ioo SCIENCE PROGRESS 



The size of these cells is a measure of the heat-treatment which 

 the steel has undergone. This is explained by the behaviour of 

 such steel when heated above the recalescence point of 6go°. 

 At a high temperature, the iron-carbide complex (sorbite or 

 pearlite) acts as a solvent for the ferrite of which the cell-walls 

 are composed ; the crystal grains thus produced grow, like the 

 grains of pure iron, during the annealing process. When the 

 steel is again cooled, the excess of ferrite is no longer held in a 

 homogeneous condition and becomes visible in the first instance 

 at the boundaries of the grains. The size of the cells is an 

 indication of the size of the crystal grains present at a high 

 temperature and is therefore either a measure of the tempera- 

 ture at which the steel has been annealed or, if that be known, 

 of the time during which the metal has been exposed to that 

 temperature. Further, the thickness of the cell-walls is an 

 indication of the rate of cooling, as the first deposition of 

 ferrite takes place at the boundaries of the original grains and 

 any ferrite subsequently deposited must appear in scattered 

 granules within the cell if cooling be rapid but become attached 

 to the cell-wall as an internal thickening if sufficient time be 

 given to allow of free diffusion through the solid mass ; a thin 

 cell-wall is therefore evidence of rapid cooling. 1 If the com- 

 position of the steel and especially its carbon-content be known, 

 an inspection of the micro-sections gives a complete knowledge 

 of the heat-treatment to which the steel has been subjected, 

 knowledge which is ot the utmost value when rails are con- 

 cerned, the relationship between heat-treatment and the physical 

 and mechanical properties on which the life of the rail depends 

 being now well known. 



The deposition of any substance present in excess during 

 the cooling of a homogeneous solid along the boundaries of the 

 crystal grains is not peculiar to steel. It is also observed in 

 alloys of the Muntz-metal class. An alloy of this kind, heated 

 to such a temperature as to be wholly or almost wholly con- 

 verted into the /3-constituent, has crystal grains of a size which 

 depends both on the time and temperature of annealing. 

 During cooling the a-constituent crystallises at the boundaries 

 of the grains and the extent to which thickening of the cell-walls 

 takes place by diffusion depends on the rate of cooling. 



The last point to be considered in the present article is the 



1 See H. M. Howe, Internat. Zeitsch. Metallographie, 191 2, 2, 13. 



