462 DAY AND shepherd: water and magmatic gases 



water in excess of the quantity required to saturate the atmos- 

 phere and any quantity smaller than this would evaporate freely 

 in the comparatively dry air at 1300 meters elevation. Further- 

 more, Brun's observation that the exposure of a dew-point hygrom- 

 eter within the cloud showed less moisture present than in the 

 clear air outside, finds equally ready explanation in the fact that 

 the cloud has been shown to carry SOo and SO3, and by conse- 

 quence sulfuric acid, which are excellent dehydrating agents. 

 A cloud charged with drying agents like these must, a priori, be 

 found drier than the surrounding air which is not provided with 

 siich drying agents. In confirmation of this we find that the 

 decomposition of the surface lava, which is continually exposed 

 to these gases, is everywhere a sulfuric acid decomposition, the 

 decomposition products being sulfates (alum, gypsum) together 

 with considerable quantities of free sulfur. 



(4) If the total quantity of gas drawn into our tube system be 

 estimated at about 1000 liters, then the quantity of chlorine is 

 less than 0.02 per cent. In confirmation of this low chlorine 

 content, analyses of the lava near the volcano which has been 

 constantly exposed to the smoke cloud for 20 years or more, 

 showed no test for chlorine in a 2 gram sample.^ 



(5) It is of the highest importance to note further that the 

 nitrogen exhaled by the volcano contains no argon. Bearing in 

 mind that argon is always present in the atmosphere in measur- 

 able quantities, and that it enters into no chemical combination 

 whatever, it would seem to be a necessary conclusion that the 

 atmosphere does not contribute to the gases which are given off 

 by the lava. If atmospheric ingredients were present in the 

 lava, then surely the argon must be given off again. This must 

 apply to all atmospheric ingredients including water, whether 

 originally gaseous or condensed, for if meteoric water were to 



2 Brun appears to have tested for chlorine with a silver nitrate solution in an 

 atmosphere which is shown above to contain S, SO2 and SO3, and notes that it 

 immediately becomes clouded, but mentions no test to ascertain whether it was 

 the chloride or the sulphite which was thus precipitated. Similarly, he nowhere 

 offers a chemical analysis of the particular gases which he collected in tubes at 

 Kilauea, but contents himself with presenting two analyses of other gases pumped 

 from solid lava fragments reheated in vacuo some months afterward. (L'Exha- 

 laison Volcanique, p. 115.) 



