods. The first depended on the constancy of the 
velocity of the maximum large waves and required the absolute 
times of their passage at not less than four stations. The second 
depended on the fact that the interval between the initial fremors 
and the maximum large waves is constant for the same distance from 
the epicentre and varies with the distance. Thus, if the interval 
were given for at least three stations, the corresponding distance could 
be determined from the time-curves and tables, and the intersection 
of the circles with centres at the stations and these distances as 
radii would give the position of the epicentre. 
In the report for 1900 the first map of earthquake-origins (for 
1899) was given, the positions of the epicentres being denoted by 
small circles. By 1902, the number of origins being increased, Milne 
found it possible to define the principal regions of seismic activity. 
From this time onwards the map was a feature of each year’s 
report, retaining the same form until 1910, when greater detail 
could be attempted. The great merit of the maps is their complete- 
ness, for of the stronger shocks not one can now escape being 
recorded. On the other hand, the period that has elapsed since 
1899 is but a brief interval in the history of the earth, and all that 
the maps can provide us with is an outline of those districts in which 
crustal changes are at the present time taking place most rapidly. 
From the map for 1899-1909 we see that of every ten earthquakes 
seven belong to the districts bordering the Pacific Ocean, one to the 
strictly oceanic districts of the Indian and Atlantic Oceans, and two 
to the great terrestrial district which extends from the Apennines 
to the east of the Himalayas. 
Conclusion—In the preceding summary of Milne’s work in 
seismology, I have omitted many points of interest. As somewhat 
outside our range I have left untouched the practical applications 
to which he humanely devoted so much of his time and attention, 
and especially to the structure of earthquake-proof buildings.’ 
In common with other seismologists, Milne became absorbed for a 
time (especially 1884-6) in the study of earth-tremors. The early 
results obtained with his ‘‘ automatic tremor recorder’? were to 
some extent corrected by his later observations with the horizontal 
pendulum. Moreover, the connexion between earthquakes and 
tremors seems to be somewhat remote.2 In 1897 he read an 
The latest and most trustworthy tables are those prepared by Turner and 
published in the Circulars issued by the Seismological Committee. The bearing 
of the time-curves and especially that of the second preliminary tremors, on 
the constitution of the earth, has been discussed by several writers, especially 
-R. D. Oldham (Quart. Journ. Geol. Soc., vol. 62, 1906, pp. 456-73) and 
C. G. Knott (Proc. Roy. Soc. Edin., vol. 39, 1919, pp. 157-208). 
1 Trans., vol. 1, pt. 2, 1880, pp. 64-85; vol. 2, 1880, pp. 27-38 ; vol. 11, 
1887, pp. 115-74; vol. 14, 1889, pp. 1-246; vol. 15, 1890, pp. 163-9. 
2 Trans., vol. 7, pt. 1, 1883-4, pp. 1-15; vol. 11, 1887, pp. 1-78 ; vol. 13, 
1890, pp. 7-20; Rep., 1895, pp. 27-31. 
