WATER AND VOLCANIC ACTIVITY DAY AND SHEPHERD. 301 



VARIATIONS IN THE COMPOSITION OF THE t.ASES. 



In full accord with the positive conclusion that these particular gases 

 can not exist together in stable equilibrium at the temperature at which 

 they are found, but are in process of active reaction, the record of the 

 analyses shows their composition to vary from one tube to another. 

 Successive tubes collected from the volcano at the same time (Table 

 2, p. 289) do not show the individual gases to be present in the same 

 proportions, but rather in proportions which change with every 

 bubble which bursts from the liquid basin. 



VARIATIONS OF LAVA TEMPERATURE RESULTING FROM THE GAS REACTIONS. 



Further confirmation of tlie same conclusion is found in the observa- 

 tion, already noted elsewhere, that when the gases given off by the 

 lava increase in quantity (pi. 11), the quantity of lava (lava level in 

 the basin) remaining the same, its temperature increases, and, con- 

 versely, when less gas is discharged through the lava this temperature 

 diminishes again. During the period of our visit in 1912 this change 

 in the temperature at the surface of the active lava in the basin 

 amounted in maximum to 115° (June 13, 1912, 1,070° ; July 6, 1912, 

 1,185°), and is therefore much greater than could be explained in so 

 large a basin by fortuitous conditions of measurement. This absence 

 of equilibrium and consequent variability of composition is also in 

 accord with the observation of Perret and others at Vesuvius, that 

 the relative proportions of the gases vary greatly with the condition 

 of the crater. 



EXPLOSIVE LAVAS (BRUN). 



From the same viewpoint the laboratory observations of Brun on 

 " live " or " explosive " lavas and, in contradistinction, " dead " lavas 

 acquire new and rational significance. In all the experience of the 

 (xeophysical Laboratory with the thermal study of silicates, we have 

 found no natural rocks or minerals which did not set free gases in 

 considerable quantity when heated in the laboratory to a temper- 

 ature high enougli to melt their chief constituents. Chamberlin,^ in 

 liis elaborate series of analyses of the gases contained in rocks, seems 

 to have had the same experience. If these studies together represent 

 sufficient breadth of experience to justify a sweeping conclusion, then 

 there are no " dead " rocks, meaning thereby igneous rocks, which no 

 longer release original volatile ingredients when heated to melting. 

 On the other hand, if we admit the nearly or quite universal distribu- 

 tion of gaseous ingredients in igneous rocks, but suppose these gases 

 were in equilibrium with each other throughout the solidification 



1 R. T. Chamberlin : The gases in rocks. PubUcations of the Carntgle Institution of 

 Washington, No. 106. 



