270 Transactions of the Society. 



isolated particles, and as these do not break up the continuity of 

 the ferrite to the same extent they are not seriously detrimental. 

 Visual examination proved that apart from these markedly 

 segregated positions the MnS was in small particles distributed 

 over the whole surfaces of the three specimens. 



After etching, the specimens from E. 288 became decidedly 

 interesting. The pearlite crystals (formed by the combination of 

 the FeaC and the MD3C), and also the silvery ferrite, could be 

 observed. As seen visually the MnS enclosures in the ferrite are 

 still very noticeable owing to the difference in colour, although they 

 do not appear in quite the same sharp contrast when photographed 

 after " etching." 



The position of a second photograph (fig. 7) in the web can be 

 seen on .the plan. Here the general " banded " structure of the 

 ferrite predominates, and although it can be expected that there 

 should be some tendency in this direction, owing to the severity of 

 the rolling-stresses necessary to reduce the steel down to form this 

 comparatively thin component of the rail, this is too marked to be 

 attributable to rolling-stresses alone, and is due to segregation of 

 phosphorus towards the crystal boundaries whilst the steel i& 

 undergoing the first crystallization to form the ingot. The ferrite^ 

 rich in phosphorus, apparently resists the entry of carbon, and so- 

 would appear to confine the pearlite to areas free from phosphorus. 

 When this " banded " structure, or " ghost lines " as termed by 

 some, is formed, it is extremely persistent, and is difficult to remove 

 even by prolonged annealing. It is a frequent cause of weakness 

 in steel plates and many failures are attributed to it. 



The next photograph (fig. 8) is of a character not frequently 

 met with. The magnification was 50 diameters, and so includes the 

 structures in the segregation and also the microstructure above that 

 area, whilst an almost continuous line of ferrite divides the two. 

 The irregularity of the crystals above the ferrite line afford unmis- 

 takable evidence of incorrect thermal treatment, although the lower 

 pearlite crystals are of greater uniformity. 



The sixth photograph (fig. 9) was taken midway between the 

 depression left by the first physical test and the top of the specimen. 

 The large size of the pearlite crystals will be noted, also the fine 

 needle-like lines of ferrite in the pearlite, from which it becomes 

 apparent that the nearer we approach the top of the rail the thermal 

 treatment has produced worse effects ; indeed the condition of the 

 structures here very nearly approaches that of the steel as first 

 cast. It is typical " Widmanstatten " structure, and tends to 

 weakness. 



Of course it is well known that the smaller the crystals are the 

 better the resultant material, so long as deformation has not taken 

 place. The old-time smith had some idea of this, and consequently 

 belaboured his glowing iron right royally so that it should prove 



