404: W. Hallock — Chemical Action between Solids. 



will then continue, provided the product (liquid or gas) escapes 

 easily and does not clog the operation. In very many cases 

 substances are found to give off a vapor below their melting 

 point, and it is natural to suppose that there is a film of that 

 vapor over the surface of the body, as there is a layer of satu- 

 rated air over water. The mechanical theory of the composition 

 of matter lends plausibility to the above suggestion. If these 

 considerations are correct they foretell the regelation of sub- 

 stances like camphor and ice, without any pressure whatever. 

 That loose pieces of camphor will become welded together by 

 simple contact is well known. The operation appears to me 

 thus : In an irregular mass of camphor in an atmosphere of 

 camphor vapor, there is a constant interchange of state for 

 the molecules at the surfaces of the solid, molecules previ- 

 ously solid are getting too far off and becoming gas, and 

 molecules previously gas are beating upon the solid and staying 

 there, thus the state of equilibrium is when, as a whole, there 

 are as many molecules which fly off and become gas as fly on and 

 become solid. On a projecting point of the solid the chances 

 are in favor of more flying off than on, in a reentrant angle 

 the reverse is true. Theoretically, then, the piece ought ulti- 

 mately to become a sphere, not only by the rounding down of 

 the corners, but by the building up of the flat or reentrant 

 sides. That the corners do round off all know. If this is all 

 true we only need to bring the two pieces together and con- 

 sider them as one and the crack between them as a reentrant 

 angle, and the union is brought about as above indicated. If 

 in the above the word liquid be substituted for vapor or gas, 

 the explanation will apply to the regelation of ice in water 

 at 0° C. 



We may go even further and predict a .uniting without act- 

 ual contact and this prediction has been experimentally demon- 

 strated in the case of ice and water. A large rough block of 

 ice (about 15 lbs.) was sawed nearly in two, the slit washed out 

 and all the fine pieces removed. In this way it was possible to 

 hold two plane surfaces of ice parallel and near each other (1 to 

 2 mm ) without danger of actual contact. Into the outer edge of 

 the saw-cut a cotton wick was pressed, thus isolating the space 

 between the faces from the outside and preventing any currents 

 from circulating through the crack. The whole block was then 

 placed in water at zero and enclosed in non-conducting cases 

 and left for 25 to 30 hours. This experiment was tried three 

 times and each time a freezing across the space had taken place. 

 The whole space was not filled, but in numerous places notably 

 along just inside the wicking and up from the bottom of the 

 cut. !No doubt the regelation would have gone further if the 

 experiment could have been continued longer. The melting of 



