xv.j FORBES' GLACIER DISCOVERIES. 517 



ing the magnitude of the change could be rightly pre- 

 dicted. In fact the temperature of freezing must be 

 lowered by ^th of a degree Reaumur for each additional 

 atmosphere or pressure of 15lbs. on the square inch. 

 The brother of the first named, William Thomson, 

 verified by experiments this result to its numerical 

 details by compressing in a very strong vessel a mixture 

 of ice and water, which was found to become colder 

 and colder as the pressure rose, and exactly as much as 

 the dynamical theory of heat requires. 



' Now, if by the application of pressure a mixture of 

 water and ice becomes colder without heat having been 

 withdrawn from it, this can only happen by some free 

 heat becoming latent, in other words, by some of the ice 

 in the mixture melting, and in this lies the reason that 

 mechanical pressure produces a change in the freezing 

 point. Ice occupies more space than the water from which 

 it is produced. When water freezes in a closed vessel, 

 it bursts not merely glass flasks, but even iron bombshells. 

 Thus, therefore, because in the compressed mass of ice and 

 water some ice melts, the volume of the whole becomes 

 less, and thus the mass yields more to the pressure than 

 it could have done without this change of freezing point. 

 Here mechanical pressure, as happens in the majority of 

 cases of interaction of different natural forces, favours 

 the production of the change, melting, which is favoural >le 

 t> the development of its own action. 



' In W. Thomson's experiment, water and ice were shut 

 u p together in a closed vessel from which nothing could 

 escape. The situation- is somewhat different, when, u 

 is the case in a glacier, the water produced between 

 pieces of ice pressed together can get out through 

 cracks. '[In- ice indeed then is pressed, but not the water, 

 which escapes. The pressed iee becomes then colder, 

 in cMii-r.jUence of the lowering of its freezing point l.y 

 pressure, but the freezing point of the water, which is 

 not compressed, is n<t lowered. In the>e eiivmnsiances 

 we have ice mlder than x-ro in contact, with water at 

 the temperature gem The consc|uen-,- of this is, that 



