272 
PACIFIC SCIENCE, VoL VI, October, 1952 
damage to the Bosch-Omori seismograph 
and put it back in operation at 01:24, 27 
minutes after the first earthquake started. 
At that time the instrument was recording the 
long waves of a large earthquake. The period 
of these waves ranged from about 6 to 8 
seconds and averaged approximately 6.7 sec- 
onds. Their maximum double amplitude was 
67 millimeters, corresponding to a theoretical 
ground displacement of approximately 0.5 
millimeter. These waves continued with 
gradually decreasing amplitude until 03:20. 
Because no other earthquake at an appropri- 
ate time was observed by more distant sta- 
tions, it is believed that these long-period 
waves were the surface waves of the major 
Kona earthquake. 
The time of origin of the major earthquake 
is given in the notice of preliminary deter- 
mination of epicenter issued by the U. S. 
Coast and Geodetic Survey as 00^^56’^57.5® 
Hawaiian Standard time (10^^56”’57.5® Green- 
wich Civil time). The time of beginning of 
registration of the preliminary waves at the 
Whitney Laboratory at Kilauea was 00^^57"^ 
09.5® Hawaiian Standard time. 
The direction of the first ground movement 
at Kilauea Caldera was east-southeast and up, 
that at the Mauna Loa station was east- 
northeast, and that at the Kealakekua station 
was east-northeast. At the Kealakekua station 
the north-south component was only slightly 
damaged, but on the east-west component 
the suspensions were broken and the weight 
dropped on the floor 2 feet west of the pier. 
The Kona seismograph, at Konawaena 
School (Fig. 1), was restored to operation 
at 15:15 on August 23. Previous to that time, 
location of the points of origin of the after- 
shocks on an instrumental basis was uncer- 
tain because of the very short base of the 
triangle formed by the intersection of lines 
from the earthquake foci to the other stations. 
Earthquakes after that time are fairly well 
located because of the control given by the 
Kealakekua seismograph. Most of these were 
located by means of data from four stations: 
Kealakekua, Mauna Loa, Hilo, and Whitney 
(Kilauea) . 
Locations of the epicenters of aftershocks 
which occurred after 15:15 on August 23 
with serial number greater than 190 are shown 
in Figure 1. Thirty-three such aftershocks 
have been located with small probable error. 
Most of them fall on or close to a fault that 
runs out to sea in a west-northwesterly di- 
rection along the northern edge of Kealake- 
kua Bay. The existence of this fault, partly 
buried by later lava flows, has been recognized 
for many years (Dana, 1890: 30; Stearns and 
Macdonald, 1946: 37, pi. 1). At its eastern 
end it bends southward, and the writers have 
suspected that the abnormally steep lower 
western slope of Mauna Loa inland from the 
highway for 15 miles or more south of Cap- 
tain Cook is a fault scarp deeply buried by 
later lava flows. An interesting partial con- 
firmation of this theory is furnished by the 
location of the epicenters of several after- 
shocks along this line (Fig. 1). 
The frequency of aftershocks decreased 
rapidly from August 23 to September 4. As 
is shown in Figure 2, the average frequency 
then decreased very slowly until the end of 
September. No figure is available for Sep- 
tember 7 because of mechanical failure in the 
recorder at the Kealakekua station. The ap- 
parent depth of origin of the aftershocks 
ranged from 3 to 12 miles, most being about 
6 or 7 miles. No progressive change of depth 
with passage of time is apparent. 
Fig. 2, Graph showing frequency of aftershocks to 
the end of September, 1951. 
