In carefully reviewing the various facts which I have been able to collect since the publication of my paper on the microscopical Structure of Crystals, I have been led to believe that many peculiarities in the structure of granitic rocks may be explained by a principle, which, as far as I am aware, has not hitherto attracted attention. It appears to me extremely probable that the critical point in the consolidation of granitic rocks is very closely connected with the critical temperature at which highly heated and compressed steam would condense into an equal volume of highly heated and expanded liquid water.

As I pointed out in the paper already mentioned, rocks melted under great pressure probably contain water either dissolved as a gas in a liquid, or in the state of a fused hydrate. On cooling down to a lower temperature, the crystallizing out of anhydrous minerals would almost necessarily set free this previously combined water. As long as the temperature was above the critical point it would necessarily remain more or less disseminated in the rock as highly compressed steam, but as soon as the temperature fell below the critical point, it would condense into highly expanded liquid water.