-1859] WRITINGS OF JOSEPH HENRY. 



201 



the greatest amount of space in a given surface. All other 

 conditions being equal, we should choose this form of ex- 

 cavation for preserving ice ; but on account of the difficulty 

 of lining a pit of this shape, we may select the next most 

 economical form, which is the cylindrical. It is scarcely 

 necessary to mention in this connection the fact that, in 

 order to succeed in preserving ice, it should be well pro- 

 tected from the surrounding earth and air by strata of non- 

 conducting materials, such as straw, powdered charcoal, or 

 saw-dust, the greater the thickness of which, the better the 

 purpose in view will be answered. The house should also 

 (as an additional precaution) be shaded above by trees, and 

 have the cover painted white, to reflect back the more intense 

 rays which may reach it indirectly. Morever the ice should 

 not be suffered to rest upon the bare ground below, but on 

 double floors, between which a non-conducting substance 

 is placed, communicating by holes with a deep pit or drain 

 through which the water from the melted ice may percolate. 

 We have stated that water at 39°*1 begins to expand, and 

 that this expansion increases until solidification takes place. 

 The force exerted by this expansion is immensely great, 

 being sufficient to burst a cannon or to cause water to pass 

 in the form of a fine frost through the pores of solid metal. 

 When however this expansion is opposed by a sufficient ex- 

 ternal pressure the water is not converted into a solid at 

 thirty-two degrees, but assumes this condition at a lower 

 temperature; a piece of ice therefore at thirty-two degrees 

 subjected to a great pressure ought to be converted into a 

 liquid ; and this may serve to explain a fact frequently 

 noticed, that pieces of ice thrown upon each other adhere at 

 the points of contact — the percussion changing these surfaces 

 from a solid to a liquid, which immediately afterwards solid- 

 ifies again. But this cause is scarcely sufficient to explain 

 the very remarkable fact that if two lumps of ice be placed so 

 as to present two flat surfaces and these be pressed together 

 they will unite as one mass; and this will take place even 

 in hot water while the external surface is rapidly melting. 

 The pressure necessary to bring them into contact would no 



