EVOLUTION AND MUTABILITY OF .MATTER., 273 



that a disc of copper could be welded to a disc of tin 

 by pressing them strongly one against the other. Up 

 to a certain distance from the surfaces of contact 

 a real alloy is formed; a layer of bronze of a certain 

 thickness unites the two metals, and this could not 

 take place did not the particles of both metals 

 mutually interpenetrate. 



§ 4. Internal Activity of Alloys. 



Structure of Alloys. — Metallic alloys have a re- 

 markable structure, which is essentially mobile, and 

 which we have only now begun to understand by the 

 aid of the microscope. Microscopical examination 

 justifies to a certain degree Coulomb's conjecture. 

 That illustrious physicist explained the physical pro- 

 perties of metals by imagining them to be formed of 

 two kinds of elements — integral particles, to which the 

 metal owes its elastic properties, and a cement which 

 binds the particles, and to which it owes its coherence. 

 M. Brillouin has also taken up this hypothesis of 

 duality of structure. The metal is supposed to be 

 formed of very small, isolated, crystalline grains, 

 embedded in an almost continuous network of viscous 

 matter. A more or less compact mass surrounding 

 more or less distinct crystals is the conception which 

 may be formed of an alloy, ' 



Changes of Structure produced by Deforming 

 Agencies. — It has been shown that profound changes 

 of crystalline structure can be produced by various 

 mechanical means, such as hammering, and the 

 stretching of metallic bars carried to the point of 

 rupture. Some of these changes are very slow, and 

 it is only after months and years that they are com- 



