256 THE NEW WORLD OF SCIENCE 



the still molten interior upon the already solid white-hot jacket 

 of steel which encases it, becomes thus richer in carbon, phos- 

 phorus, and sulphur than the preceding layer, so that the con- 

 tent of these elements increases progressively from the skin 

 to the axis of the completely solidified ingot. 



But this is not the worst of it. The steel solidifies not in 

 successive layers like the leaves of a gigantic onion, but rather 

 in great columnar or pine-tree crystals protruding out at any 

 given moment into the still molten interior. As solidification 

 proceeds these trees grow not only at their tips but also at the 

 ends of their tree-like branches, which thus in time inter- 

 lace, and thus landlock part of the enriched mother liquor. 

 The result is that, when solidification is complete, the ingot 

 as a whole has a dendritic structure, with these elements con- 

 centrated in part between the trunks and branches of the pine 

 trees, and in part concentrated progressively towards the axis 

 of the ingot, or more strictly towards the last freezing part. 

 This structure may be likened to the veining of marble. In 

 each case the mass is substantially free from cavities, and even 

 from porosity, but it is coarsely heterogeneous. 



The carbon and phosphorus which are thus concentrated 

 embrittle the metal locally, giving rise to brittle regions scat- 

 tered through the mass, somewhat like brittle links in an other- 

 wise ductile chain, lessening the resistance of the whole to 

 shock. 



The main purpose of forging is to lessen this heterogeneous- 

 ness by a species of kneading which mixes up the various 

 parts, and in particular lessens the distances which diffusion 

 has to cover in order to give uniformity. Kneading thus being 

 a good thing, give us plenty of it. It was most readily given 

 by reducing the cross section of the ingot under the hydraulic 

 press, and simultaneously lengthening it. But in order that 

 this cross section should thus be reduced greatly it must ini- 

 tially be much greater than that of the finished piece, formerly 

 four times as great. In trade language, there was a reduction 

 of four to one. Even before the war many of us insisted 



