AhK'. 1 8, i8So] 



NATURE 



69 



exerted on a large mass of metal by a trace of another metal or 

 metalloid — that is, by a quantity so small that it appears to be 

 out of all proportion to the mass in which it is distributed. 



I think it may safely be asserted that in no other branch of 

 applied science has the operator to deal with quantities that are 

 at once s-o vast and so minute ; and the course will not have 

 proceeded far before you will recognise this fact. 



It may be that the trace to be extracted is alone of value — as, 

 for instance, the few grains of gold that can be profitably 

 extracted from each ton of a material, which, though containim; 

 only one part of gold in five millions by volume, is thereby 

 entitled to be regarded as an auriferous de|)osit that can be pr.ifit 

 ably worked ; or it may be the minute percentage of a metalloid 

 which must be extracted in order that the physical properties of 

 a large ma-s of metal may not be entirely altered. 



[Numerous instances of the influence of small traces of metals 

 and metalloids, including the following, were then given:—] 



In 1S66 Graham showed,^ by experiments with which I had 

 the privilege of being connected, that the presence of occluded 

 gases in metals often exerts a marked influence on their molecular 

 structure. In the case of iron he urged that metallurgists should 

 study the effects of occluded gases, more especially carbonic 

 oxide, the weight of which, according to his experiments, could 

 not exceed the jV per cent, of the weight of iron in which it was 

 present. The significance of such facts is now under considera- 

 tion by a Committee of the Institution of Mechanical Engineers," 

 and the question of the presence of gas in steel, either occluded 

 or retained in the form of bubbles, is further being investigated 

 by Chernoff,' Miiller,'' and others. 



M. Nyst, of tlie Brussels Mint, has lately found that the 

 presence of yW per cent, of silicon in standard gold will so affect 

 its molecular grouping as to render it pos-ible for a thin strip 

 to bend by its own weight, as zinc would, in the flame of a 

 candle. 



The growing importance of physical research in connection 

 with metallurgy is shown by the fact that physical metliods are 

 now constantly appealed to by those interested in metallurgy, 

 more especially in the case of iron and steel. We are told, for 

 instance, that the hardness of steel may be correctly inferred 

 from a numerical determination of its coercive force ; '^ it is 

 sought to establish the actual nature of the change in the mode 

 of existence of the carbon in steel that accompanies hardening by 

 determining its thermo-electric properties ; '' and the hope is held 

 out' to us tbat the time will soon come when boiler-makers will 

 electrically test their plates, possibly by the aid of the induction- 

 balance, just as they now test them for ductility and tenacity. I 

 can only add the expression of a belief that this powerful weapon 

 of molecular research which Prof. Hughes has given us will yield 

 good results in the hands of some of you. 



The results of mechanical tests are also of the highest impor- 

 tance. Not long since the appearance of the fracture of a 

 sample of metal was considered to afford trustworthy and suffi- 

 cient evidence as to its nature and properties ; but such rough 

 methods have given place, in the hands of Kirkaldy and others, 

 to the rigorous physical and mechanical investigation to which 

 metals must now be submitted as a matter of ordinary routine. 

 The results, tabulated or plotted into curves, which mark the 

 influence of each constituent or impurity, form permanent 

 records of the greatest value.* 



It has only been possible for me to indicate the more impor- 

 tant conditions affecting the successful practice of metallurgy. 

 I have traced the relation between technical and scientific 

 workers ; but there is yet another condition of somewhat recent 

 growth. The enormous scale on which operations are now con- 

 ducted renders it more necessary than formerly for those engaged 

 in metallurgical enterprise to seek the aid of capitalists. The 

 result is that a large share in the control of many important 

 works falls to the non-scientific members of the Board of Direc- 

 tors, men of high commercial ability, but whose knowledge of the 

 importance of scientific work is necessarily limited. It is true that 



<■ Phil. Trans. 1866, p. 43S. 



~ First Report of the Committee on the Hardening, Tempering, and 

 Annealing of Steel, 1879. 



3 "On the Structure of Cast Steel Ingots." Translated for the Institution 

 of Mechanical Engineers by W. Anderson, C.E (1879). 



** Berichtc der deiitschen chemisch^n Gcselhcha/t, 1879. No. xii. 93 : 

 Claser' s Annalenfiir Gewerbe und Bauwesen, August, i83o, p. 138. 



s Trcve and Durassier, Comp. rend., t. !xyx. (1875), p. 799 ; Watten 

 hofen, yournai n{ the Iron and Steel Institute, 1879, No. i, p. 305. 



6 Barus, P/iiV. Mag. [5], vol. viii. p. 341. 



^ W. H. Johnson, Chemical Ne'MS, vol. xUi. (i83o), p. 70. 



8 V. Deshayes, "Classement et Emploi des Aciers " (Paris, 1880); also 

 B71II. Chem. Soc. torn. xxxi. (1879), p. 16S. 



they may recognise the necessity for scientific aid in the works 

 with which they are connected, but they are too often unconscious 

 of the labour and difficulty that are involved in the attainment 

 of accurate scientific knowledge. I am convinced, however, 

 tliat facts are gradually compelling them to recognise that the 

 value of a metal may entirely depend on whether it does or does 

 not contain a trace of impurity, and that the exact method of 

 treatment to be adopted depends much on the character of the 

 materials employed ; they will therefore examine more carefully 

 than they have hitherto done the qualifications of men to whom 

 important duties are entrusted, and will insist that the services 

 of only adequately trained metallurgists shall be secured. 



I shall have to direct your attention to the minute care with 

 which details affecting commercial interests are now investi- 

 gated ; ^ and your success will further depend on the facility 

 wMth which you are able to use the ** tools of thought " furnished 

 by chemistry, physics, and mechanics. Whether you will ever 

 possess the tact and judgment necessary to direct such works as 

 Dowlais with an army of 10,000 people, obviously depends on 

 personal qualifications which I can but little influence. 



1 venture to hope that you will, by original research, add to 

 the general advance of science, for, as the late Prof. Clifford has 

 reminded us, what have often proved to be the most useful parts 

 of science have been investigated for the sake of truth, and not 

 for their usefulness. 



Dr. Percy found metallurgy practised in this country mainly 

 as an empirical art. He may well feel, to borrow the words of 

 an old writer, that in his hands " the business of metallurgy and 

 essaying has not only been illustrated but also improv'd, 

 amended, and enrich'd '' ; for his works contain a record of its 

 progress, his teaching and researches have secured it a scientific 

 basis, and he has trained a body of scientific workers, in whose 

 hands the immediate future of metallurgy to a great extent rests. 

 Bearing in mind how much the progress of our science means to 

 England, I cannot but be conscious that, in attemptmg to 

 continue this work, I undertake a grave responsibility. 



ON AN EXPERIMENTAL ILLUSTRATION OF 



MINIMUM ENERGY - 

 'T'HIS illustration consists of a liquid gyrostat of exactly the 

 same construction as that described and represented by the 

 annexed drawing, repeated from Nature, February i, 1877, 

 p. 297, 298, with the difference that the figure of the shell is 

 prolate instead of oblate. The experiment was in fact conducted 

 with the actual apparatus which was exhibited to the British 

 Association at Glasgow in 1876, altered by the substitution of a 



shell iiaving its equatorial diameter about -j\ of its axial diameter, 

 for the shell with axial diameter /j of equatorial diameter which 

 was used when the apparatus was shown as a successful gyrostat. 



I In illustration of this see an exhaustive mathematical paper on the values 

 of iron ores, by Prof. A, Habets : Cttyper's Revue Uiiiziersclle des Mines 

 (1877), t. i. p. 504. 



= BySirWilliam Thomson, F.R.S. British Association, Swansea, Section A. 



