564 TRANSACTIONS OP SECTION B. 



The second line of inquiry disclosed what traces of blowholes remained in 

 the boiler plate by cutting very thin slices lengthwise and crosswise from that 

 plate, mirror-polishing them, and then bending them double hi such a way that 

 any blowhole traces present ought to gape open like the cards of a bent pack. 

 Had there been no welding of blowholes this bending should have disclosed 

 unwelded seams about 3.5 inches long and 1.3 inch wide. In point of fact 

 the traces detected were so short as to indicate strongly that a very great degree 

 of welding had occurred. It had seemed to me extremely probable, antecedently, 

 that such welding ought to occur ; but here some very competent writers had 

 differed from me. The longest single trace was 0"7 inch long. Only one 

 important ' string ' of such traces was found and this was only 0-3 inch long. 

 Further the scantiness of these relics of blowholes tends to show that the blowhole 

 gases have been reabsorbed by the metal to a very great degree. I suggest that 

 such relics of blowholes as have persisted represent in most cases spots where 

 the reabsorption of the gas has become complete after the temperature has fallen 

 too low to permit welding. I therefore suggest prolonging the exposure to a 

 temperature above the welding-point, so as to complete the reabsorption of gas 

 while the metal is still weldable. 



The reabsorption of the blowhole gases and the welding of the blowholes 

 ought to be promoted rather by the practice of ' reheating ' than by that of 

 ' direct rolling. ' In the former the ingot is rolled part way towards its final shape, 

 and the resultant ' bloom ' is then reheated before further rolling ; in the latter 

 the ingot is rolled to its final shape, such as a rail or a boiler plate, at a single 

 heat. During the early part of the rolling the metal surrounding each blowhole 

 should become strongly charged with gas reabsorbed from that blowhole because 

 of the enormous pressure caused by rolling. To reheat the bloom exposes it for 

 a long time to a high temperature in the heating furnace, during which time the 

 gas dissolved in the metal immediately' around the blowhole has an opportunity 

 to diffuse away, leaving that metal free to absorb a new lot of gas from the 

 blowhole. And further, the high temperature of the rolling immediately after this 

 reheating facilitates both the further absorption and outward diffusion of whatever 

 gas remains in the blowholes and the welding up of their sides. 



4. The Provident Use of Coal. By Professor H. B. Armstrong, F.R.S. 



5. The Influence of Chemical Composition and Thermal Treatment on 

 Properties of Steel. By Professor A. McWilliam, A.R.S.M., M.Met. 



Sheffield is pre-eminently the home of the manufacture of special steels. 

 and she attains her ends by strict attention to ultimate chemical composition, 

 by insuring the presence of the most desirable amounts of beneficial substances, 

 and the practicable minimum of deleterious contents, be they the well-understood, 

 such as sulphur and phosphorus, or the more or less mysterious, such as oxygen ; 

 and also to those thermal changes called heat treatment, that are produced, 

 sometimes slowly, as in annealing, sometimes Buddenly, as in quenching, and 

 that vary the chemical composition or the physical nature of the several con- 

 stituents, and the mechanical and other properties of their aggregate, the finished 

 steel. Carbon is the chief of the elements that are varied to obtain the different 

 properties required in steels, and typical examples were given of the change in 

 properties produced by the gradual increase in the content of carbon, which, 

 at least in normal steels, is present as carbide of iron (Fe 3 C). Other elements 

 are added to vary the properties of the steel, probably mainly by their influence 

 on the nature, the composition, or the distribution of the carbide in the steel. 

 The properties of a series of steels with varying carbon contents, but containing 

 in addition about one per cent, manganese, were shown, and a similar series con- 

 taining two per cent, chromium, all from results quite recently obtained by the 

 author and Mr. E. J. Barnes. Examples of steels of similar carbon contents 

 containing (1) no special element, (2) one per cent, manganese, (3) two per cent, 

 chromium, (4) vanadium and chromium, (5) nickel and chromium, were tabulated 

 to show the comparative influence of these various additions. 



The effects of such heat treitments as long annealing and quenching, followed 



