Key points are not available for this paper at this time.
Summary and Conclusions A comparison was made of all the periods of local earthquakes entered in the record books, and this showed that the P wave of 0.3-sec. period occurred a maximum of 156 times, and a secondary peak for the period of 0.5 sec. occurred 89 times. The S wave of 0.5-sec. period had a maximum of 129 occurrences, and a secondary peak for 0.8-sec. period had 100. This suggested that in any earthquake the ratio of the period of the S to the P wave was inversely as their velocities, or as the square root of three. The maxima just given appear to hold for such waves from all depths of origin. It had been noted previously that large amplitudes and periods occur together. The upper limits of the amplitudes of the P and S waves of local shocks were found to vary with the cube of the periods. Different results were found for the variation of epicentral shocks in California and Japan. The difference may be caused by the difference in physical characteristics of the underlying crustal rock. While these studies in Hawaii were made on shocks of intensities I to IV, Rossi-Forel, they show promise of giving information about the waves to be expected in destructive earthquakes. The sectorial lines may be raised by new data, but in each region should approach some unknown lines as a limit. Formulas were used to correct the observed waves to those of standard displacement and consequent period. These periods were plotted with respect to distance and depth, with no reliable result. A tendency was shown for the period of P waves to increase with distance more rapidly than the period of S waves, whereas observations of more distant earthquakes would suggest the opposite. Study of the ratios of the amplitudes of the P to the S wave (AP/SS) showed no distance effect. The formulas from the previous amplitude-period study suggest that this ratio should not vary with the local distance. For Hawaii the ratio averages about 15 per cent. About 60 per cent of the foci are less than 5 km. deep, 70 per cent less than 10 km. deep. Very few appear to have originated at 60 or more km. depth. The decline in numbers of earthquakes with depth is a rapidly decreasing exponential function. Most of the deep earthquakes are under Mauna Loa and the Kilauea southeast rift zone. A large number of the located shallow foci are in and near the Kilauea crater. Possibly this is an increase that should be expected near any active volcanic crater, but it may be due to the close network of stations about Kilauea crater. The magnitude of the shock is not a function of the location either areally or in depth; that is, large earthquakes may be expected in any part of the island and near-by sea bottom and at all depths to at least 60 km. A method of classifying the earthquake records is based on the number of P or S waves shown on the seismogram, which indicate the key number from one to seven. A map of Hawaii was constructed showing the areas in which the different types of shock had originated. The first type, K-1, occurs either central to Mauna Loa or within 50 to 60 km. radius of the seismograph. Type K-2 is not recorded from northwest Hawaii. Type K-3 does not occur close to the instruments. Types K-4 to K-7 are noted to occur at somewhat greater distances, and to date have been observed only from small outlying areas. Earthquake records of simple character are generally near the area of deep-focus shocks and near the seismographs, so that the waves come in at a steep angle. Earthquakes under Kilauea crater are generally simple. As the foci become more distant and shallow they also become more complicated in type. These criteria should help in designating phases and consequent locations, but they are not final, and may be of no help beyond 100 km. The number of phases in some of the records of outlying earthquakes suggest a complexity of structure in the island mass and the near-by sea bottom. The locations near and on the extension of rifts and in pronounced lines and zones suggest a larger and more numerous system of rifts than has previously been mapped. The resulting pattern of rifts about Mauna Loa is roughly an asterisk. The main accent is on the visible active rifts to the southwest and the east-northeast of Mokuaweoweo. These rifts have apparently controlled most of the island's seismicity in the immediate past.
Austin E. Jones (Sat,) studied this question.