O'Reilly — On Gaseous Products of Great Eruptions. 27 



tions made at sea must be limited to a relatively small number of 

 points or zones — that the polar regions must be perhaps for ever closed 

 to observation. Again, the constitution of the upper parts of the 

 atmosphere, above 7 miles = 11,000 m. height, are quite unknown to 

 us, and will probably ever remain so, since no living being can exist 

 at that height. Lastly, that account must be taken of the porosity 

 of the surface rocks and soil, and of the ocean, which can absorb 

 and retain quantities of gases, variable relatively to temperature 

 and pressure. Thus supposing the volume of the atmosphere to be 

 actually doubled by volcanic emission at a given moment, it does 

 not at all follow that the barometer would show that increase of 

 volume in totality and at once, since the pressure on the surface of 

 the earth would cause a certain portion to be taken up by the soil 

 and rocks, and a certain other part by the water. In this respect, 

 indeed, we should perhaps look to the ocean as a far more reliable 

 witness to variations of volume in our atmosphere ; and were the 

 analyses of ocean water sufficiently numerous, both as regards 

 local distribution and depths, and extended over a sufficiently long 

 period of time, they would manifest by changes in the quantities 

 of contained gases much more accurately, and with much more 

 chances of sound comparison, variations in the volume of the 

 atmosphere, than would barometric measurement. Here there 

 is room to remark that the quantity of gases contained in the 

 ocean and other waters must be in intimate relation with their 

 organic life, and that, consequently, the greater or lesser abun- 

 dance of fossils in certain formations must bear some relation to 

 the quantity and nature of the gases contained in the sea in which 

 they were deposited, and these gases were in relation to the volume 

 and constitution of the then atmosphere. Thus we have probably, 

 in the fossils of the different formations, real measures of the 

 atmospheres, corresponding to the periods of their deposition. 



Were it possible to determine directly the gases given forth 

 from any one of the existing active volcanoes, no more valuable 

 scientific work could be attempted, but the difficulties are evidently 

 immense, if not insurmountable, unless in the case of some small 

 volcanic cones, where it might be possible to make such an attempt. 

 But these difficulties only enhance the value of all measurements 

 and determinations of the emissions of gaseous hot springs and 



