ON SEISMOLOGICAL INVESTIGATIONS. 91 
We have here sixteen instances in which the azimuth of an origin 
determined from the maxima of North-South and East-West motion 
approximately agrees with the azimuth as measured on a globe. There 
are, however, in the same interval of time twenty-six instances where 
no such agreement exists, and this I find to be the case for all the 
large records obtained during the latter half of 1909. The inference 
is that the main portion of teleseismic motion, like that of macro- 
seismic motion, generally takes place in directions independent of the 
azimuth of its origin. 
V. On the Relative Duration of Two Rectangular Components of 
Earth-movement at a given Station. 
The records I refer to were made at Shide by a pair of light horizon- 
tal Milne pendulums mounted on the same cast-iron frame, and 
installed upon a brick column. One of these recorded N.-S. motion 
and the other E.-W. motion. The duration of the movements of 
the latter practically agreed with the duration recorded by a similar and 
similarly oriented pendulum on a separate column. From this it is 
inferred that although two light pendulums are carried on one stand 
they had no sensible effect upon each other’s movements. 1 have 
divided the records into the following four groups, the natural period of 
the pendulums being different in each group :— 
Group 1.—May 25, 1901, to January 1, 1902. Period of N.-S. boom 
19 seconds, and K.-W. boom 17 seconds. Ten large earthquakes had a 
total duration of 1,341 minutes fcr N.-S. movements, and 1,313 for 
E.-W. motion. Pendulum with the longest period moved for the 
longest time, but the difference is very small. 
Group 2.—¥February 27 to December 28, 1903. Period of N.-S. 
boom 20 seconds, and H.-W. boom 17 seconds.’ Eleven large earth- 
quakes had a total duration of 730 minutes for N.-S. motion, and 725 
for K.-W. motion. 
Group 3.—January 20 tc June 27, 1904. Period of N.-S. boom 
20 seconds, and E.-W. boom 30 seconds. Nine large earthquakes had 
a total duration of 1,036 minutes for N.-S. motion, and 1,067 for 8.-\W. 
motion. Here again the pendulum with the longest period was dis- 
turbed for the greatest length of time. 
Group 4.—July 24 to October 9, 1904. Period of both pendulums 
25 seconds. The total duration for eight earthquakes was for N.-S. 
motion 725 minutes, and for E.-W. motion 732 minutes. In this 
instance pendulums with similar periods have been kept in motion for 
equal intervals of time. Nineteen large earthquakes between July 3 
and December 10, 1909, show a similar result. 
From the above notes it might be inferred that the apparent dura- 
tion of a teleseism largely depends on the sensibility of the recording 
apparatus to tilting. A detailed examination of these records, how- 
ever, distinctly shows that this is not the case, and that a pendulum 
with a short period is frequently in movement for a longer interval of 
time than one with a long period. Amongst the earthquakes referred 
