METALLURGY. (!RON AND STEEL.) 



357 



tinuous spectrum was much stronger. Many lines 

 and bands new to the Bessemer flame spectra were 

 observed. Twenty-six plates of spectra were 

 photographed. The spectra increased in intensity 

 as the blow proceeded in the first stage, a re- 

 sult which could come only from a corresponding 

 increase in the temperature of the bath of metal 

 and of the flame. Difficulty was experienced in 

 the identification of some of the lines and bands; 

 some were due to uncommon elements, and others 

 were relatively much stronger than a study of the 

 oxyhydrogen flame and other spectra of the same 

 metals had hitherto shown. Line spectra were not 

 observed in the open-hearth furnace a circum- 

 stance attributed mainly to the fact that the at- 

 mosphere of the furnace is oxidizing, when only 

 sodium gives a spectrum approaching in intensity 

 that which it gives in the reducing flame. Among 

 other differences between the phenomena of the 

 basic Bessemer blow and those of the acid process, 

 are that in the former a flame is visible from the 

 commencement of the blowing ; that volatilization 

 of metal occurs largely at an early period in the 

 blow ; that a very large amount of fume is formed 

 toward the close of the second period; that the 

 overblow is characterized by a very powerful illu- 

 mination from what appears to be a brilliant yel- 

 low flame; and that the spectra of flames from 

 the first stage differ from those of the acid process 

 in several particulars the manganese bands be- 

 ing relatively feeble, and lines of elements not 

 usually associated w r ith Bessemer metal being 

 present. Differences were noticed in the intensity 

 of metallic lines, it varying with the amount of 

 metal in the charge, and also with changes of tem- 

 perature; as the temperature of the flame rose, 

 some lines faded almost away, and others became 

 stronger. The changes were more marked in the 

 arc spectrum and still more in the spark spec- 

 trum of iron. A new line of potassium, with 

 somewhat widely- variable intensity, was ob- 

 served. 



A new iron process described by A. v. Forselles 

 consists in melting in the blast-furnace a mix- 

 ture of apatite, or some other phosphatic rock, 

 with charcoal, scrap-iron, and suitable fluxes. 

 The products obtained are a slag rich in phos- 

 phorus, which is sold as manure, and a pig-iron 

 that contains phosphorus and can be made use 

 of in the Thomas or basic steel process. 



Brinell's method of determining hardness and 

 other properties of iron and steel, as described by 

 Axel Wahlberg, of Stockholm, consists in forcing, 

 by means of pressure, a hardened steel ball into 

 the material tested, so as to cause an impression, 

 the diameter of which is then to be measured, in 

 order to obtain the spherical area of the concavity. 

 The quotient resulting from dividing the maxi- 

 mum pressure by this area will represent what is 

 called by Brinell a hardness number, indicating, ac- 

 cording to him, the amount of pressure (kilo- 

 grams per square millimeter) to which the ma- 

 terial in hand has been subjected. With this 

 method several researches have been carried 

 out, detailed particulars of which are given. 

 They relate to the determination of the hardness 

 of various metals, controlling forging tests, and 

 to the hardness of iron and steel. Under this 

 head, experiments were made to ascertain the in- 

 fluence of the percentage of carbon on the harden- 

 ing capacity of different quenching liquids, the 

 influence of the temperature of the quenching 

 liquid on the hardening result, and the influence 

 of different hardening temperatures. Other re- 

 searches described dealt with an attempt to ascer- 

 tain the homogeneity of iron and steel, the degree 

 of annealing, the influence of cold-working deter- 



mination of the yield point, ?ilt<-i natr -itrcs.4 and 

 elongation, and tests oi bhnik . i,i ....> him els. 



Observations on the prodnd, >.i -,\ ii-iuens- 



Martin furnace in which tin i.inlly 



present are recorded by A. ZU^JMM ; i j i; i't 



a metal containing O.o5 per cen'c. oi iiu \> ' 

 cent, of antimony, 0.03 per cent, ot ;u 

 0.182 per cent, of copper, rolled quite sal 

 to plate. On further rolling to sheet, crack- ; 

 to develop on the edges of the sheet, hut they 

 could not, as it proved, be attributed to red-.^hoii 

 ness. Bars rolled from the ingot folded completely 

 over under the cold hand test, and showed a 

 tensile strength of 40 kilograms, with an exten- 

 sion of from 31 to 34 per cent. It appears to be 

 proved from this that the presence of 0.55 per 

 cent, of tin in iron or steel does not affect its 

 malleability, tenacity, or extensibility. The weld- 

 ing qualities at most are unproved. In other ex- 

 periments bearing on the same matter reported in 

 Stahl und Eisen and the Chemiker Zeitung, sam- 

 ples of iron containing from 0.1 to 0.63 per cent, 

 of tin, when tested, all forged well. On rolling, 

 the samples containing more tin split more readily 

 at the edges. All except the one with 0.63 per 

 cent, of tin, rolled well. They bent satisfactorily, 

 both hot and cold. The tin was irregularly dis- 

 tributed in the ingots. Crucible steel was pre- 

 pared, containing 0.23, 0.50, 0.68, and 0.62 per 

 cent, of tin. All samples forged well, but the last 

 was somewhat red-short. Some of them would 

 weld. The breaking stress was 72.3-73.9 kilo- 

 grams per square millimeter. 



Experiments are described by F. H. Williams, 

 of Pennsylvania, which go to show that copper 

 reduces the corrosion of soft steel to within that 

 of wrought-iron, and has a similar effect upon the 

 susceptibility of the latter metal. 



In a paper on the influence of copper on steel 

 rails and plates read before the Iron and Steel In- 

 stitute in May, J. E. Stead and John Evans re- 

 mark that it is generally thought that copper has 

 a very deleterious effect upon the metal, so that 

 engineers when buying steel frequently specify 

 that it must be absent. The observations of the 

 authors show that this general opinion is errone- 

 ous, and prove that between 0.5 and 1.3 per cent, 

 of copper has no deleterious effect upon either the 

 hot or cold property of steel; that a very large 

 amount (2 per cent.) makes the steel more liable 

 to be overheated; and that in small quantities 

 it slightly raises the tenacity and elastic limit, 

 but, unlike phosphorus, does not sensibly make 

 the steel liable to fracture under sudden shock. 

 Like carbon, it reduces the power of the steel to 

 extend under stress; but this is not pronounced 

 when the quantity is small. The effect is more 

 marked when large quantities of copper are pres- 

 ent. The authors furthermore prove that if the 

 evidence of the open-hearth steel trial can be 

 confirmed, copper, instead of producing red-short- 

 ness, has the contrary effect of changing red- 

 short steel so that it will roll without cracking. 



In a discussion on the properites of steel con- 

 taining nickel published in the Report of the Con- 

 gres International des Methodes d'Essai, the prin- 

 cipal changes in passing from ordinary steel to 

 steel containing a considerable proportion of 

 nickel are defined as being the lowering of the 

 temperature of transformation of the carbon, the 

 fusion of two of the transformations, and the ex- 

 aggeration of the phenomena accompanying the 

 double point. In ferro-nickels containing traces 

 of carbon, but more than 20 per cent, of nickel, 

 the transformations are determined by the nickel, 

 while the carbon acts as a retardent. The mag- 

 netic properties appear to be due to a certain 



