8o 



NATURE 



[November i6, 191 



it any molten materials, but was sufticiently violent 

 to throw down one side of a volcano more than 6000 

 feet high, which had been lonjj^ in the " Solfatarra 



J'/ioto.] W- Amicrson 



Fig. 3. — " Pillow Lava," formed by the molten material coming in contact with sea-water. The 



cnilled surfaces form a scum which is stretched and distended by the 

 fjrccd f irward behind, intoth; great p.llowlike mass:s shown above. 



Stage." The quantity of material thus displaced was 

 calculated by Profs. .Sekiya and Kikuchi, the com- 

 missioners appointed by the Japanese Government to 

 investigate the phenomena, to be no less than 2782 

 millions of tons ! European geologists had long been 

 familiar with the fact that among the miniature 

 volcanoes of the Eifel district there are examples in 

 which scarcely any ejection of igneous matter has 

 taken place, the vents being surrounded by fragments 

 of slate and other underlying rocks ; and the experi- 

 ments of Daubree and Sir Andrew Noble have shown 

 that heated gases, under intense pressure, are capable 

 of drilling holes through the hardest rocks. But in 

 the case of Bandaisan, the "hurricane-blasts" hurled 

 along the debris at the rate of forty-eight miles an 

 hour, the hurtling masses of which were gradually 

 reduced to the condition of sand, and covered an aiv.i 

 of twenty-seven square miles, .\lthough " steam ' is 

 said to have been seen rising to the height of 4200 

 feet above the "crater," the materials overwhelming 

 the country were dry, except where they crossed 

 lakes or rivers. But towards the end of the eruption 

 scalding mud is said to have fallen and destroyed 

 many lives ; it is noteworthy, however, that a survivor 

 declared that the gases in which he was enveloped 

 were not of a poisonous character. From the direc- 

 tion in which the materials were thrown, there is 

 some ground for the belief that the outrush of gas 

 was not, as in Krakatoa, vertical in direction, but 

 more or less oblique. The result of the eruption was 

 to produce a great cavity which strikingly resembles 

 the Val del Bove of Etna and the Caldera of Palma. 

 In iqo2 there occurred, in the West-Indian islands 

 of Martinique and St. Vincent, extraordinary out- 

 bursts of volcanic violence, which were carefully 

 studied by English, French, and American geologists. 

 In these cases the great destruction of life and pro- 

 perty was caused by enormous volumes of super- 

 heated gas, charged with solid particles of all sizes, 

 which rushed down the valleys leading from the 

 mountain summits to the sea. These mixtures of gas 

 and dust behaved like liquid torrents, and travelled at 



NO. 2194, VOL. 88] 



a rate of more than eighty miles an hour, sv 

 everything before them. Their destructive ac 

 living beings appears to have been in part du. 

 the heated condition of the gas an(. 

 part to its suffocating character. 

 Martinique another very curio 

 phenomenon was exhibited : the 11 

 of molten rock consolidating in 

 vent of the volcano was gradu 

 pushed up by forces from below i; 

 it formed a great "spine" which 

 tained an elevation of nearly 1 

 feet; but after many changes, du 

 both movement and disintegraii 

 extending over more than a y<"- 

 " spine " finallv disappear- 



It has generally been assumed 

 geologists that the chief componen 

 the gas emanations — alike in Kr. 

 toa, in Bandaisan, and in Martini 

 — was water gas; and this seem«(: 

 be indicated, not only by the \ 

 clouds overhanging volcanic vt 

 but by the torrents of rain wli 

 mingling with finely commini 

 rock, gave rise to the "muddy lav. 

 often so much more destructive 1 

 the "fiery lavas" of active volcai 

 It is true that deposits of chlori' 

 sulphur, and other products arc 

 volcanic vents indicate the present 

 v.'irious other gases in the upri~ 

 columns, but these have usually been regarded as ci 

 subordinate to the high-pressure steam to which 

 most important action in volcanic eruptions 

 usuallv been attributed. 



liquid material 





m' 



PAoto.] [£■ O. H<r-r 



Fig. 4. — The great "spine," composed of andesitic lava, which w. 

 forced up during the eruption of 1903 to ihe height of near 

 laoo feet above the crater-rim of Mont Pelee in Martinique. 



It has long been known that lavas after li 

 consolidation are found to contain, occluded in their , 



