TRANSACTIONS OF SECTION B. 613 



the shrinking, is afforded by an examination of tlie action of reap:ents on cotton 

 yai-n. Thus, hanks of a long staple yarn having a mean breakiog strength of 

 4]7'4 ± 21 grams were immersed loose in caustic soda (sp. gr. 1-342) and saturated 

 barium mercuric iodide solution, and the following changes in the breaking load of 

 the yarn and the lengths of the hanks were found to result : — 



Caustic Soda. — Mean breaking load, 52G-:i ± 3-8 grams : shrinkao-e, from 

 66-0 to 44-8 cm. ^ , „ , 



Barium Mercuric Iodide. — Mean breaking lead, 526-6 ± 3-3 grams ; shrinkage, 

 from 66-0 to 48-9 cm. 



Although the shrinkage and the increase in the breaking load brought about 

 by these two reagents are so nearly the same, yet on immersing hanks under tension 

 in these solutions and washing whilst still under strain the hank treated with 

 soda acquires a brilliant lustre, whilst that treated with the iodide exhibits only 

 a trace more lustre than the untreated yarn. The explanation of this result is 

 found in the fact that caustic soda causes rapid untwisting of the fibre, whilst 

 barium mercuric iodide does not cause untwisting. 



The authors giye a list of reagents which bring about two of the three effects 

 shown to be essential to the production of lustre — namely, swelling, shrinking, and 

 untwi.sting— -and find that ' lustreing ' cannot be eftected with' such reagents ; 

 several solutions are known, however, which cause the three effects, and with the 

 aid of such liquids the lustre can always be produced. 



5. Stead's recent Researches as to the Causes and Prevention of 

 Brittleness in Steel. By Professor T. Turner, M.Sc. 



After briefly referring to the nature of a eutectic, and the characteristic 

 microstructure of such bodies, as pointed out by Osmond, the author outlined the 

 structure of steel. Special reference was made to the properties and distribution 

 of farrite and pearlitein metal, when used in its natural state for constructional 

 purposes, and containing about 0-45 per cent, of carbon. A short summary was 

 then given of the work of Brinell, Heyn, Stansfield, and of Stead and Richards in 

 reference to brittleness caused by heating steel either for a short period to a high 

 temperature, or for a longer time at a lower temperature (900° 0.). The crystalline 

 character and brittleness so produced can be at once removed, in most cases, by 

 heating to slightly under 900° 0. The structure of steel of good quality is, 

 therefore, largely dependent on the rate of cooling through the point Ac^. Details 

 were also given of the work of Stead and Richards on the production of sorbite in 

 steel. The maximum quantity of sorbite is obtained by cooling the heated steel 

 rapidly until its temperature is below the critical points, and then tempering 

 either by external heat, or, in the case of rails and other similar large objects, 

 by the internal heat of the partly cooled steel. Rails which have been rendered 

 sorbitic in this way have a higher tensile strength and greater wearing power 

 than ordinary rails. Sorbitic steel, when tested by repeated reversals of stress, 

 also shows much greater toughness and endurance. A number of photographs 

 were exhibited which showed very plainly that the microstructure of the sorbitic 

 portion of asteel rail is quite difierent from that of the rest of the steel. The 

 normal portion consists of a heterogeneous mixture of ferrite and pearlite, while 

 the sorbitic portion is almost perfectly homogeneous. 



The papers to which special reference was made were read at the Iron and 

 Steel Institute, September 1903, and are as follows : — 



1. The Burning and Overheating of Steel, by A. Stansfield. 



2. The Restoration of Dangerously Crystalline Steel by Heat Trevtment, by 

 J. E. Stead and A. W. Richards. 



3. Sorbitic Steel Rails, by J. E. Stead and A. W. Richards. 



