ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 259 



and growth jof the ferrite crystals occur on anneahng, at temperatures 

 below the critical range, a low-carbon steel which has been previously 

 strained in any way. A. E. AVhite and H. F. Wood describe experi- 

 ments with a steel containing • 1 1 p.c. carbon, which had been pre- 

 viously locally deformed by making a Brinell indentation with a ball 

 of 5 mm, diam. under a pressure of 800 kilogrm., with the object of 

 determining the relation between time and temperature governing 

 crystal growth on annealing at temperatures below the critical range. 

 Photomicrographs are given showing (1) the structures of the steel in 

 the neighbourhood of the indentation, and that of the unstrained metal 

 at a distance from the indentation ; (2) the corresponding structures 

 after annealing at various temperatures from 850° C. to 550° C. for 

 varying lengths of time. The strained crystals only showed any change 

 in structure. At temperatures down to 700° C. recrystallization took 

 place in a fraction of a minute. At 675° C. a period of eight minutes 

 was required. Below 675° C. the time required for crystal growth is 

 considered to vary in geometrical progression for every fall in tempera- 

 ture of 25° C. : e.g. at 550° C. annealing for thirty-five hours was re- 

 quired before crystal growth was noticed. Annealing for longer periods 

 than the critical period for each temperature produces very coarse 

 structures. 



The bearing of these results upon the causes of frequent failures 

 among low-carbon steel boiler tubes is next considered. The possible 

 overheating of a tube by the deposition of scale on the water side 

 might result in rapid crystal growth and the weakness associated with 

 a coarse crystallization. Photomicrographs are given of the structure of 

 a failed tube, showing that in this instance crystal growth had clearly 

 taken place, the structure being exceedingly coarse on the fire side. To 

 prevent failures the deposition of scale and consequent overheating 

 should be avoided, or steel of a higher carbon-content might be 

 employed. 



< I = ^ I » 



