ERUPTION OF USU VOLCANO — MINAKAMI ET AL. 265 



here run from north to south. As a result of the upheaval of the 

 ground that now took place the roads and other passages were cut off 

 by cracks and dislocations in the ground, rendering the area im- 

 passable. The underground waters, too, were considerably disturbed. 

 Villagers in the disturbed area no longer had water for drinking and 

 irrigation, while in the adjoining villages the wells and fountains 

 overflowed. 



From March to the middle of April the rising area stretched along 

 the Kami-Osaru-Yanagihara road for a distance of 4 kilometers. The 

 greatest rise, as determined by leveling, was 16 meters, or at a rate of 

 30 centimeters per day. Just at this stage the locus of rising shifted 

 about a kilometer toward the north from Yanagihara to Hukaba and 

 the adjoining cornfield at Kumantubo, where numerous cracks of vari- 

 ous size and disposition were formed. 



As soon as the rising area had shifted northward, the earthquakes, 

 which are believed to have had an extremely shallow origin, occurred 

 directly below Hukaba. After the middle of June more than 100 

 shocks per day were experienced there and reached 250 shocks on the 

 day preceding the first paroxysmal explosion, June 23. It is needless 

 to say that during this early stage of activity, which had now en- 

 dured for 6 months, the inhabitants of the area were extremely un- 

 easy, wondering when volcanic explosions would occur and where the 

 craters would open. The maximum upheaval during this stage 

 reached a height of 50 meters. 



STAGE OF PAROXYSMAL ERUPTION 



On June 23, 1944, the eruptive column of the first explosion was ob- 

 served, rising noiselessly from the northwestern corner of the corn* 

 field at Kumantubo. This explosion gradually increased in intensity, 

 hurling mud, ashes, and blocks of rock from a newly formed crater. 

 Simultaneously a mud stream flowed from the crater, pouring into a 

 nearby depression to form a pond of mud and hot water. After 

 several hours of recurring mud flows the eruption came to an end, 

 leaving a calm pool of hot, muddy water in the bottom of the crater. 

 After a few hours of quiescence, steam again began to rise and activity 

 was resumed with the spouting of muddy water. The activity con- 

 tinued variable ; at its height blocks were thrown as high as 800 meters. 



A strong explosion of similar type occurred on the morning of June 

 27, when a second vent opened near the first. Activity in the first 

 crater gradually diminished, the muddy explosion shifting to the 

 new vent. 



The eruptive activity bore a striking resemblance to the phenom- 

 enon of a geyser. It appears that these muddy eruptions were not 

 caused by the pressure of gases emitted by the fresh lava but by the 

 pressure of steam generated by the heat of the new magma. Since 



