ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 119 



tension, the potential of six reached a constant value shortly after 

 fracture. The difference between the initial and final potentials varied 

 from 0-0019 to 0*0077 volt. The conclusion is drawn that even 

 beyond the elastic limit the corrosion of iron is not greatly affected by 

 stress. 



Hard and Soft States in Ductile Metals.* — Gt. T. Beilby, in con- 

 tinuation of his previous work on this subject, has sought to define 

 more accurately the temperature range over which crystallisation takes 

 place in metals hardened by cold work. Hard drawn wires of gold, 

 silver and copper were heated to various temperatures. Observations 

 were made of the microstructure, the mechanical stability (by determining 

 the load which would give a permanent extension of 1 p.c), the E.M.F. 

 given by a thermocouple consisting of a hard wire and a wire previously 

 heated to the given temperature. The change in elasticity was deter- 

 mined by taking the pitch of the note given by reed vibrators of different 

 metals annealed at various temperatures. The following are among the 

 author's conclusions. The most severe mechanical working of a metal 

 always produces a mixed structure of the hard and soft phases. It has 

 not yet been found possible to produce a homogeneous specimen of 

 metal entirely in the hard state. The temperature ranges over which 

 (1) re-crystallisation, (2) loss of mechanical stability, (3) development of 

 thermal E.M.F. between wires in the hard and soft states, (4) complete 

 restoration of elasticity in hardened metal occur, coincide with each other 

 closely. The maximum amount of change in gold, silver and copper 

 occurs between 200° and 800° C. The change is essentially the develop- 

 ment of the crystalline from the non-crystalline condition. 



Densities and Specific Heats of Some Alloys of Iron. J — From 

 measurements made on a large number of alloys, quenched in water from 

 a bright red heat, W. Brown has determined the effect upon the specific 

 volume and specific heat of iron, of additions of carbon, manganese, 

 nickel, tungsten, silicon, chromium, copper, cobalt and aluminium. The 

 results are expressed as change per 1 p.c. of added element. By applying 

 these results to the calculation of dissipation of energy per cycle in 

 armature cores, the superiority for this purpose of silicon steel to pure 

 iron or other alloys is demonstrated. 



Alloys of Iron with Molybdenum 4— Lautsch and G. Tammann 

 have sought to determine the equilibrium diagram. The metals melted 

 in magnesia tubes were heated to 1800°-1850° C, and the protected 

 thermocouple inserted when the temperature had fallen to 1600° C. 

 Alloys with more than 70 p.c. molybdenum could not be made homo- 

 geneous in this way, the molybdenum not dissolving completely. Abnor- 

 malities apparent in the curve, which theoretically cannot occur in a 

 two-component system, have led the authors to put forward the hypo- 

 thesis that owing to the slow formation of a compound the system must 

 be considered as one of three components — iron, molybdenum and the 



* Proc. Roy. Soc, Series A, Ixxix. (1907) pp. 463-80 (12 figs.). See also 

 Nature, lxxvi. (1907) pp. 572-4 (2 figs.). 



t Trans. Roy. Dublin Soc., ix. (1907) pp. 59-84 (6 figs.). 

 X Zeitschr. Anorg. Chem., lv. (1907) pp. 386-401 (18 figs.). 



