Travel Times of Seismic Sea Waves— Zetler 
central positions are on or off shore. The 
method used in this study necessitated a 
given depth of water to make a computation 
possible, and therefore the times were com¬ 
puted to the ocean deeps which are near the 
plotted epicenters. The comparison of ob¬ 
served and computed travel times would 
seem to indicate that this procedure is reason¬ 
able. 
Three seismic sea waves which originated 
near the Japanese island of Honshu have 
been recorded on the tide record at Hono¬ 
lulu. The observed travel times exceed the 
computed times by 14, 23, and 49 minutes. 
The first two differences are sizable but not 
necessarily serious. There are several fac¬ 
tors which can contribute to a small con¬ 
sistent variation. Accurate computations of 
travel time require adequate and reliable 
soundings. Some areas in the Pacific are in¬ 
adequately surveyed, and the accuracy of 
some soundings in other areas is question¬ 
able. For the 120-mile travel path sections 
which contain large variations in depth, the 
procedure was to use a mean depth after 
rejecting occasional extremely shallow 
depths. This technique and the resulting 
mean depths are somewhat subjective and 
could lead to small inaccuracies in travel 
time. 
The seismic disturbance off the south coast 
of Honshu on December 7, 1944, created a 
sea wave which arrived at Honolulu 9 hours 
and 9 minutes later, whereas the computed 
travel time for the path is 8 hours and 20 
minutes. The epicenter of the December 20, 
1946, sea wave is approximately 170 nautical 
miles west of the 1944 epicenter, the water 
between the two is relatively shallow, and 
great circle paths from the two to Honolulu 
would virtually coincide. The difference of 
only 3 minutes between the observed travel 
times cannot be explained by some of the 
difficulties mentioned above. It seems prob¬ 
able that some of the 49-minute difference 
must be attributed to tide gage operation, 
187 
tide record interpretation, the seismological 
determination of epicenter, or a combina¬ 
tion of these factors. 
A serious misinterpretation of a tide mari- 
gram can take place if the waves are small in 
amplitude. The first wave is usually smaller 
than those immediately following it and may 
not be recognizable because of the seiche. 
Therefore a time of arrival which appears 
to be definite may refer to the second wave 
rather than the first. With an observed 
travel time greater than its true value, the 
time difference, computed minus observed, 
is large negatively. 
Besides giving the travel time of a seismic 
sea wave from an epicenter, the chart may 
also be used in conjunction with apparatus 
which detects the sea wave as it moves 
toward the Hawaiian Islands. Midway 
Island, for example, lies approximately on 
a great circle arc between Honshu and Hono¬ 
lulu. If a wave were received at Midway 
from an epicenter near Honshu, a warning 
could be sent to Honolulu. The travel time 
difference between Midway and Honolulu 
of about 2 hours and 50 minutes is readily 
obtained from the chart. The necessary 
condition for using the chart in this fashion 
is that the outpost lie close to the great circle 
course from the epicenter to Honolulu. 
There are a number of other islands in the 
Pacific Ocean whose positions make possible 
their use as detector outposts for waves from 
other directions. 
Travel time to other places in the Ha¬ 
waiian Islands can be estimated with the aid 
of the time curves near the islands. For 
example, a seismic sea wave originating near 
Chile would reach Hilo about a half-hour 
before it reached Honolulu, but a sea wave 
from southeast Alaska would arrive at both 
places at approximately the same time. 
Table 1 gives the geographic position of 
epicenter, the Greenwich time of the seismic 
disturbance, the location of the gage which 
recorded the receipt of the seismic sea wave 
