RECENT WORK ON VOLCANOES 89 



lifted and from the resulting crack gas was being vigorously 

 expelled, rising with a violent whirling motion like that of a 

 water-spout. The gas and fumes were insoluble in air. At the 

 moment of the explosion not much could be seen, but from the 

 number and velocity of the ejected blocks it was evident that 

 the nearest chimney had entirely emptied itself. The rim on 

 which he stood was swept with fumes, but there was no con- 

 densation of moisture on the cool surface of the rocks. On 

 another occasion Brun thrust his geological hammer into the 

 uprushing stream of gas and no water was condensed on the 

 bright metallic surface. 



In a neighbouring volcano, Bromo, the continued explosions 

 prevented Brun from looking down into the crater. So he 

 caused a little platform to be cut in the loose cinders just under 

 the rim on the outside ; on this he established his battery of 

 thermometers, barometers and hygrometers, and also a little 

 pump which had attached to it a long train of glass tubes 

 connected by indiarubber joints, which was dangled into the 

 crater. When an explosion took place the hygrometer showed 

 no excess of moisture in the air. I can, however, find no 

 account of an analysis of the gas thus collected directly from 

 the throat of the volcano by the pump. Elsewhere Brun relies 

 on the gases occluded in the lavas ; these are expelled on 

 heating the rock to a certain temperature above the melting- 

 point. Plutonic rocks and lavas which have been in existence 

 for long geological periods are " dead," and do not contain, or 

 have lost, occluded gases. Recent lavas when heated to their 

 explosion-point suddenly give off with tremendous violence 

 large quantities of chlorides — magnesium, iron, and silicon — 

 together with ammonium chloride, carbon dioxide, carbon 

 monoxide, marsh gas, chlorine, hydrogen chloride, and less 

 frequently sulphur dioxide and sulphuretted hydrogen, and 

 lastly, hydrogen and nitrogen, but neither oxvgen nor water. 

 Gautier (7) points out, however, that the gases of fumaroles are 

 generally hydrous. But then fumaroles belong to a late stage 

 of the volcano, when the activity is dormant and water from 

 the surrounding rocks can percolate and attain to the hot centre of 

 the volcano and thence be driven up to the surface of the earth. 



Fouqu6's analyses of the gases from Santorin in 1866 (8), 

 although collected from the surface of sulphurous water in a 

 fissure, contained only traces of oxygen but nearly 30 per cent. 



