24 
PACIFIC SCIENCE, January, 1947 
The time of arrival of the waves in the 
Hawaiian Islands is known with certainty 
only at Honolulu. The record of the Hono¬ 
lulu tide gage (Fig. 2) shows that the first 
rise started at about 6:33 a.m. (Green, C. K., 
1946: 491), though the exact time can¬ 
not be stated closer than 2 or 3 minutes. 
The drum of the water-stage recorder at the 
Waimea River, on Kauai, revolves too slowly 
to give an accurate indication of time, but the 
first rise appears to have started there at 
about 5:55. At Hilo, electric clocks were 
stopped at 7:06, and a brief power failure 
occurred at 7:18. These have been inter¬ 
preted by Powers (1946: 2), probably cor¬ 
rectly, as the time of arrival of two wave 
crests at Hilo. From other considerations, 
discussed briefly elsewhere (Shepard, Mac¬ 
donald, and Cox, in preparation), it appears 
probable, however, that the crest at 7:06 
was the second wave at Hilo, not the first. 
If so, allowing for the observed 15-minute 
interval between later waves, the first rise 
at Hilo probably started at about 6:45. Com¬ 
puted from these times of arrival, the ap¬ 
proximate average speed of the tsunami from 
its origin to Honolulu and Hilo was, respec¬ 
tively, 490 and 498 miles an hour. On en¬ 
tering shallow water the waves decreased 
greatly in speed. The waves moving up 
Kawela Bay, on Oahu, were estimated by 
Shepard to be moving only about 15 miles 
an hour. Similar low speeds near shore were 
reported by other observers, and are com¬ 
parable to the speed of 20 miles an hour 
recorded in San Francisco Bay (Green, 
C. K., 1946: 492). 
The interval between the first and third 
wave crests, as recorded on the Honolulu tide 
gage (Fig. 2), was about 25 minutes, indi- 
HOURS 
Fig. 2. Record produced on the tide gage in Honolulu Harbor by the tsunami of April 1, 1946. 
