516 SUMMARY OF CUERENT RESEARCHES RELATING TO 



oxide, prepared by melting copper in an open crucible so that it oxidises 

 freely, is recommended as a fixed point for calibrating thermocouples, 

 and may be taken as 1062°. 



Rail Corrugation.* — G. Moyle discusses the singular phenomenon 

 of roaring rails, i.e. rails which in use develop furrows across the running 

 head, causing a deafening noise when a train passes over. The ascer- 

 tained facts are briefly stated, and a list is given of the numerous causes 

 suggested. A report on three " roarers," drawn up at the Cooper's Hill 

 laboratory, states that the unevenness of surface is not due to the 

 chemical composition or physical state of the rail but to some local 

 cause. The author considers that the results of investigation are meagre 

 and unsatisfactory. Microscopic examination appears to throw no light 

 on the cause. 



Copper Steels.t — P. Breuil has investigated two series of steels con- 

 taining copper, the carbon in the first being about • 15 p.c, in the second 

 about 0*35 p.c. The copper in the members of each series was 0*5, 1,2, 

 4, 8, 16 and 32 p.c. Analyses of the top and bottom of the ingots 

 showed that the copper was uniformly distributed, except in the ingot 

 of the second series containing 32 p.c. copper. This practically con- 

 sisted of two portions. The copper content varied from 21-75 p.c. 

 The fractures of the ingots with 8 p.c. or more of copper showed a red 

 coloration. The hardness of the steels increased with increase in copper 

 content. Peculiarities in the position of the critical points were noted. 



The Crystallography of Iron.l — F. Osmond and G. Cartaud put 

 forward an explanation of the structure of martensite. When a small 

 piece of ordinary steel containing manganese is quenched from about 

 1100° in cold water, cracks may be developed. Around the cracks are 

 very fine twin crystals. The microstructure of a polished section is 

 exactly similar to that of martensite in carbon steels. The partial trans- 

 formation of y iron into /? iron occurring during quenching produces 

 stresses, owing to change of volume. These stresses cause the formation 

 of twin crystals in great numbers, parallel in any one grain to the four 

 pairs of faces of the octahedron ; whence the frequency of square 

 figures and equilateral triangles. The marked resemblance between the 

 structures of martensite and of meteoric iron is pointed out. 



Critical Points of Steel. § — P. Fournel has succeeded in detecting 

 Aj and A 2 as well as A 3 by the variation of electrical resistance of steel 

 with temperature. Wires 0'3 mm. diam. 30 cm. long were wound on 

 mica and heated in vacuo in an electric resistance furnace. In series 

 with the wire was a standard 1-ohm resistance ; a current of a few 

 hundredths of an ampere was passed through, and the difference of 

 potential between the two ends of the wire, and of the two ends of 

 the standard resistance, measured by a potentiometer. Temperatures 

 were measured by a thermocouple. Figures and curves for five steels 

 are given by the author. 



* Tramway and Railway World, xix. (1906) pp. 558-61 (9 figs.). 



t Comptes Rendus, cxlii. (1906) pp. 1421-4. 



t Op. cit., cxliii. (1906) pp. 44-6. § Tom. cit., pp. 46-9 (1 fig.). 



