New Hebrides, of 19 13-1 ^ 539 



Niauafau in the Tonga Islands in 1886. In 1894 was the first great 

 eruption of Ambrym since about 1820 ; in 1902 began the eruptions 

 of Savaii, which were repeated in 1905 and 1906 ; in 1902 happened 

 the explosive eruptions of Mts. Collima and Santa Maria in 

 Central America, followed by those of Mts. Pelee and St. Yincent 

 in the Atlantic border of the West Indian area; in 1906 occurred the 

 eruptions of Topia and Fanua-lai; in 1908 that of Puna in Hawaii, 

 and in 1909 that of Korintzi, Sumatra; in 1912 the explosion of 

 Katmai in Alaska caused sunset glows in Europe ; Sangir Island, on 

 the edge of the Pacific, between the Philippines and Celebes, broke 

 into eruption in March, 1913, followed by that of Ambrym in 

 December of the same year. 



The two eruptions with which it is natural to compare the Ambrym 

 outbreak of 1913 are that of Tarawera, which it resembles by the 

 formation of a volcanic trough by explosions along a fissure, and that 

 of Sakura-jima in Kyusu, Southern Japan, in 1914. That eruption 

 was practically contemporary with the outbreak of Ambrym, and it 

 has been described in detail by Professor Koto (1916, 1 and 2). 

 It began towards the end of 1914 with small earthquakes and with 

 eruptions from Kirishima, 34 miles north of Sakura-jima, in 

 November and December, 1913; on January 12, 1914, Sakura-jima 

 was devastated by an explosive eruption which lasted for nine hours ; 

 it began from the high central crater of the island, followed by the 

 opening of two vents on the flanks of the island, the three vents 

 being situated on a line running approximately east and west. Lava 

 streams flowed from these vents, and one stream crossed the Strait of 

 Seto and thus converted the island of Sakura-jima into a peninsula. 

 The composition of the Sakura-jima lava is shown by the analysis, 

 No. 5 on the table, p. 14 ; it is an andesite, though it contains small 

 quantities of corroded and partially resorbed olivine. Professor Koto 

 states that when he first saw the eruption he considered that the 

 three linear vents were along one fissure traversing the island, and 

 he says that the belief in such a fissure is universal (Koto, 1916, 2, 

 p. 138). He has, however, himself come to the conclusion that the 

 evidence for this fissure is not convincing. He regards Sakura-jima 

 as built up around three centres, and that the recent eruption found 

 outlets at the weakest points, as determined by the compound 

 structure of the island. The fact that these were linear he apparently 

 regards as a coincidence. In an interesting transverse section through 

 the island he explains the recent eruption as discharging a central 

 chamber partly through a vertical central pipe, but mostly through 

 the lateral outlets at lower levels. There appears nothing inconsistent 

 in this arrangement with the determination of the position of the 

 lateral and central vents by a plane of weakness along an east and 

 west fissure. The existence of such a fissure can only be a matter of 

 conjecture as regards Sakura-jima, but the evidence for it at Ambrym 

 appears conclusive. 



7. References. 



Beresfokd (C. W. de la Poer). 1884. "Note on the Ascent of Ambrym 

 Volcano in the New Hebrides " : Proc. Boy. Geogr. Soc, vol. vi, 

 pp. 129-32. 



