GEOPHYSICAL LABORATORY. 129 



of the expedition sent to the Katmai region in 1919 by the National Geo- 

 graphic Society, in collaboration with the Geophysical Laboratory. It 

 contains a general description of the fumarolic area of the Katmai region, 

 and also the distinctive characteristics of the fumaroles located in the 

 pumiceous deposit of the Valley of Ten Thousand Smokes and its branches. 

 A detailed account is also given of the field and laboratory work, together 

 with deductions based thereon. 



The highest fumarole temperatures (650° maximum) followed a discon- 

 tinuous zone about Baked and Broken Mountains, with a salient extending 

 out into the main valley from Broken Mountain. The borders of the fuma- 

 rolic area were usually characterized by low temperatures. A perusal of 

 all the temperature data obtained by the expeditions of 1917, 1918, and 

 1919 strongly indicates that the fumarolic activity is declining. This is 

 not at all surprising when one considers the great cooling effect of the drain- 

 age water from the mountain slopes that finds its way into the pumice 

 and reappears as steam issuing from the fumaroles, or as vapor rising from 

 the valley floor, as well as in the hot streams that drain the valley. The fact 

 that great fumarolic activity and unusually high temperatures still persisted 

 in a number of the fumaroles 7 years after the eruption is much more sur- 

 prising. It shows definitely that the heat given off by the fumaroles must 

 be derived from a deeper source than the ejected pumice. 



Steam is the predominant constituent in all the gases examined, and varies 

 from 99.45 to 99.85 per cent by volume. The most important of the other 

 gases are HC1, C0 2 , H 2 S, N 2 , HF, and sometimes CH 4 . Minor constituents 

 are 2 , CO, A, and NH 3 . Ammonia generally occurred in the form of chloride 

 and probably fluoride. Sulphur dioxide may occur in some places, but was not 

 detected. Judging by the incrustations, sulphur must be a constituent of 

 the gases and, more rarely, the sulphide of arsenic. 



Samples of soluble gases collected by Shipley (1918) and examined in the 

 Geophysical Laboratory indicate that the percentage of HC1 reaches, on the 

 average, the same order of magnitude as C0 2 and HF, and comparable 

 with that of H 2 S and N 2 . This is borne out by the relatively large amounts 

 of fluorine found as fluoride in the incrustations and is unique in the annals 

 of volcanology. 



The ratio of argon to nitrogen is practically the same in the gases as it is 

 in the atmosphere, and the inference is that they are both of atmospheric 

 origin. If so, the associated oxygen has been largely absorbed. The amount 

 of heat generated in the oxidation of the volcanic gases by this small volume 

 of oxygen can be only an insignificant portion of the total heat given out by 

 the fumaroles. 



The water-vapor in the gases is largely derived from surface water, as 

 appears from the following observations: (1) The quantity of water- vapor in 

 the gases varies considerably and the variations appear to be closely related 

 to local variations in drainage in nearly all cases. (2) The porous nature of 

 the pumiceous deposit permits the absorption of a great amount of the drain- 

 age from the mountain slopes. A calculation based on annual precipitation 

 and on the amount of water carried out by the streams shows that much of 

 the drainage does not reach the streams and is in all probability returned to 

 the atmosphere as vapor through the fumaroles. This is also borne out by 



