zero is not entirely established; but we may suppose, with the author of the Traité de Thermométrie, that in glass, as in alloys, are to be found compounds which vary according to the temperature. At each temperature glass tends to assume a determinate composition and a corresponding state of equilibrium; but the previous temperature to which it has been subjected clearly has an influence on the rapidity with which it attains its state of repose. The effect of variation is more marked when we observe glass of more complicated composition. We can understand that those which contain comparable quantities of the two alkalies, soda and potash, may be more subject to these modifications than those having a more simple composition based on a single alkali.
Effects of Annealing.—A piece of brass wire that has been drawn and then heated is the scene of certain very remarkable internal changes, and these have been only recently recognized. The violent treatment of the metallic thread in forcing it through the hole in the die has crushed the crystalline particles; the interior state of the wire is that of broken crystals embedded in a granular mass. Heating changes all that. The crystals separate, repair themselves, and are built up again; they are then hard, geometrical bodies, in an amorphous, relatively soft and plastic mass; their number keeps on increasing; equilibrium is not established until the entire mass is crystallized. We may imagine how many displacements, enormous when compared with their dimensions, the molecules have to undergo when passing through the resisting mass, and arranging themselves in definite places in the crystalline structures.
