WATER AND VOLCANIC ACTIVITY DAY AND SHEPHERD. 



291 



activity which we are studying. It follows directly from this that 

 the chemical analyses of volcano gases offered bj' Briin in support of 

 his conclusion that the exhalations fi'om Kilauea and other volcanoes 

 are anhydrous, also carry on their face the clear proof that his con- 

 clusion is untenable. Neither COo nor SOo can be associated with 

 free hydrogen at temperatures in the vicinity of 1,000° without the 

 formation of water. The two analyses offered by Brun of the gases 

 given off on reheating the Kilauea lava show them to be chiefly made 

 up of precisely these ingredients (loc. cit., p. 115) : 



Brim's analyses of gases ohtained from Kilauea lava ichen heated in vacuo. 



HCl Traces. 



SO2- 

 C0=- 

 CO- 



1.7 



50.8 



17. .'5 



28.5 



1.2 



CI (libre) 5.58 



HCl 1.71 



SO. 4. 72 



CO2 G9. 09 



CO -___ ^ [^ 11. 60 



CH4 0, 85 



H2 6. 10 



N2 0.3 



It is, furthermore, noticeable that the analyses here offered by Brun 

 as representative of the gases emanating from Halemaumau do not in 

 any way agree with the composition of the gas which we obtained 

 from liquid lava. He finds, for example, in one analysis more than 

 7 per cent of chlorine in one form or another where we find 0.02 per 

 cent or less. He obtained about 5 per cent of SO2, whereas SOo dur- 

 ing our visit was perhaps the most notable gas evolved from this 

 crater. 



THE RELATIVE QUANTITIES OF THE CONSTITUENT GASES. 



Leaving now the question of the identity of the gases discharged by 

 the laA'a at Halemaumau, we should perhaps turn for a moment to 

 the consideration of their relative quantity, which, as alread}'- inti- 

 mated, is not so well establislied by our samples because of the unex- 

 pected presence and condensation of the Avater in the collecting tubes. 



In addition to the presence of the iron tube, and its possible effect 

 on the total quantity of water and of free sulphur collected in our 

 tubes, which has been discussed above, our analyses of the gases 

 (p. 289) are subject to a second limitation which is at once obvious. 

 When the pumping had been completed, the collecting tubes each con- 

 tained a quantity of the condensed water, in which the fixed gases 

 are individually soluble in varying degree both during and after 

 cooling. There is also some reaction between the gases and water. 

 The long period which elapsed between the date of collecting the 

 gases and their analysis in Washington after the close of the field 

 season — ^nearly a year— gave opportunity for these readjustments to 

 proceed practically to completion. The SOj, for example, has gone 



