] 72 A. Imamura 



of the former has been very slow but not without some results, such as 

 I shall now describe. 



2. Earthquake prediction. Several methods for the solution of 

 the problem of earthquake prediction have been proposed and tried by 

 different investigators but almost all have, so far as I am aware, ended 

 in failure. Japanese seismologists, especially the late Prof. Omori, 

 have, however, conveniently divided the subject into two parts; first, 

 the place of occurrence, and second, the time. Of these, the first was 

 investigated by Prof. Omori, with definite success, and his results were 

 published under the title of the " Seismic zone." His investigation oi' 

 the seismic zone of the Pacific coast of America and of that of 

 Italy led him to predict the Valparaiso Earthquake in 1906, and tw(3 

 Italian earthquakes, iiamely the Messina-Reggio Earthquake in 1908 

 and the Avezzano Eartliquake in 1915. 



The seismic zones in certain localities can be simply determined by 

 the geological formation and the topographical features ; nevertheless 

 it is very necessary to know the space distribution of past earthquakes 

 in the particular locality in order to be able to forecast the places where 

 large earthquakes are sooner or later likely to occur. In fact, the ac- 

 curacy of the determination of seismic zones depends upon that of the 

 determination of the position and. area covered by past earthquakes. 



Let us first suppose the site of a future earthquake to be given ; 

 then the problem as to when the earthquake will occur presents itself. 

 It is apparently very hard to solve this question, but it does not seem 

 to be impossible, especially when we consider such indications as 

 foreshocks, slow changes of land levels and other phenomena, which 

 are often manifested in or near the origin of a big earthquake beioie 

 it takes place. For this reason, the study of seismic zones now plays 

 a very important part in investigations undertaken in countries subject 

 to earthquakes. 



3. Catalogue of Japanese earthquakes. The history of Japanese 

 earthquakes dates back as far as the fifth year of Emperor Inkyo, i. e. 

 416 A. D. The luiperial Earthquake Investigation Committee has 

 compiled a catalogue containing more than 2200 earthquakes which 

 have taken place during the successive 14.51 years^\ Of these records, 

 the part relating to the years ])revious to 1595 is considered to be 

 rather imperfect, so much so that it is liable to give an inaccurrate 



1) Eeport (in Japanese) of the Imp. Earthquake Inv. Comiu., No. 46 



