on the principal Geijsirs of Iceland. 401 



The corresponding values of/ are as follow : — 



Third and fourth experiment . . . /=136"'151 

 Maximum temperature at the bottom 127 



Difference . . 9"-151 



Fifth experiment /=]35°'398 



Minimum temperature at the bottom 122°'5 



Difference . . 12°-898 



First experiment /=135°-31 



Temperature at the bottom . . . 122°'5 



Difference . . 12°-81 



Second experiment /=136°"28 



Temperature at the bottom . . . 123°-60 



Difference . . 12°-68 



Thus, at the point to which the lower thermometer can 

 penetrate, the water does not attain the temperature of ebulli- 

 tion which it should have under the pressure to which it is 

 subjected ; but it approaches that temperature in proportion 

 as a great eruption is near. We may thence conclude that 

 the centre of heat which raises the temperature of the water 

 of the Geysir is situated at a certain distance below the pre- 

 sumed bottom of the central well, and we may deduce from 

 this a plausible explanation of the manner in which the erup- 

 tions are determined. 



Let us in fact suppose that the column of water in the cen- 

 tral basin communicates, by a long and sinuous channel, with 

 the space, be it what it may, which receives the direct action 

 of the subteri-anean heat : after an eruption during which a 

 projection of a great quantity of water and vapour has taken 

 place, the lower parts of the liquid mass are cooled, and the 

 steam which is formed in the reservoir submitted to the action 

 of the heat has a less tension than that at which the weight of 

 the central column and that of the atmosphere are in equili- 

 brium ; this vapour, as it forms, becomes condensed in con- 

 tact with the water which fills the sinuous channel, and it im- 

 parts to that water its latent heat. The increase of tempera- 

 ture of the water of the channel is transmitted by degrees to the 

 lower part of the central column where the thermometer 

 can reach ; but this increase is retarded by the atmospheric 

 air and the other gases which accompany the vapour; how- 

 ever, at the lapse of a shorter or longer time, the water of the 

 channel must boil, and the steam which continues to form can- 

 not longer condense there; this vapour must therefore accu- 

 mulate, and acquire a gradually increasing tension, until this 

 /-•////. Matr. S. 3. Vol. 30. No. 203. June 1 847. 2 E 



