REPORT OF THE CHIEF ASTRONOMER 711 



SESSIONAL PAPER No. 25a 



slowly. A special aggregate or swarm of them would therefore exert a' strong 

 buoyant effect on the mass of magma in which they are entangled. The prin- 

 ciples of fluid dynamics show that the mass of specially vesiculated magma 

 would rush up the conduit at comparatively high speed. Arrived at the surface, 

 it parts with much of its dilating gas, grows heavier, and sinks. Its place is 

 taken by a later uprushing mass ; the rhythmic action is more or less continuous. 

 Since a gas phase and a liquid phase are essential to the process, this powerful 

 method of circulation may be called 'two-phase convection.' 



Again the writer must here omit detailed arguments on a subject which 

 seems to him of considerable importance. Suffice it only to remark, first, that 

 two-phase convection is visible throughout the activity of the lava lakes of 

 Hawaii and Savaii, and has been observed in the lava of the craters at Vesuvius 

 and Etna. Secondly, the hypothesis seems to be well supported by actual calcu- 

 lation of its efficiency, as will be indicated in the special paper. 



That paper will also sketch the arguments for the view that much of the 

 heat emanating at a central vent is not primary but is the product of chemical 

 leactions, chiefly among the gases, in the lava column. Herewith we have 

 partial explanation for the long lives of many volcanoes. Their vents are kept 

 cpen, partly because of the manufacture of heat in the conduit by exothermic 

 gaseous reactions; and partly through the convective transfer of heat by the 

 formation of a gas phase in the lava column. Since in both respects the presence 

 of magmatic gases is vital to the continuance of activity, this view of the essen- 

 tial nature of eruptivity at central vents may be called the ' gas-fluxing hypo- 

 thesis.' 



Gas-fluxing explains the localization of the vent. In the parent abyssal 

 injection the gases rise and collect about points in the roof of the magma 

 chamber. The highest point in the roof will, in the end, attract the rising gas 

 most effectively. As the gas tension increases, the strength of the roof at the 

 point of special gas accumulation may be overcome and an explosion opens a 

 vent (a diatreme) to the Earth's surface. Or, if a fissure is opened above the 

 point of gas accumulation, it may be enlarged to vent size, first by outrushing 

 hot gas and then by two-phase convection. 



In the process of time every central vent must become more or less perfectly 

 cylindrical — a solution form. In this respect, as well as in the small size of these 

 vents, even in the mightiest cones, the gas-fluxing hypothesis is supported by 

 the facts of nature. 



Gas-fluxing also explains the periodicity of the activity at central vents. 

 A long period of activity tends to exhaust the supply of gas in the conduit and 

 immediately below it, that is, in the uppermost part of the magma chamber. 

 Hence two-phase convection tends to slow down. The powerful radiation in the 

 crater finally causes the lava to freeze at the surface and a solid plug of greater 

 or less depth is formed. This plug must be removed before activity can be 

 resumed. 



The removal of the plug is not clue primarily to explosion. A volcano 

 may be dormant for scores of years, so that not even mild solfataric action 

 persists in the crater. In such a case the plug must be thick. On the average 



