614 SUMMARY OF CURRENT RESEARCHES RELATING TO 



corrosion is considered to be an electrolytic action between the carbides 

 and. the carbon-free material (ferrite or solid solution) which together 

 constitute the structure of the steel. The carbides act as anodes of the 

 electrolytic couples, and are therefore preserved while the ferrite or solid 

 solution is attacked, and passes into solution in the corroding medium. 

 Photomicrographs of (1) steel containing 1-25 p.c. carbon, and (2) steel 

 containing 0'7;-5 p.c. carbon and 21 "5 p.c. tungsten, after corrosion are 

 given to illustrate this point. Each shows massive carbide intact. In 

 a similar way, pearlite does not corrode as a whole, but as a mixture of 

 ferrite and carbide, and only tlie ferrite is attacked. The disappearance 

 of the carbide from pearlite which usually occurs in corroding pearlitic 

 steels is due to mechanical loss. In the case of pure iron, consisting of 

 ferrite only, electrolytic action between the ferrite crystals and the inter- 

 crystalline amorphous cement is put forward as the chief cause of 

 corrosion. Here the amorphous material acts as kathode, and passes 

 into solution more rapidly than the ferrite. Evidence of this is shown 

 in a photomicrograph of pure iron after corrosion in sodium chloride 

 solution, showing accelerated action along the crystal boundaries. The 

 concentration of the solid solution is the controlling factor in deciding 

 the rate of corrosion of a steel, since the higher the concentration the 

 lower is the electromotive force set up by the solid solution in a corroding 

 medium. The properties of special steels are reviewed, and shown to 

 confirm these views. Elements like molybdenum, vanadium, and 

 tungsten, which form part of the carbide and do not enter into the solid 

 solution until a high percentage is present, exert little effect on the rate 

 of corrosion, while elements such as chromium and nickel, which enter 

 into solid solution from the first, retard corrosion. 



Surface Tension Effects in Metals.* — A theory, based upon the 

 existence of films of metallic amorphous material or cement between the 

 crystalline grains of metals, is elaborated by F. C. Thompson. This 

 amorphous material is regarded as an under-cooled liquid, and the deduc- 

 tion made that surface tension forces operate between crystals separated 

 by thin films of such material. The great and unexpected strength of 

 the crystal junctions in ductile metals is associated with these surface 

 tension forces. The growth of crystal grains at high temperatures 

 during annealing is explained by the endeavour of these surface forces 

 within the mass of metal to reduce the area of the intercrystalline 

 boundaries. The mechanical and other properties of motals are con- 

 sidered and interpreted in the light of these ideas. The elastic limit of 

 a metal is reached when the attraction of the surface tension forces over 

 a given area of a specimen under test is just overcome. The larger 

 area over which surface tension attractions can occur provide an explana- 

 tion of the well-known fact that a fine crystal structure possesses a 

 higher elastic limit than a coarse crystal structure. The equation 



2 T 



E = — T- (where E = elastic limit, T = the surface tension of the 



amorphous material, .c? = thickness of the film) is deduced. Accurate 

 * Journ. Iron and Steel Institute, xciii. (1916) pp. 155-92 (7 figs.). 



