458 DAY AND shepherd: water and magmatic gases 



the smoke cloud (100 meters distant from the point of emer- 

 gence), and he observed further that a dew-point hygrometer 

 exposed within the cloud indicated a smaller moisture content 

 than in the clear air near-by. Some other observations were 

 offered pointing to the same conclusion, but the above citations 

 fairly serve to show the character of the observations which led 

 to the conclusion that water had no part in these exhalations. 



Jt has seemed to the writers, in considering this problem, that 

 evidence of this kind is unconvincing. Such evidence may serve 

 to prove that the cloud above Kilauea does not consist entirely 

 of steam, but it is a very different matter to conclude from it 

 tiiat no steam is emitted, particularly in view of the fact that the 

 extensive hygrometer measurements offered by Brun all show that 

 the observations were made in an unsaturated atmosphere, and 

 the further fact that the temperature of the gases at the moment 

 of emission was not far from 1200? The latter point seemed to 

 us of vital importance in any attempt to identify volcanic gases, 

 for nearly all the recorded analyses of the gases contained in 

 volcanic rocks^ include gases (e.g., CO or H2) which undergo im- 

 mediate alteration and lose their identity if released into the air 

 at such a temperature as 1200? We therefore undertook the 

 somewhat hazardous task of going io the bottom of the crater 

 itself and collecting the gases direct from the liquid lava before 

 they had come in contact with the air at all. 



The gases were collected under the following conditions. A 

 lava fountain broke thru the floor of the crater beside the lava 

 lake and by its own spattering quickly built for itself an en- 

 closing dike. When this dike had grown to a completely enclos- 

 ing dome, an ideal gas collector was provided by the volcano 

 itself; for this dome, because of the continual bursting of great gas 

 bubbles within, was lined with the same liquid lava from which 

 the gas was being released. From the narrow cracks in the 

 sides of the dome, sheets of pale blue flame could be seen burn- 

 ing at night, which indicated (1) an excess pressure within, and 

 in consequence (2) that the gases released from the liquid lava 

 first came in contact with the air upon emerging from the dome. 



2 R. T. Chamberlin, The gases in rocks, Publications of the Carnegie Institu- 

 tion of Washington, No. 106. 1908. 



