WATER ANALYSES 593 



It is not improbable that most of the alkalies, lime, and alumina have 

 resulted from the action of the acid liquid on the glass tubing, but it is 

 of the greatest importance to establish the fact that the entire quantity 

 of gas pumped through the "wash bottle" yielded no more than 0.4 gram 

 of chlorine. If this be calculated in the form of gas, it will correspond 

 at most to 0.02 per cent, assuming that approximately 1,000 liters of gas 

 were drawn into the tubes. Fluorine seems to be present in about twice 

 this quantity, but in no sense can these halogens be regarded as forming 

 more than a very minor part of the crater exhalation. 



Rare Gases 



In the progress of the analyses, after all the active gases had been re- 

 moved from the several tubes analyzed, there remained an inactive residue 

 which, of course, consisted mainly of nitrogen, but which might be sup- 

 posed to contain traces of argon, helium, or other of the rare inert gases, 

 should any such chance to have been present in the volcano emanation. 

 For the determination or detection of these several of the residues were 

 brought together in a spark tube and exposed for several hours, in the 

 presence of oxygen, to an alternating current spark discharge of consider- 

 able intensity. When the volume of residual gas could no longer be dimin- 

 ished by this means, there remained a final residue of 75 cubic centimeters 

 of gas, which was forwarded to Prof. R. W. Wood, of Johns Hopkins, who 

 very kindly offered 15 to make a spectroscopic examination of it for traces 

 of the rare gases. The search yielded a decisive negative result. No lines 

 characteristic of any of the rare gases could be found with the spectre-* 

 scope, nor did exposure to charcoal at the temperature of liquid air leave 

 any residue whatever. The gas examined was, therefore, all nitrogen. Sub- 

 sequently the residues (15 cubic centimeters) from another group of the 

 tubes were treated in the same way and forwarded to Professor Wood, 

 who was again able -to detect nothing but pure nitrogen. It appears to be 

 definitely established, therefore, that the gases collected from Halemau- 

 mau in May contain nitrogen but no argon. This affords a most desirable 

 confirmation of our belief that the volcano gases were successfully col- 

 lected before they had come in contact with atmospheric air at all, and 

 were therefore entirely uncontaminated either by reaction or admixture 

 with it. It also offers support to the view that the volcanic nitrogen is 



16 We desire to take this opportunity to thank Professor Wood for courteously offering 

 to examine these gas residues. The Geophysical Laboratory possesses neither the equip- 

 ment nor the special experience pegessary to undertake a spectroscopic study of this 

 critical character. 



