THE GEEAT ]:RITTI0N OF SAKURA-JIMA IN 1914. 87 



hira flous carefully (iroidcd this Si'dion of Sohira-jima. Even the 

 temperature of the hot-spring in Fiirusato was not greatly affected, 

 l)eing 45^ C. which was only 2^^ C. liigher than in former clays 

 {see p. 47). 



Jan. 21st fpjj£ Thikd Phase. The sccond phase of the eruption 

 (luring the interval from the 14th to the 20tli was now closed. 

 The city side was quiet ; lava still moving forwards into the 

 sea. The writer took the train on tlie 21st for Tokyo, so the diary 

 for the following days is not from his personal observations. It is 

 said that from 7.5 p.m., a strong explosion and roaring occurred, 

 continuing throughout the whole night. The sea glared from the 

 lava flow. Ash fell on the city. About a thousand people again 

 fled to the west from the city. 



Nothing was known of the east side. But it nmst have 

 been more active than on the west side, throwing pumice and ash 



A 



to the opposite coast of Osumi. 



-Tan. 22Nr) r^\yQ yr^iYv of asli OU tho city coutiuued from last night 

 (118 grams per ^(\. m, 2 mm. thick). Day^^ and night there was 

 constant roaring. Nothing particular was known from either side. 

 The lateral lava margin was said to have been pushing forwards 

 one foot'^ an hour near Hakamagoshi on the west. 



Jan. 23ed From midnight it began to roar, giving a sound like 

 a landslide. This was probably caused by the widening of pits on 

 the east side. The citizens passed a restless night ; the afternoon 

 was misty, due to floating ash (21.5 grams per sq. 772.) ; a strong 

 earthquake occurred at 9 A,:\r., the day was rather quiet due to 

 the outpouring of lava, but at night it again roai-ed. 



1) About the sjk titer »■rupthii'.s observed on this day the reader is referred to tlie SumriKiry. 



2) Vldfi ante, p. 76, footnote 2. 



