54 Earthquake at the Philippine Islands, 
these two elements, the magnitude and direction of the oblique 
undulations. 
From the indications of these two instruments the following 
results were obtained, during the successive days of the great 
earthquake. We do not attach to them an absolute value, 
since the seismometers cannot make perfectly correct observa- 
tions, when such movements are of great violence and compli- 
cation. Yet we believe that they afford quite a good registra- 
tion of the phenomena, and will be useful for comparison with 
those of other earthquakes. The facts obtained are as follows. 
The vibrations began during the months of April and May, 
in the northern provinces of Luzon. The center of oscillation, 
between Lepanto and Abra, in the central Cordillera of 
Luzon, in latitude 16° 22’ N., and longitude 127° E. from the 
Spanish Observatory of San Fernando. At first the move 
want of care in the method of taking the observations, exact- 
the purpose. 
Early in July some vibrations were felt; yet from the 5th 
to the 14th none were recorded at Manila for any point on the 
islan 
On the 14th, at 12" 53’ p. w., when a storm from the north- 
east of Luzon was threatened, as indicated by an extraordinary 
fall of the barometer, the first shock occurred in which it was 
observed that there were two centers of oscillation (see figure 1); 
one in the second quadrant from the point where the oscillation 
of the pendulum of the horizontal seismometer commenced, an 
the other in the third, in which the oscillation of this first move 
ment—mainly horizontal in direction—ended. The total am- 
plitude of the oscillation reached 5° 25’. The horizontal pen- 
ulum left inscribed a cross whose arms intersected at a right 
angle, the first set, from 8. 55° E. to N. 55° W., and the other, 
m 8. 40° W. i. 
The first impulse was in the direction from S.E. to N.W., 
and the amplitude of the oscillation in this direction covered a0 
are of 5° 25’, all apparently a result of the first shock ; then 
the pendulum was violently oscillated in a direction perpendic- 
ular to this, and with an amplitude a little less than in the 
former case. The index of the vertical seismometer WS 
moved 4™™ from its position. 
