JAPANESE EARTHQUAKES — HECK 217 



peated triangulation and leveling, long-continued and preseismic tilt, 

 and all kinds of periodicities. The earthquake of 1930 in northern 

 Idu indicates how difficult it is to correlate crustal movement with 

 earthquakes. A case has been made from occurrences in earthquakes 

 of the Japan Sea coast for preseismic tilt, from 30 minutes to 5 hours 

 before the earthquake. The study of periodicities has not brought 

 significant results, except that there is an interval of several cen- 

 turies between major earthquakes of any region. The outstanding 

 examples of great success and great failure in prediction are credited 

 to Omori. After the California earthquake of April 18, 1906, he 

 predicted that the next great shocks in the circumpacific belt would 

 occur in South America and Alaska. On the same day, August 16, 

 1906, there were great earthquakes in Valparaiso, Chile, and the Aleu- 

 tian Islands. In the spring of 1923, Omori expressed the view that 

 there would be no great earthquake in the Kwanto region for many 

 years, and the great shock came on September 1. 



The Japanese have gone in strongly for earthquake-resistant struc- 

 tures, and miscellaneous measures have been taken. 7 Heights of 

 buildings have been restricted, simple structures insisted on, and water 

 supply that is not likely to be affected by earthquakes has been es- 

 tablished in Tokyo at least. The effectiveness of these measures has 

 not been tested, and the description of the 1923 shock has brought out 

 many things to guard against. 



7 Willard Price, Unknown Japan, Nat. Geogr. Mag., vol. 82, pp. 225-252, 1942. 



