THE (iÜEAT ERUPTION OE SAKURA-JIMA IN 1914. 109 



and gases in the picture are generated in such a tunnel, and gusli 

 out at the end. "J'he channels which present the appearance of 

 natural levees are really collapsed tunnels with broken-in roofs, the 

 vault material being partially carried forwards with the flowing 

 liquid rock. The event is said to have happened some time between 

 the end of March and the beginning of April, 1915, i.e., after a 

 lapse of fourteen months after the great eruption. 



Another feature, no less striking, is what the writer calls by 

 the name of the gas-blowing horn at the largest tunnel's end, as 

 may be seen in one white spot in the background of PL XII. 

 Fig. 1 (/?). An enlarged and nearer view of it is the lower picture 

 (Fig. 2 (/?)). This volcanic spouting horn is something like that 

 often found in cliff-caves exposed to stormy coasts and like that 

 it is generated in an analogous way, diftering only in mode, which 

 is here by a dry, not a wet process. Pent-up vapor and gases in 

 the lava-tunnel forced their way out, causing this remarkable 

 volcanic blowing horn at the water's edge. It is a special kind of 

 fumaroles. 



A relinquished lava-grotto is found near Cape Nagasaki 

 (Geologic Map, gr) on the southeastern shore, where it opens at 

 the termination of a lava -tunnel of an ancient flow at the water's 

 edge. Here a cold " infernal " wind, so say the people, gently 

 blows throughout the year from the deep interior. The tromba 

 on land as well as in the sea [postea, p. 115) may be explained, 

 according to the writer's opinion, on the same principle. 



In the western field (p. 105) the lava flowed down from under 

 a partially consolidated supercrust (on the front of PL XII. Fig. 3) 

 toward the sea, where it deployed into two fan-shaped areas with 

 an indentation between them, the head of which marks the posi- 

 tion of the buried Karasu-jima. The back of the main front 



