270 Reviews — Dr. A. Brim's Volcanic Researches. 



defined. For commoulv occurring types the following values have 

 been obtained: 826° C., 861° C, 877°C, 883°C. ) 886°C., 888°C, 896 C C. 



It is thus tolerably certain that tbe temperature of the Krakatoa 

 magma in the paroxysm of 1883 was approximately 880° C. 



A contrast which presents itself in the behaviour of rocks in 

 general wben heated allows Bran to divide them into two classes, 

 the active and the dead. The active rocks include onh T comparatively 

 recent, unoxidized, volcanic products; all other rocks are dead. 



Active basic rocks when melted liberate gas and volatile salts, which 

 latter condense to form white fumes. The liberation of gas and 

 vapour often swells the molten mass to such an extent that it over- 

 flows the crucible, furnishing a miniature fuming lava-flow. By 

 cooling and reheating it is sometimes possible to repeat this experiment 

 several times. A block of the 1904 lava of Vesuvius was still 'active' 

 after five such heatings. 



Active acid roclcs are much more violent when heated and give rise 

 to veritable explosions. 



Dead roclcs, such as schist, granite, or gabbro, behave quite 

 differently. On heating they give off gas, and then at a higher 

 temperature melt quietly. There is no trace of the magnificent 

 expansion characteristic of the active rocks. 



The tremendous pressure developed on heating obsidian to its 

 explosion temperature can be illustrated by performing the experiment 

 on a piece of obsidian enclosed in a box of refractory steel. "When 

 the explosion temperature is reached the box is burst and the magma 

 is extruded in a manner which recalls the formation of the spines of 

 Montague Pelee and Merapi (Java). 



As a general rule a magma, at any rate when once it has been 

 extruded at the surface, is markedly heterogeneous in regard to its gas- 

 producing capacity. On a field of lava there are comparatively few 

 points from which fumes issue in really great abundance. Again, 

 certain thin bands in obsidian may flash up into froth on heating 

 while other adjoining layers remain glassy. 



The volume of gas liberated in exploding an obsidian can be 

 ascertained approximately from a comparison of the density of the 

 obsidian with that of the resultant pumice. To take two striking 

 examples : Lipari obsidian of density 2 - 358 corresponds with a pumice 

 of density 0*224 ; and Krakatoa obsidian of density 2 - 36 corresponds 

 with a pumice of density - 416. In the case of Krakatoa this means 

 that 1 kilog. of the obsidian must yield 1,577 c.c. of gas at 880° GV 

 (the explosion temperature), or 373 c.c. reckoned at 0° C. 



Here Bran turns aside to give an account of the collection, 

 extraction and analysis of the various volcanic gases. 



Collection in the field. There are two main methods. "Wherever 

 possible the gases are pumped through a system of tubes, which may 

 be made of lead where the temperature is low or of glass where it 

 is high. A rubber pump is used, and the gases, once they are cooled 

 down, are collected over mercury, or, as the case may be, examined in 

 a hygrometric chamber. The glass tubes employed are each a metre 

 long, and sometimes as many as forty of them have been connected 

 end to end, their joints wrapped in sheets of lead or aluminium bound 



