The warning system' is a co-operative undertaking. It binds into a 

 network existing Coast and Geodetic Survey seismograph and tide sta- 

 tions b}- using facilities of the U.S. Navy, Army, Air Force, and Civil 

 Aeronautics Administration. These agencies, along with the Coast 

 Guard, the Territorial Government, the City of Honolulu, and others, 

 will broadcast warnings by radio, siren, airplane, and an}^ other means 

 necessary to clear the beaches and waterfront areas of threatened 

 communities. 



Seismograph stations in the system are the Coast Survey stations at 

 Tucson, x\rizona ; Sitka, Alaska ; Fairbanks, Alaska ; and Honolulu. 

 To strengthen the system, the California University observatories at 

 Pasadena and Berkeley will doubtless participate from time to time. 



Tide stations that can be used for wave detection include three in the 

 Aleutians, one each at Midway, Johnston Island, Palmyra Island, Pago 

 Pago, Kwajalein, Hilo, Honolulu, and others, including numerous stations 

 on the mainland of North and South America. 



xMthough the warning system is not yet established in a complete 

 form, it is in operation, and the quake of 8th September, 1948, in the 

 region of the Tonga Islands provided a test of its efficacy. In this inst- 

 ance perhaps more time was consumed in obtaining seismograph reports 

 than was desirable, but the time available to the central stations was 

 sufficient to obtain reports from tide stations and others in the path of 

 a possible wave. The travel-time chart showed that it would take a wave 

 about 6 hours and 35 minutes to travel from the epicentre to Honolulu. 

 As the time approached for the wave to arrive, planes took to the air in 

 readiness to clear beach and waterfront areas, and all agencies were on 

 the alert and readj^ to broadcast warnings. As the wave reports came in 

 it became clear that though there probably was a wave, it was very smaU. 

 The " Stand b}^ " order was therefore rescinded and the " All clear " was 

 given. When the record was later removed from the Honolulu tide gauge, 

 it showed that a seismic sea wave had arrived about on time. Its height 

 was 6 in. 



So far as the public was concerned, the warning system operated 

 satisfactorily, but from the point of view of those at the centre the test 

 indicated the desirability of some adjustments and additional studies. 

 The basic plan, however, proved to be well designed to accomplish the 

 task for which the system was created. 



WAVE ANALYSIS AND PREDICTION 



By G. E. R. Deacon, Admiralty Research Laboratory, England 



[A bstract] 



As soon as continous wave records were examined the fact that the 

 observed wave profile is the resultant of a number of wave trains had to 

 be recognized, and methods had to be developed to detect and. measure 

 each invisible wave-train to obtain the wave-spectrum. The examination 

 of many such wave-spectra and the relevant wind charts shows that a 

 wind blowing over the sea produces waves of all lengths up to a maximum 

 that depends on the greatest wind strength. These component wave- 

 trains travel independently across the ocean with the group velocities 

 appropriate to their periods ; the short waves generated at the beginning 



362 



