Nov. 14, 1689] 



NATURE 



37 



Victory consumed only 325 pounds of powder, that from 

 the Victoria will consume 3000 pounds. The most for- 

 midable projectiles belong to the highly carburized class 

 of steel. Shells contain o"8 to 094 per cent, of carbon, 

 and, in addition, some of these have 0*94 to 2 per cent, 

 of chromium. The firm of Holtzer shows, in the Paris 

 Exhibition, a shell which pierced a steel plate 10 inches 

 thick, and was found, nearly 8co yards from the plate, 

 entire and without flaw, its point alone being slightly dis- 

 torted. Compound armour-plate with steel face, which 

 face contains o"8 per cent, of carbon, is, however, more 

 difficult to pierce than a simple plate of steel. 



[A prominent feature in the " trophy," among the class 

 of highly carburized steels which contain over ^"jj per 

 cent, of carbon, was a fine suspended wire y%^^ of an 

 inch diameter, of remarkable strength, supporting a weight 

 of i\ cwt., or a load of nearly 160 tons to the square 

 inch. The strength of the same steel undrawn^ would 

 not exceed 50 tons to the square inch. A similar wire 

 manufactured by the steel company of Firminy attracted 

 much attention in the Paris Exhibition by supporting a 

 shell weighing i8co lbs., or a load of 158 tons per square 

 inch.] 



Lastly, I will refer to the highly carburized steel used 

 for the manufacture of dies. Such a steel should contain 

 0"8 to I per cent, of carbon, and no manganese. It is 

 usual to water-harden and temper them to a straw colour, 

 and a really good die will strike 40,000 coins of average 

 dimensions without being fractured or deformed ; but I 

 am safe in saying that if the steel contained y\j percent, 

 too much carbon, it would not strike 100 pieces without 

 cracking, and if it contained ,-5 per cent, too little carbon, 

 it would probably be hopelessly distorted, and its engraved 

 surface destroyed, in the attempt to strike a single coin. 



The above examples will be sufficient to show how 

 diverse are the properties which carbon confers upon iron, 

 but as Faraday said, in 1822, " It is not improbable that 

 there may be other bodies besides charcoal capable of 

 giving to iron the properties of steel." The strange thing is 

 that we do not know with any certainty whether, in the 

 absence of carbon, other elements do play the part of 

 that metalloid, in enabling iron to be hardened by rapid 

 cooling. Take the case of chromium, for instance : 

 chromium-carbon steels can, as is well known, be ener- 

 getically hardened, but Busek ^ has recently asserted that 

 the addition of chromium to iron in the absence of car- 

 bon does not enable the iron to be hardened by rapid 

 cooling. So far as I can see, it is only by employing the 

 electrical method of Pepys that a decision can be arrived 

 at as to the hardening properties of elements other than 

 carbon. 



A few words must be devoted to the consideration of the 

 colours which, as I said (see attte, p. 11), direct the artist 

 in tempering or reducing the hardness of steel to any deter- 

 minate standard. The technical treatises usually give — 

 not always accurately, as Reiser ^ has shown — a scale of 

 temperature ranging from 220" to 330"^, at which various 

 tints appear, passing from very pale yellow to brown yellow, 

 purples, and blues, to blue tinged with green, and finally to 

 grey. Barus and StrouhaF point out that it is possible 

 that the colour of the oxide film may afford an indication 

 of the temper of steel of far greater critical sensitiveness 

 than has hitherto been supposed. It is, however, at 

 present uncertain how far time, temperature, and colour 

 are correlated, but the question is being investigated by 

 Mr. Turner, formerly one of my own students at the 

 School of Mines. 



That the colours produced are really due to oxidation 

 was shown by Sir Humphry Davy in i8i3,'» but the nature 



' Stahl und Eisen, ix. 1889, p. 728. 



^ "Das Harten des Stahles," p. 78 (Leipzig, 1881). See also Loewenherz, 

 Zeitschrift fiir Instruinentenknnde, ix., 1889, p. 322. 



3 Bull. U.S . Ceo. Survey, No. 27, 18S6, p. 51. 



^ Sir Humphry Davey, Thomson's Ann. Phil., i., 1813, p. 131 ; quoted 

 by Turner, Proc. Phil. Soc, Birmingham, vi., 1889, part 2. 



of the film has been the subject of much controversy. 

 Barus points out that "the oxygen molecule does not 

 penetrate deeper than a few thousand times its own 

 dimensions,^ and that it probably passes through the film 

 by a process allied to liquid diffusion. The permeable 

 depth increases rapidly with the temperature, until at an 

 incipient red heat the film is sufficiently thick to be 

 brittle and liable to rupture, whereupon the present phe- 

 nomenon ceases, or is repeated in irregular succession. 



Looking back over all the facts we have dealt with, it 

 will be evident that two sets of considerations are of 

 special importance : (i) those which belong to the rela- 

 tions of carbon and iron, and (2) those which contem- 

 plate molecular change in the iron itself. The first ot 

 these has been deliberately subordinated to the second, 

 although it would have been possible to have written 

 much in support of the view that carburized iron is an 

 alloy of carbon and iron, and to have traced with Guthrie 

 the analogies which alloys, in cooling, present to cooling 

 masses of igneous rocks, such as granite, which, as the 

 temperature of the mass falls, throws off " atomically 

 definite " - bodies, leaving behind a fluid mass of indefi 

 nite composition, from which the quartz and feldspar 

 solidify before the mica. This view has been developed 

 with much ability in relation to carburized iron by Prof 

 Howe, of Boston, who even suggests mineralogical 

 names, such as " cementite," " perlite," and " ferrite," for 

 the various associations of carbon and iron. 



I am far from wishing to ignore the interest presented 

 by such analogies, but I believe that the possibility of mole- 

 cular change in the iron itself, which results in its passage 

 into a distinctive form of iron, is at present the more im- 

 portant subject for consideration, not merely in relation 

 to iron, but as regards the wider question of allotropy in 

 metals generally. 



Many facts noted in spectroscopic work will have, as 

 Lockyer has shown, indicated the high probability that 

 the molecular structure of a metal like iron is gradually 

 simplified as higher temperatures are employed. These 

 various simplifications may be regarded as allotropic 

 modifications. 



The question of molecular change in solid metals 

 urgently demands continued and rigorous investigation. 

 Every chemist knows how much his science has 

 gained, and what important discoveries have been made 

 in it, by the recognition of the fact that the elements 

 act on each other in accordance with the great law of 

 Mendeleeff which states that the properties of the elements 

 are periodic functions of their atomic weights. I firmly 

 believe that it will be shown that the relation between 

 small quantities of elements and the masses in which they 

 are hidden is not at variance with the same law. I have 

 elsewhere tried to show ■' that this may be true, by exa- 

 mining the effect of small quantities of impurity on the 

 tenacity of gold. 



In the case of iron, it is difficult to say what property 

 of the metal will be most affected by the added matter. 

 Possibly the direct connection with the periodic law will 

 be traced by the effect of a given element in retarding or 

 promoting the passage of ordinary iron to an allotropic 

 state ; but " the future of steel " will depend on the care 

 with which we investigate the nature of the influence 

 exerted by various elements on iron, and on the thermal 

 treatment to which it may most suitably be subjected. 



Is it not strange that so many researches should have 

 been devoted to the relations between carbon, hydrogen, 

 and oxygen in organic compounds, so few to the relations 

 of iron and carbon, and hardly any to iron in association 

 with other elements ? I think that the reason for the com- 

 parative neglect of metals as subjects of research arises 



' Bull. U.S. Geo. Sun>ey, No. 35* "^886, p. 51. 



2 Phil. Mag., June 1884, p. 462. 



3 Phil. Trans. Roy. Sbc., clxxi.t., i8«^, p. 339- 



