August 8, 1901] 



NA TURE 



0^0 



permit.] The welding of three miles of electrical tramway rails 

 was successfully eftected in Brunswick in May 1900. 



As regards the comparison of the use of aluminium as fuel 

 with the eleclric arc, M. Camille Matignon (Moiiilmr Srieit- 

 tifitjHC Qiiesntmlk, No. 702, Juin 1900, p. 357 et seq.), in a 

 very interesting discourse recently delivered in Paris, has insti- 

 tuted a comparison between the Goldschmidt process and 

 the electric furnace. Quoting Moissan ( " Le Four electrique," 

 p. 19), he shows that in reducing titanic acid by carbon 

 in the electric furnace having a "laboratory space" of Soo 

 cubic centimetres, 300 horse-power absolute were employed, 

 producing per second 190,500 calories by burning I 'oS kilograms 

 of aluminium. On the other hand, by burning 3'2 kilograms of 

 ferric cxide during one minute in a crucible of about the same 



Fig 6.— Section of the welded test piece {Fig. 7), showing crystals passing 

 .across the line of weld, A b. Magnification 140 diameters. 



capacity as the labiratory of the electric furnace, the rate of 

 evolution of heat is equivalent to 375 horse-power absolute ; the 

 latter process does not, however, work continuously, but could 

 readdy be made to do so. It should be pointed out that an 

 impure variety of aluminium can be used, and that if the heat 

 needed to etifect a given operation is but moderate, the aluminium 

 •may be diluted by the presence of an inert substance. 



The photomicrograph (Fig. 6) is from a little test piece of 

 wrought iron (Fig. 7) which was cut in two. The carefully 

 faced surfaces were then clamped together, and I united them 

 into an excellent weld, without any previous experience in con- 

 ducting such an operation. No line of demarcation can be seen, 

 and the crystals pass over the line A B, which I know by 

 measurement to be that of the actual weld. 



C: 



The very hot molten iron may be used in a somewhat different 

 way for repairing defective castings. In this case the slag is care- 

 fully poured off the fluid iron in the crucible and the iron is then 

 poured into the defective part in the casting vvhich it is required 

 to mend, a guiding rim of some refractory material being pro- 

 vided. By mixing other metallic oxides with the iron oxide, the 

 metals they contain are reduced and alloy themselves with the 

 iron, and the composition of the defective casting can thus be 

 matched. In connection with the repairs of fractured or defective 

 steel castings, the possibility of producing directly steel of a suit- 

 able degree of carburisation is important. This may readily be 

 effected by mixing fragments of cast iron with the "thermit," 

 thus 70 to 90 grammes of cast iron mixed with 1000 grammes of 



NO. 1658, VOL. 64] 



thermit gives a very fine-grained and workable steel. One useful 

 application of the process is for locally softening hardened armour 

 plates in the positions where the bolts and screws have to be 

 inserted through the holes drilled to admit them. This is 

 effected by placing a little fluid " thermit " on the spot where the 

 plate has to be drilled and the heat softens the hardened surface. 

 It should also be remembered that, with reference to the repairs 

 of defective parts of machinery, a suitable admixture of metallic 

 oxides with the ferric oxide, such as those of chromium, nickel 

 or manganese, may be reduced together with the iron derived 

 from the ferric oxide. Richly carburised iron may be added to 

 the molten mass, and in this way any quality ofsteelmaybe 

 produced. 



This latter reference to metallic oxides reminds us of the 

 original use for which the finely divided aluminium was em- 

 ployed, namely, as a reducing agent for the rarer metals and not 

 for the sake of the heat evolved by the reaction. This portion 

 of the subject I dealt with at the Royal Institution six years ago, 

 but there have been great advances since. It would have been 

 tedious to have conducted the experiments before you, as the 

 crucibles would have taken so long to cool; but in each of these 

 crucibles, which will now be broken open, I hope to find a small 

 mass of metal, which, until now, has not left the spot in which 

 it was reduced. [About a pound of nickel and a pound of cobalt 

 were then produced from the respective crucibles in which they 

 had been reduced]. 



Manganese and chromium containing only small quantities 

 of carbon are now produced on a large scale for industiial 

 use. As regards the reduction of metals and alloys from their 

 oxides by burning aluminium, the following are the most recent 

 results that have been obtained {S/ahl mid Eiseii, March 24, 

 1901). The use of carbon-free chromium in connection with 

 the metallurgy of steel is an exceedingly useful development of 

 the methods we have considered. Hitherto, the addition or 

 ferro-chrome to steel has involved a loss of from 20 to 25 per 

 cent, of the chromium, while with pure chromium the loss is 

 slight. Moreover, the addition of ferro-chrome incidentally 

 raises the percentage of carbon, and steel containing, for 

 instance, 2"5 per cent, of chromium should not have more than 

 fromo'i5 to 0'20 per cent, of carbon, and this can only be 

 attained by the use of pure chromium. In the manufacture, also, 

 of tool steel, the percentage of chromium may reach from 6 to 10 

 per cent, and even higher, a result which is only rendered possible 

 by the use of pure chromium. In the same way, in connection 

 with the metallurgy of copper, the possibility of providing 

 carbon-free manganese is important, as is also the preparation 

 of cupro-manganese free from iron. Alloys of manganese with 

 zinc and with tin are likely to prove of value. Many uses have 

 been found for the alloy containing 80 per cent, of zinc and 20 

 per cent, of manganese, while it is anticipated that the alloy 

 containing 50 per cent, of tin and 50 per cent, of manganese will 

 also prove to be important. Use has also been found for an 

 alloy of 70 per cent, manganese and 30 per cent, chromium. 

 Ferro-titanium, with 20 to 25 per cent, of titanium, and alloys 

 of titanium and manganese containing from 30 to 35 per cent, of 

 titanium, have also been produced. Titanium, moreover, 

 absorbs nitrogen, and ferro-litanium is found to be very useful 

 in producing sound steel castings. I, quite independently o. 

 Dr. Goldschmidt, succeeded in the preparation of alloys of iron 

 with from 3 to 25 per cent, of boron, the alloy containing 3 

 per cent, of boron proving to be beautifully crystallised. Dr. 

 Goldschmidt states that definite results have not been obtained in 

 attempts to utilise it. I am still investigating this most interest- 

 ing subject. Dr. Goldschmidt has obtained ferro-vanadium, 

 the be.st results being given with steel containing o'5 per cent, 

 of vanadium. He has also prepared an alloy of lead and barium 

 containing 30 per cent, of barium, which affords an example of 

 the possibility of forming alloys of metals with those of the 

 alkaline earths by this process. 



It only remains for me to direct your attention to the nature 

 of the solid product of the combu.stion of aluminium, which is 

 alumina often of a high degree of purity, and in a specially 

 interesting form. The alumina from the reduction of oxide 

 of chromium, when it is allowed to cool, forms large ruby- 

 tinted crystalline masses, closely resembling the natural ruby. I 

 have now to show you on the screen some rubies and sapphires 

 produced as an incident of this beautiful process. The 

 blue sapphire mass is, however, only translucent, not trans- 

 parent. The ruby crystals are often very beautiful, as these 

 slides show. Rubies placed in a vacuum tube and subjected 



