xv.J FORBES' GLACIER DISCOVERIES. 501 



bed of uneven rock, and through a channel so sinuous 

 and irregular that a glacier is often embayed in a valley 

 whence it can only escape by an aperture of half its 

 actual width ? On all mechanical principles we answer 

 that it is impossible. We may add, that many small 

 glaciers are seen to rest upon slopes of from 20 to 30 

 without taking an accelerated motion ; and this is con- 

 formable to the known laws of friction. It is known, 

 for instance, to architects that hewn stones, finely dressed 

 with plane surfaces, will not slide over one another until 

 the slope exceeds 30. 



' 2. The dilatation theory is founded on i mistake as to 

 physical fact. I am sorry to put it in this way, but it 

 is unavoidable ; and the respectable author of the only 

 intelligible or precise account of the theory will, I hope, 



use me for pointing it out. 



"'The maximum temperature which a glacier can 

 have," observes M. de Charpentier, "is Centigrade, or 

 32Fahr., and the water in its fissures is kept liquid only 

 by the small quantity of heat which reaches it by the 

 surface water and by the surrounding air. Take away 

 this sole cause of heat, i.e., let the surface be frozen, and 

 the water in the ice must congeal." Now, this is a pure 

 fallacy ; for the fact of the latent heat of water is en- 

 tirely overlooked. The latent heat of water expresses 

 the fact, that when that fluid is reduced to 32 it does 

 not immediately solidify, but that the abstraction, not of 

 " a small quantity of heat," but a very large quantity 

 indeed, is necessary to convert the water at 32 into ice 

 at 32. Not a great deal less heat must be abstracted 

 than the difference of the heat of boiling water and that 

 at common temperatures. The fallacy, then, consists in 

 : Admitting all the premises, the ice at 32 (it is 

 allowed that in summer, during the period of infiltra- 

 tion, it cannot be lower) is traversed by fissures extend- 

 ing to a great depth (for otherwise the dilatation would 

 be only 8Uj< rii i.ih, filled with surface water at 32. 

 Night approaches, and the surfaee fi ,md water 



ceases to be conveyed to the interior. Then, says the 



