172 



rature of 2 R. His measurements had reference to observed 

 temperatures of the ice varying from 2'3 R. to 22 R. After 

 applying the requisite corrections, it resulted from them that the 

 coefficient of expansion of ice is for 1 R. 



OOOOG46G, 



which is nearly twice as great as the coefficient of dilatation of lead, 

 and more than twice as great as that of any other solid. 



We do not know the modulus of elasticity of ice, or the pressure 

 under which it disintegrates. 



If it were as elastic as slate and did not resist crushing more than 

 hard brick, a block of it placed wit]: its ends between two immoveable 

 obstacles, would crumble when its temperature was raised one degree 

 of Fahrenheit. It is its great dilatability which gives to ice this ten- 

 dency to disintegrate, when, not being free to dilate, its temperature 

 is raised*, even so slightly as this. 



If the block of ice experimented on by Schumacher had been 

 placed upon a plank inclined at the same angle as that used in the 

 experiment with the lead was, and if its under side had been coated 

 with lead-foil so as to give it the same friction on the plank as the 

 lead had, then, under the same variations of temperature as the lead 

 experienced, it could not but have descended as the lead did, but 

 twice as fast, because its dilatability is twice as great. 



We may conceive such a block of ice to be made up of thin plates 

 parallel to its upper surface, such as plates of glass would be, if glass 

 were as dilatable as ice and as friable, and if it possessed that pro- 

 perty of passing from a disintegrated into a solid state, which in ice 

 is called regelation. If we put the adherence of these plates to one 

 another in the place of friction, and conceive the variations of external 

 temperature (or the effects of solar radiation) to reach them in suc- 

 cession, each one being dilated or contracted independently of the 

 rest, then each would descend by a motion proper to itself, and also by 

 reason of the descents of those subjacent to it. The extremities of 

 the plates would under these circumstances overlap, and the descent 

 of each, proper to itself, would be increased by the overlappings of 

 those beneath it. 



* Agassiz describes a disintegration of the transparent ice of the blue bands 

 of glaciers when laid bare, which appears to be due to its expansion. Bulletin Un. 

 de Geneve, vol. xliv. p. 142. 



