304 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1913. 



supported by measured data, that the rainfall at Halemaumau is 

 even smaller than that recorded at the Volcano House, for at an equal 

 distance to leeward of Halemaumau the country is desert and prac- 

 tically without rainfall. The present crater lies in the midst of this 

 transition zone from 78 inches to zero. Be that as it may, there is 

 a more or less abundant rainfall at Kilauea, even though the aggre- 

 gate amount is much smaller than has hitherto been supposed. 



There is a further fact of observation which may be cited in this 

 connection. Wells have been bored on the sugar plantations at ele- 

 vations up to 2,000 feet on Hawaii and on the other islands. In 

 these borings water is invariably met with (so far as we were able 

 to learn) at sea level only. The water is ordinarily fresh, but a 

 heavy draft on it always has the effect of increasing its salt content, 

 and some of the wells have been permanently ruined for irrigation 

 purposes by this means. 



So far as the conditions surrounding this volcano are concerned, 

 therefore, Avater in some form would seem to be very widely dis- 

 tributed except on the high mountains, and as freely available as 

 silica for active participation in any form of volcanic activity. In 

 the present preliminary survey of the situation it tlierefore appears 

 as if any attempt to assign the water found in the lava to one or other 

 of these three conceivable sources, or, perhaps better, to justify any 

 specific distribution of it among the three conceivable sources, must 

 be based on assumptions of a somewhat arbitrary and hypothetical 

 character. Nevertheless, there are some indications which inevitably 

 give direction to the probabilities which an individual observer may 

 fix on. First and most important, in our opinion, is the fact that 

 the nitrogen found in the emanation is free from argon. It is plain 

 that if atmospheric water is to reach a hot lava column at a tem- 

 perature of 1,000° or higher it must do so as a gas, and therefore 

 on the same terms as other atmospheric gases. Argon is invariably 

 contained in the air in measurable quantity and forms no chemical 

 compounds. Whence it follows that if the gases of the atmosphere 

 had reached the liquid lava in any manner whatsoever the argon 

 would be released with the others, but no trace of argon was found. 



The second difficulty is to conceive a mechanism whereby atmos- 

 pheric or surface water of whatever origin (for example, the sea) 

 can make its way into a lava column or basin at a temperature of 

 1 .000° or more. The Daubree experiment, whereby water vapor was 

 found to make its way through 2 centimeters of sandstone against 

 an excess pressure within, though often quoted in this connection, 

 does not help us to a solution of it. The force which was active in 



