NA TURE 



[May 5, 1904 



In part iii., on the mechanical treatment of steel, by 

 Mr. J. W. Hall, the excursions into theoretical matters 

 of pure metallurg:y are not always happy, but the other 

 parts are treated as one would expect from an engineer 

 of his enthusiasm and experience. The development of 

 various types of mills, examples of modern plant, forg- 

 ing bv the hammer and by the press, all seem excel- 

 lently treated, while the case for and against ffuid com- 

 pression is made very clear. Several Sheffield firms, 

 however, make high speed steels, generally acknow- 

 ledged to be much more than " nearly equal to those 

 at Bethlehem." 



Part iv., by .Mr. Harbord, on finished steel, treats 

 of the metal steel itself, its mechanical properties, the 

 relations of iron and carbon, influence of other 

 elements, effect of heat treatment, and the micro- 

 scopical examination of steel. Mechanical testing 

 makes a good chapter, but why use the erroneous 

 term " tensile strain " instead of " maximum 

 stress "? 



To the chapter on iron and carbon many will eagerly 

 turn, because of the paramount importance of the sub- 

 ject in everyday work, its great historical interest, and 

 it may be also because of recent controversy. Perhaps, 

 therefore, one expects too much, but it must be con- 

 fessed that the author does not seem to have risen to 

 the occasion, and gives only a not too excellent compila- 

 tion where one expected a sorting out and a grappling 

 with the question. The well known diamond and iron 

 in vacuo experiment is taken as proof that carbon 

 can be transferred to iron without the intervention of 

 gas, whereas it is now well known that steel which 

 has ceased to give off gas even at 1000° C. will give 

 off more if heated to 1200° C. After a description of 

 temper (or annealing) carbon, " the existence of this 

 form of carbon has not been confirmed by other in- 

 vestigators " is a rather startling statement. The re- 

 search on " The Influence of Carbon on Iron," pub- 

 lished by Prof. Arnold in 1895, and acknowledged by 

 all to be a classic, is quoted, and with it some recent 

 work of the author's own which only seems to obscure 

 the subject. The author hardly deals fairly with his 

 readers in withholding the tests of his steels as re- 

 ceived, and the first three are given here (see sixth 

 report .Alloys Research Committee) side by side with 

 ordinary commercial samples taken at random. It 

 will be clear that they are a very undesirable series as 

 a basis for such a research. The three numbers re- 

 present maximum stress in tons per square inch, 

 elongation per cent, on two inches, and reduction of 

 area per cent, respectively : — 



Ilarbord ... 33 13 25 

 Commercial 35 2S 47 

 Harbord ... 37 6'3 19 



The author seems almost alone at the present date 

 in the opinion that " hardening carbon is possibly 

 merely free carbon dissolved in iron," as even Stans- 

 field, Osmond and Stead all favour the idea of carbide 

 dissolved in iron, and the author ou^ht to have told 

 his readers this. He also says that the carbon theory 

 does not explain the critical points in pure iron and 

 NO. I 80 1, VOL. 70] 



loss of magnetism, but surely he must know that the 

 main function of the carbon theory is to explain the 

 hardening of steel. He is hardly up to date on some 

 important matters, as he takes Arnold's mere sugges- 

 tion of long ago as to Ar^, and leaves the student to 

 think that this is the present well defined position, 

 whereas both sides are agreed as to hr^ being the real 

 carbon change point, and to carbonists it represents the 

 formation or decomposition of a substance correspond- 

 ing to the formula Fe„jC. All are agreed also that the 

 purest iron shows Ar., and Ar,, and Prof. Arnold's 

 theory with regard to -Xr, is quite sufficient to account 

 for the disappearance of magnetism. The author 

 should remember that the serious question is this, 

 " Does a flint hard allotropic iron exist? " The allo- 

 tropists are more happy than the carbonists in that 

 they have a crisp explanation for Arj, if the use of 

 another Greek letter as a prefix can be said to give 

 satisfaction to any practical worker in the field. The 

 solution theory is given in detail, but the author wisely 

 dismisses the application of the phase rule to the 

 problems of steel by reference to original papers, and 

 the student who endeavours to follow these is in need 

 of sympathy if he be well acquainted with the known 

 facts. 



The extremely difficult subject of the influence of 

 various elements on steels is well considered, but the 

 author implies that Le Chatelier was the pioneer in 

 our knowledge of sulphide in steel, while everyone 

 should know that that honour belongs to Prof. Arnold. 

 Heat treatment of steel is discussed in twenty-seven 

 pages, and the last chapter deals with the microscopical 

 examination of steel, and several methods of prepar- 

 ation and etching are well described, but Fig. 448 

 should either be altered to suit opaque objects or 

 removed. 



The volume closes with 100 photomicrographs, four 

 useful appendices, and a good inde.x. The structures 

 shown in several of the photomicrographs are not in 

 accord with the writer's experience, but that might be 

 due to abnormal crystallisation in the original steels, 

 which seem to be identical with those used for the 

 sixth report already mentioned. Several errors have 

 been noted, but there is only room to indicate a few 

 as examples : — p. 53, " wild metal which pipes in the 

 moulds"; p. 101, incorrect definition of a heat unit; 

 p. 227, the hardening power of liquids is said to be a 

 function of their specific heats, whereas their conduc- 

 tivities are more important, as witness mercury 

 compared with water; " microphotograph " all 

 through the work instead of photomicrograph. 

 On p. 680 0-8 carbon steel is indicated as 

 saturated, while on p. 681 it is o-g, and the footnote 

 to p. 6S4, " The latest research has shown that it 

 should be o-g," P'ives a wrong impression. It had no 

 need to be shown after 1895, ^^ Arnold made it quite 

 clear then, and others, perhaps working on impure 

 steels, claimed o-8, but now they have seen their error 

 and o-g is accepted almost universally. On tlie whole, 

 however, the book is to be recommended as the best 

 available on the metallurgy of steel. 



A. McWiLLIAM. 



