114 SUMMARY OF CURRENT RESEARCHES RELATING TO 



study of the causes of crystallisation causing brittleness, developed by 

 cementation ; (4) study of deformation caused by cementation, appearing 

 either after that operation or after the subsequent quenching. 



The Composition of the Eutectic Copper-Copper Oxide.* — E. Heyn 

 defends the accuracy of his statement that the eutectic contains ;^ • 5 p.c. 

 oxide. With 5 p.c. oxide (the eutectic composition given by Bejean) 

 the alloy is clearly hyper-eutectic. This is shown by the distinct segre- 

 gation of a 5 p.c. alloy — a pure eutectic shows no segregation. The 

 author suggests that Dejean's high results are due to faulty sampling of 

 the ingots for analysis. 



Pournel's Researches and the Lower Limit of A2.t — F. Osmond, 

 remarking on the difficulty of determining the temperature at which, on 

 cooUng, the magnetic transformation of iron ends, indicates the bearing 

 of Fournel's results (embodied in two papers here given in full) on 

 the question. Fournel determined the critical points by measuring 

 electrical resistance throughout the range of temperature. Osmond 

 considers that, in the curves showing relation between resistance and 

 temperature, the point at which the curve ceases to be a straight line 

 marks the lower limit of the range A 2. This temperature varied from 

 ?jr)O°C.-l()0°C. in the steels used in Fournel's investigations. Osmond 

 gives a diagram showing the proportions of the a, ji and y modifications 

 contained in iron at different temperatures. 



Nickel-Silicon Steels.J— A further instalment of L. Guillet's work 

 on quaternary steels. Forty-nine alloys, prepared to show the effect 

 of increasing amounts of silicon on the three classes of nickel steels 

 (pearlitic, martensitic and y-iron) were subjected to exhaustive micro- 

 scopic examination and were tested mechanically. Therresults appear to 

 agree substantially with the deductions drawn from the author's researches 

 on nickel steels and silicon steels, subject to the following modifications. 

 (1) The addition of silicon tends {a) to hinder the formation of martensite ; 

 {h) to cause a y-iron nickel steel near the border of the martensitic class 

 to become martensitic ; {c) to increase the maximum stress and elastic 

 limit, and diminish elongation and resistance to shock, of pearlitic steels. 

 Size of grain is diminished ; {d) to improve the mechanical qualities of 

 y-iron steels by raising the maximum stress ; (e) to raise the position 

 of the thermal critical points. (2) The presence of nickel tends [a) to 

 counteract the effect of silicon in causing graphite separation, both in 

 the original condition and after annealing at 900" C. ; {h) to cause 

 graphite to be more readily separated in high carbon than in low carbon 

 steels. White areas visible in y-iron steels of sufficiently high silicon 

 content appear to be a silicide of iron or nickel. Some of the steels 

 containing large amounts of nickel and silicon show extraordinarily 

 complex micro structures. The addition of more than 2 p.c. silicon to 

 nickel steels does not appear to offer any advantages. 



Etching Reagents for Steel.§— Kourbatoff states that failures met 

 •with in using one of his etching solutions have been due to the employ- 



* Rev. Metallurgie, iii. (1906) pp. 543-4. 



t Tom. cit., pp. 551-7 (3 figs.). See also this Journal, 1905, p. 516. 



X Tom. cit., pp. 558-77 (19 figs.). 



§ Tom. cit., p. 648. See also this Jom-nal, 1905, p. 392, and 1906, p. 635. 



