518 SUMMARY OF CURRENT RESEARCHES RELATING TO 



with phosphorus, the ferrite bands or ghosts frequently contain ex- 

 cessive quantities of sulphide inclusions. The presence of such inclu- 

 sions embedded in ferrite has given rise to the mistaken view that the 

 inclusions are the cause of the ghosts, having acted as centres of crystal- 

 lization for the ferrite. The true cause of ghosts is the tendency of 

 the carbon, when in solution, to leave the high-phosphorus regions and 

 to concentrate in the low-phosphorus parts, during cooling. At tem- 

 peratures little below the solidus, the point of incipient fusion, phosphorus 

 diffuses and in time becomes uniformly distributed in steel, but at lower 

 temperatures diffusion is very slow. The unequal distribution of 

 phosphorus also retards somewhat the diffusion of carbon in steel when 

 heated above the critical range. When a steel, heated above the critical 

 range, has its carbon uniformly diffused, but its phosphorus varying in 

 concentration from point to point, ferrite first forms on cooling in the 

 regions of maximum phosphorus content. Thus ghosts may reappear 

 after repeated heating to high temperatures. 



The heterogeneous character of wrought irons, through unequal dis- 

 tribution of phosphorus, has been indicated by the cupric reagent 



The paper is illustrated with numerous photomicrographs, some of 

 which are reproductions in colour of colour-photomicrographs of heat- 

 tinted specimens. 



Detection of Burning in Steel.* — The heating of steel to such a 

 high temperature that it cannot be rolled or forged without breaking up is 

 technically known as " burning," and is coincident with incipient fusion. 

 J. E. Stead has previously shown that when incipient fusion occurs the 

 portions which melt first are rich in phosphorus, and persist after cooling 

 as globules or envelopes round the crystals. He has now applied the 

 new cupric reagent (see preceding abstract) to the detection of these 

 minute high-phosphorus segregates, the presence of which is an unfailing- 

 indication that the steel has been " burnt." A steel plate containing a 

 layer of segregate (carbon 0*5, phosphorus 0' 13) sandwiched between 

 layers of purer material (carbon ■ 3, phosphorus ■ 07 p.c.) was heated 

 at one end until that end melted. After cooling, sections were cut at 

 parts heated to different temperatures. The globules and envelopes first 

 appeared in the central high -phosphorus region, showing that the less 

 pure material became burnt at a lower temperature than the outer and 

 more pure layers. Burning is not in any way an effect of oxidation, but 

 may cause the formation of cracks which oxidize very rapidly. 



Ancient Iron from Ceylon. f — W. Rosenhaiu has examined a small 

 fragment of a link from a chain which had hung in the open air in 

 Ceylon for probably more than two thousand years. Polished sec- 

 tions showed included slag much resembling that present in common 

 wrought-iron, but drawn out to a less extent. All etching reagents tried 

 were found to be slower in their action on the Ceylon iron than on 

 modern Staffordshire wrought-iron, suggesting a lesser corrodibility of 

 the ancient iron. The etched sections presented the usual features of 



* Journ. Iron and Steel. Inst, xci. (1915, 1) pp. 398-408 (17 figs.). 



f Nat. Phys. Lab. Collected Researches, xii. (1915) pp. 153-60 (5 figs.). 



