Contributions fromThe Museum of History axd Technology:Paper 3 The Beginnings of Cheap Steel Philip If. Bishop STEEL BEFORE THE 1850s 29BESSEMER AND HIS COMPETITORS 30ROBERT MUSHET 33EBBW VALE AND THE BESSEMER PROCESS 35MUSHET AND BESSEMER 37WILLIAM Kelly's air-boiling process 42CONCLUSIONS 46 27471274?59 3 THE BEGINNINGS OF CHEAP STEEL Bv Philip W. Bishop Other inroitors claimed a part in the inven-tioii of the Bessemer process of making steel. Here,the contemporary discussion in the technical pressis re-examined to throw light on the relations ofthese various claimants to the iron and steel industryof their time, as having a possible connection withthe antagonism shown by the ironmasters towardBessemer' s ideas.The Author: Philip W. Bishop /s curator ofarts and manufactures. Museum of History and Tech-nology, in the Smithsonian Institution's UnitedStates National Museum. THK DEVELOPMENT of the world's i)rociuctivc re-sources during the 19th century, accelerated ingeneral by major innovations in the field of power,transportation, and irxiiles, was retarded by theoccurrence of certain bottlenecks. One of theseaffected the flow of suitable and economical raw-materials to the machine tool and transportationindustries: in spite of a rapid growth of iron produc-tion, the methods of making steel remained as theywere in the previous century; and outputs remained negligible.In the decade 1855-1865, this situation w-as com-pletely changed in Great Britain and in Europe generally; and when tiie United States emerged fromthe Civil War, that country found itself in a positionto take advantage of the European innovations andto start a [)eriod of growth which, in the next 50 years,was to establish lier as the world's largest producer of steel.This study reviews the controversy as to the originof the process which, for more than 35 years ' pro- ' From 1870 through 1907, "Bessemer" production ac-counted for not less than 50 percent of United States steelproduction. From 1880 through 1895, 80 percent of all steelcame from this source: Historical Statistics of the United .States1789-1945 (Washington, U. S. Department of Commerce,Bureau of the Census, 1949), Table- J. 165-170 at p. 187.28 BULLETIN 218: CONTRIBUTIONS FROM THE MUSEUM OF HISTORY AND TECHNOLOGY \idcd the greater part of the stt-i'l production of theI'nited States. It concerns four men for whom])riority of invention in one or more aspects of the])rocess has been claimed.The process consists in forcing through molten castiron, held in a vessel called a converter, a stream ofcold air under pressure. The combination of theo.xygen in the air with the silicon and carbon in themetal raises the temperature of the latter in a spec-tacular way and after "blowing" for a certain period,cHminates the carbon from the metal. Since steel ofvarious qualities demands the inclusion of from 0.15to 1.70 percent of carbon, the blow has to be terini-nated before the elimination of the whole carboncontent; or if the carbon content has been eliminatedthe appropriate percentage of carbon has to be putback. This latter operation is carried out by addinga precise quantity of manganiferous pig-iron (.spiegel-eisen) or ferromanganese, the manganese serving toremove the oxygen, which has combined with theiron during the blow.The controversy which surrounded its do\elo|3inentconcerned two aspects of the process: The use of thecold air blast to raise the temperature of the moltenmetal, and the application of manganese to overcomethe problem of control of the carbon and oxygencontent.Bessemer, who began his experiments in the makingof iron and steel in 1854, secured his first jjatent inGreat Britain in January 1855, and was persuaded topresent information al)out his discovery to a meetingof the British Association for the Advancement ofScience held at Cheltenham, Gloucestershire, in .?\ugust 1856. His title "The Manufacture of Ironwithout Fuel"" was given wide jiublicity in GreatBritain and in the United States. Among those whowrote to the papers to contest Bessemer's theorieswere several claimants to ])riority of invention.Two men claimed that they had antici|3ated Bes-semer in the invention of a method of treating moltenmetal with air-blasts for the purpose of "purifying" ordecarbonizing iron. Both were Americans. JosephGilbert Martien, of Newark, New Jersey, who at thetime of Bessemer's address was working at the plantof the Ebbw \'alc Iron Works, in South Wales,secured a provisional patent a few days before Bes-semer obtained one of his series of patents for makingcast steel, a circumstance which provided ammunitionfor those who wished to dispute Bessemer's somewhatspectacular claims. William Kelly, an ironmaster olEddyville, Kentucky, brought into action by an P.A,PER j: BEGINNING.S OF CHE.M' .SIKIX .\inerican report of Bessemer's British A.ssocialionpaper, opposed the granting of a United Statespatent to Bessemer and substantiated, to the satis-faction of the C^ommissioner of Patents, his claim topriority in the "air boiling" process.A third man, this one a Scot resident in England,intervened to claim that he had devised the meanswhereby Martien's and Bessemer's ideas could \x:made practical. He was Robert Mushet of Coleford,Gloucestershire, a metallurgist and self-appointed "sage" of the British iron and steel industry who abowas associated with the Ebbw Vale Iron Works as aconsultant. He, like his American contemporaries,has become established in the pui)lic mind as oneupon whom Henry Bessemer was dependent for theorigin and success of his process. Since Bessemer wasthe only one of the group to make money from theexpansion of the steel industry consequent upon theintroduction of the new technique, the suspicion hasremained that he cxijloited the inventions of theothers, if indeed he did not steal them.In this study, based largely upon the contemporarydiscussion in the technical press, the relation of thefour men to each other is re-examined and an attemjjtis made to place the controversy of 1855-1865 infocus. The necessitv for a reapprai.>al arises from thefact that today's references to the origin of Bessemer steel ' often contain chronological and other inac-curacies arising in many cases from a dependence onsecondary and sometimes unreliaiile sources. .As a result, Kelly's contribution has, perhaps, been over-emphasized, with the effect of derogating from thework of another .American, .\lexander Lyman Holley,who more than any man is entitled to credit forestablishing Bessemer steel in .\merica.' Steel Before the 1850's In spite of a rapid increase in the use of machines andthe overwhelming demand for iron products for theexpanding railroads, the use of steel had expanded -See especially mateiial distributed by the American Ironand Steel Institute in connection with its celebration of thecentennial of Steel: "Steel centennial (19S7), prcia informa-tion,"' prepared by Hill and Knowllon. Inc., and released bythe Institute as of May 1, 1957.5 Holley's work is outside the scope of this paper. Belatedly,his biography is now beinp written. It can hardly fail to sub-stantiate the contention that during his short life (1832-1882)Holley, who negotiated the purchase of the .American rightsto Bessemer's process, also adapted his methods to the .Americanscene and laid a substantial part of the foundation for themodern .American steel industry. 29 litilf prior lo 185i>. The im-lhoils ol production wen- still lari^cly those of a century earlier. Slow prepara-tion of the steel by cementation or in cruciljlcs meanta disproportionate consumption of fuel and a resultinghigh cost. Production in small ciuantities preventedthe adoption of steel in uses wltich required largeinitial masses of metal. Steel was. in fact, a luxuryproduct.The work of Reaumur and, especially, of Huntsman,whose development of cast steel after 1740 secured aninternational reputation for Sheffield, had establishedthe cementation and crucible processes as the primarysource of cast steel, for nearly 100 years. JosiahMarshall Heath's patents of 1839, were the first devel-opments in the direction of cheaper steel, his processleading to a reduction of from 30 to 40 percent in theprice of good steel in the Sheffield market. ' Heath'ssecret was the addition to the charge of from 1 to 3percent of carburet of manganese ^ as a deoxidizer.Heath's failure to word his patent .so as to cover alsohis method of producing carburet of manganese led tothe efTective breakdown of that patent and to thegeneral adoption of his process without payment oflicense or royalty. In sjiile of this reduction in thecost of its production, steel remained, until after themidpoint of the century, an insignificant item in theoutput of the iron and steel industry, being used prin-cipally in the manufacture of cutlery and edge tools.The stimulus towards new methods of making steeland, indeed, of making new steels came curiouslyenough from outside the established industry, from aman who was not an ironmaster?Henry Bessemer.The way in which Bes.semcr challenged the trade wasitself unusual. There are few cases in which a strangerto an industry has taken the risk of giving a descriptionof a new process in a pul)lic forum like a meeting of theBritish Association for the Advancement of Science.He challenged the trade, not only to attack his theoriesbut to produce evidence from their own plants thatthey could provide an alternative means of satisfyingan emergent demand. Whether or not Bessemer isentitled to claim priority of invention, one can butagree with the ironmaster who said: " "Mr. Bessemerhas raised such a spirit of en(|uiry throughout . . . the ' Andrew Ure, Dictionary of arts, manujacturrs and mines. NewYork, 1856, p. 735. '.Sec abridgement of British patent 8021 of 1839 quoted byJames S. Jeans, Steel, London, 1880, p. 28 ff. It is not clearthat Heath was aware of the precise chemical effect of the use ofmanfianesc in this way. ? Mining Journal, 1857, vol. 27, p. 465. land as nuist Icid lo an improved syslciii ol manu-facture.''Bessemc-r and his CompetitorsHenry Bessemer (1813-1898), an Englishman ofFrench extraction, was the son of a mechanicalengineer with a special interest in mctalhugy. I lisenvironment and his unusual ability to synthesize hisob.servation and experience enabled Bessemer to begina career of invention by registering his first patent atthe age of 2.S. His active experimenting continued uniil his death, although the public record of his results ended wiih a patent issued on the day beforehis .seventieth birthday. A total of 117 Britishpatents " bear his name, not all of them, by any means,successful in the sense of producing a substantialincome. Curiously, Bessemer's financial stability wasassured by the success of an invention he did notpatent. This was a process of making bronze powderand gold paint, until the 1830's a secret held inGermany. Bessemer's substitute for an expensiveimported product, in the then state of the patent laws,would have failed to give him an adequate reward ifhe had been unable to keep his process secret. Toassure this reward, he had to design, assemble, andorganize a plant capable of operation with a minimmnof hired labor and with close .security control. Thefact that he kept the method secret for 40 years,suggests that his machinery * (Bessemer describes it as virtually automatic in operation) represented anappreciation of coordinated design greatly in advanceof his time. His experience must have directly con-tributed to his conception of his steel process not as ametallurgical trick but as an industrial process; forwhen the time came, Bessemer patented his discoveryas a process rather than as a formula.In the light of subsequent developments, it isnecessary to consider Bessemer's attitude toward thepatent privilege. He describes his secret gold paintas an example of "what the public has had to pay fornot being able to give . . . security to the inventor"in a situation where the production of the material "could not be identified as having been made by anyparticular form of mechanism.'"* The inability toobtain a patent over the method of production meantthat the disclosure of his formula, necessary for patentspecification, would openly in\ile competitors, in- ~ Sir Henry Bessemer, F. R. S., an autobiography, London. 1905,p. 332. ' Ibid., p. 59 fir. ? Ibid., p. 82.30 BULLETIN 218: CONTRIBUTIONS FROM THK MUSLIM OK HISTORY .XND TECHNOLOGY chidina; the Germans, to evolve their own teehnic|iies.Bessemer coneludes: '"Had the invention been paK'nled, it would have becomepublic property in fovnieen years from the dale of thepatent, after which period the public would have beenable to buy bronze powder at its present [i.e., ca. i8go]market i)rice, viz from two shillings and three pence totwo shillings and nine pence per pound. But this important secret was kept for about ihirly-five years and the |)ublichad to pay excessively high prices for twenis-one yearslonger than they would have done had the invention becomepublic propert)' in fourteen years, as it would have been ifpatented. Even this does not represent all the disad-vantages resulting from secret manufacture. While everydetail of production was a profound secret, there were noimprovements made by the outside public in any one of themachines employed during the whole thirty-five years;whereas during the fourteen years, if the invention had beenpatented, there would, in all probability have been manyimproved machines invented and many novel featuresapplied to totally different manufactures.\\ hile these words, to some extent, were the rational-izations of an old man, Bessemer's career showed thathis philosophy had a practical foundation; and, ifthis was indeed his belief, the cpi.sode explains inlarge measure Bessemer's later insistence on thelegal niceties of the patent procedure. The effect ofthis will be seen.Bessemer's intervention in the field of iron and steel was preceded by a period of experiments in themanufacture of glass. Here Bessemer claims to havemade glass for the first time in the open hearth of areverberatory furnace." His work in glass manu-facture at least gave him considerable experience inthe problems of fusion under high temperatures andprovided some support for his later claim that inapplying the reverberatory finnace to the manufactureof malleable iron as describetl in his first patent ofJanuary 185.S, he had in some manner anticipatedthe work of C. W. Siemens and F.niil Martin. '- '? Ibid., p. 83. " Ibi ' * -?^? r f I'igurc I. ? Bessemer's Design for a Converter, as Shown in I . S Patent 16082. Thispatent, dated November 11, 1856, corresponds with British patent 356. dated February 12,1856. The more familiar design of converter appeared first in British patent 578, March i,i860. The contrast with Kelly's schematic drawing in Fig. 2 (p. 42) is noticeable. on August 13, 1856.'** Bessemer described his firstconverter and its operation in some detail. .\ltliiniL;iihe was soon to realize that he '"too readih allowedmyself to bring my inventions tinder public notice,"'"Bessemer had now thrown out a challenge whicheventually had to be taken up, regardless of thestrength of the vested interests involved. The prov- '^ Bcsscmcr's paper was reported in 'ihf Times, London,August 14, 1856. By the time the Transactions of the BritishAssociation were prepared for publication, the controversyaroused by Bessemcr's claim to manufacture "malleable ironand steel without fuel" had broken out and it was decidednot to report the paper. Dredge (o/<. cil., footnote 1.S, p. 915)describes this decision as "sagacious." '? B?-ssemer, op. cil. (footnote 7), p. 164. ocation came I'roin his claims that the [jrodiict of thefirst staye of tlie conxersion was the equivalent olcharcoal iron, the processes following the smeltiniii)eing conducted without contact with, or the use ot.any mineral fuel: and that further blowing could beused to produce an\' quality of metal, that is, a steelwith any desired percentage of carbon. Yet, theprincipal irritant 10 the complacency of the iron-master must have been Bessemcr's attack on anindustry which had gone on increasing the size of itssmelting furnaces, thus improving the uniformity ofits pig-iron, without modifying the puddling process,which at best could handle no more than 400 to 500pounds of iron at a time, divided into the "homeo-32 BULLETI.N 218: CONTRIBITIONS FROM THE MUSEl'M OF HISTORY AND TECHNOLOGY pathic doses" of 70 or 80 pounds capable of beinghandled by human labor.-" Bcsscmcr's claim to "do" 800 pounds of metal in 30 minutes against thepuddlins; furnace's output of 500 pounds in twohours was calculated to arouse the opposition of thosewho feared the loss of capital invested in puddlingfurnaces and of those who suspected that their jobsmight be in jeopardy. The ensuing criticism ofBessemer has to be interpreted, therefore, with thisin mind; not by any means was it entirely based onobjective consideration of the method or the product.-'Within a month of his address, Bessemer had soldlicenses to several ironmasters (outside Sheffield) andso provided himself with capital with which to con-tinue his development work; but he refused to sell hispatents outright to the Ebbw Vale Iron Works andby this action, as will be seen, he created an enemyfor himself.The three years between 1856 and 1859. whenBessemer opened his own steel works in Sheffield, wereoccupied in tracing the causes of his initial difficulties.There was continued controversy in the technicalpress. Bessemer (unless he used a nom-de-plume) tookno part in it and remained silent until he madeanother public appearance before the Instituii(jn ofCivil Engineers in London (May 1859). By this timeBessemer's process was accepted as a practical one,and the claims of Robert Mushet to share in hisachievement was becoming clamorous.ROBERT MUSHETRobert (Forester) Mushet (1811-1891), born in theForest of Dean, Gloucestershire, of a Scots father(David, 1772-1847) himself a noted contributor tothe metallurgy of iron and steel, is, like the .AmericanWilliam Kelly, considered by many to ha\"c been avictim of Bessemer's astuteness?or villainy. Becauseof Robert Mushet's preference for the quiet of Cole-ford, many important facts about his career are lack-ing; but even if his physical life was that of a recluse,his frequent and verbose contributions to the corre-spondence columns of the technical press made himwell-known to the iron trade. It is from these lettersthat he must be judged.In view of his propensity to intervene pontifically in every discussion concerning the manufacture of ironand steel, it is somewhat surprising that he refrainedfrom comment on Bessemer's British A.ssociaiionaddress of August 1856 for more than fourteen months.The debate was opened over the signature of hisbrother David who shared the family facility withthe pen.^^ Recognizing Bessemer's invention as a "congruous appendage to [the] now highly developedpowers of the blast furnace" which he describes as "tooconvenient, too powerful and too capable of furtherdevelopment to be superseded by any retrogradeprocess," David Mushet greeted Bessemer's discoveryas "one of the greatest operations ever devised inmetallurgy." " A month later, however, DavidMushet had so modified his opinion of Bessemer as tocome to the conclusion that the latter "must indeed beclassed with the most unfortunate inventors." Hegave as his reason for this turnabout his discovery thatJoseph Martien had demonstrated his process of "purifying" metal successfully and had indeed beengranted a provisional patent a month before Bessemer.The sharp practice of Martien's patent lawyer, Mushetclaimed, had deprived him of an opportunity ofproving priority of invention against Bessemer.Mushet was convinced that Martien's was the first inthe field.-'Robert Mushet's campaign on behalf of his ownclaims to have made the Bessemer process effectivewas introduced in October 1857, two years after thebeginning of Bessemer's experiment and after oneyear of silence on Bessemer's part. Writing as "Sideros" -' he gave credit to Martien for "the great =0 The Times, London, .AuRiist 14, 1856. -' David Mushet rt-cognizcd that Bossi-mor's great featurewas this effort to "raise the after processes ... to a levelcommensurate with the preceding case"" (Mining Journal,1856, p. 599). " See Mining Journal, 1857, vol. 27, pp. 839 and 853. DavidMushet withdrew from the discussion after 1858 and his relapseinto obscurity is only broken by an appeal for funds for thefamily of Henry Cort. A biographer of the Mushets is of theopinion that Robert Mushet wrote these letters and obtainedDavid's signature to them (Fred M. Osborn, The slory of theMushels, London, 1952, p. 44, footnote). The similarity in thestyle of the two brothers is extraordinary enough to support thisidea. If this is so. Robert Mushet who disagreed with himselfas "Sideros"' was also in controversy with himself writing as "David." -' Mining Journal, 1856, vol. 26, p. 567. -* Ibid., pp. 631 and 647. The case of Martien will be dis-cussed below (p. 36). David Mushet had overlooked Bessemer'spatent ofJanuary 10, 1855. -' Mining Journal, 1857, vol. 2"", p. 723. Robert Mushet wasa constant correspondent of the Mining Journal from 1848. Theadoption of a pseudonym, peculiar apparently to 1857-1858(see Dictionary oj national biography, vol. 39, p. 429), enabledhim to carry on two debates at a time and also to sing his ownpraises. PAPER .3: BEGINNINGS OF CHEAP STEEL discovery that pig-iron can, whilst in the (luid slate, bepurified ... by forcing currents of air underit . . . ," though Marticn had failed to observe theuse of temperature by the "dellation of the ironitself;and for discovering that ? when the carbon has been all. or nearly all. dissipated,the temperature increases to an almost inconceivable extent,so that the mass, when containing only as much carbon as is requisite to constitute with it cast steel . . . still retains aperfect degree of fluidity.This, says "Sidcros," was no new observation; "it hadbeen before the metallurgical world, both practicaland scientific, for centuries," but Bessemer was thefirst to show^ that this generation of heat could beattained by blowing cold air through the melted iron.Mushet goes on to show, however, thai the steel thusproduced by Bessemer w-as not commercially valuablebecause the sulphur and phosphorous remained, andthe dispersion of oxide of iron through the mass "imported to it the inveterate hot-short quality whichno subsequent operation could e.\pel." '"Sideros"concludes that Bessemer's discovery was "at least fora time" now shelved and arrested in its progress; andit had been left "to an individual of the name ofMushet" to show^ that if "fluid metallic manganese"were combined with the fluid Bessemer iron, the por-tion of manganese thus alloyed would unite with theoxygen of the oxide and pass off as slag, removing thehot-short quality of the iron. Robert Mushet haddemonstrated his product to "Sideros" and hadpatented his discovery, though "not one print, literaryor scientific, had condescended to notice it." "Sideros" viewed Mushet's discovery as a "sparkamongst dry faggots that will one day light up a blazewhich will astonish the world when the unfortunatein\^entor can no longer reap the fruits of his life-longtoil and unflinching perseverance." In an ensuingletter he ''^ summed up the situation as he saw it:Nothing that Mr. Mushet can hereafter invent can entitlehim to the merit of Mr. Bessemer's great discovery . . .and . . . nothing that Mr. Bessemer may hereafter patentcan deprive Mr. Robert Mushet of having been the first toremove the obstacles to the success of Mr. Bessemer'sprocess.Bessemer still did not intervene in the newspaper discussion; nor had he had any serious supporters, atleast in the early stasje.""Publication in the Mining Journal o( n list of Mushet'spatents,"* evidently in response to .Sideros" complaint,now presented Bes.senier with notice of RobertMushet's acti\it\. e\en if he had not already observedhis claims as they were presented to the Patent Ollice.Mushet, said the .Mirii>ig Journal? appears to inlciid to cany on his researches from thepoint where Mr. J. G. Marticn left otf and is proceedingon the Bessemer plan of patenting each idea as it occurs tohis imaginative brain. 1 Ic proposes to make both iron and steel but does not appear to have quite decided as to thecourse of action ... to accomplish his object, and thereforeclaims various processes, some of which are never likely to realize the inventor's expectations, although decidedlynovel, whilst others are but slight modification of inventionswhich have already been tried and failed.The contemporary attitude is reflected in anothercomment by the Mining Journal:''^Although the application of chemical knowledge tothe manufacture of malleable iron cannot fail to producebeneficial results, the quality of the metal depends moreupon the mechanical than the chemical processes . . .Without wishing in any way to discourage the iron chemists,we have no hesitation in giving this as our opinion which we shall maintain until the contrary be actually proved. Withregard to steel, there may be a large field for chemicalresearch . . . however, we believe that unless the iron beof a nature adapted for the manufacture of steel by ordinaryprocesses, the purely chemical inventions will only give ametal of a very uniform quality. .\nother correspondent. William Green, was of theopinion that Mushet's "new compoimds and alloys,"promised well as an auxiliary to the Bessemer processbut that "the evil which it was intended to remove wasmore visionar\' than real." Bessemer's chief dillicultywas the phosphorus, not the oxide of iron "as Mr.Mushet assumes." This, Bessemer no doubt wotdddeal with in due course, but meanwhile he did well "to concentrate his energies upon the steel opcra- *' Ibid., p. 823. Mushet's distinction between an inventorand a patentee is indicative of the disdain of a son of DavidMushet for an amateur (see also p. 886). -" One William Green had commented extensively on DasidMushet's early praise of the Bessemer process and on hissudden reversal in favor of Martien soon after Bessemer's BritishAssociation address {Mechanics' Magazinf, 1856, vol. 65, p. 373ff.l. Green wrote from Caledonian Road, and the proximityto Baxter House, Bessemer's London headquarters, su?;gests thepossibility that Green was writing for Bessemer.2' Mtmns; Journal, 1857, vol. 27, p. 764. =? Ibid , p. 764. 34 BULLETIN 218: CONTRIBUTIONS FROM THE MUSEUM OI- HISTORY AND TECHNOLOGY lions," after which he wimld have time to tackle "thedifficiihies which have so far retarded the ironoperations." '"'Mushet ^' claims to have taken out his ])atent ofSeptember 22, 1856, co\-crino- the famous "triplecompound," after he?had fully ascenaincd, upon the ordinary scale of manu-facture that air-purified cast-iron, when treated as set forthin my specifications, would afford tough malleable iron . . . I found, however, that the remelting of the coke pig-iron,in contact with coke fuel, hardened the iron too much,and it became evident that an air-furnace was more properfor my purpose . . . [the difficulties] arose, not from anvdefect in my process, but were owing to the small quantityof the metal operated upon and the imperfect arrangement ofthe purifying vessel, which ought to be so constituted thatit may be turned upon an axis, the blast taken off, the alloyadded and the steel poured out through a spout . . , Sucha purifying vessel Mr. Bessemer has delineated in one of his patents.Mushet also claimed to have designed his own "purifying and mixing" furnace, of 20-ton capacity,which he had submitted to the Ebbw \'ale IronWorks "many months ago, " without comment fromthem. There is an intriguing reference to the pain-ful subject of two patents not proceeded with, andnot discussed "in the a\aricious hope that the partiesconnected with the patents will make me honorableamends . . . these patents were suppressed withoutmy knowledge or consent. " Lest his qualificationsshould be cjuestioned, Mushet concludes: I do not profess to be an iron chemist, but I have un-doubtedly made more experiments upon the subject ofiron and steel than any man now living and I am therebyenabled to say that all I know is but little in comparisonwith what has yet to be discovered.So began Mushet's claim to have solved Be.s.semer'sproblem, a claim which was to fill the correspondencecokmnis of the engineering jotirnals for the next tenyears. Interpretation of this correspondence is madedifficult by our ignorance of the facts concerning thecontrol of Mushet's patents. These have to be piecedtogether from his scattered references to the subject.His experiments were conducted, at least nearly upto the close of the year 1856, with the cooperationof Thomas Brown of the Ebbw Vale Iron Works. '"The price of this assistance was apparently half in-terest in Mushet's patents, though for reasons whichMushet does not explain the deed prepared to effect "/AiV., p. 791. " IhiJ., p. 770 (italics supplied). " Ihid., p. 770.PAPER .3: BEGINNINGS OF CHE.\P STEEL471274?59 4 the transfer was never executed," Mushet contin-ued, however, to regard the patents as "wholly myown, though at the same time, I am bound in honorto take no unfair advantage of the non-execution ofthat deed." .\ possible explanation of this situationma\- be found in Ebbw \'ale's activities in connectionwith Martien and Bessemer, as well as with anAustrian inventor, Uchatius,Ebbw Vale and the Bessemer Process After his British .^s.sociation address in .\ugust1856, Be-ssemer had received applications from severalironmasters for licenses, which were issued in returnfor a down payment and a nominal royalty of 25pence per ton. ,\niong those who started negotia-tions was Mr. Thomas Brown of Ebbw Vale IronWorks, one of the largest of the South Wales plants.He proposed, however, instead of a liccn.se, an out-right purchase of Bessemer's patents for ?50,000.Bessemer refused to .sell, and according to his '* ac-count?intense disappointment and anger quite got the better of[Brown] and for the moment he could not realize the factof my refusal . . . [He then] left me very abruptly, sayingin an irritated tone . . . "I'll make you see the matterdifferently yet" and slammed the door after him.David Mushet's advocacy of Martien's claim topriority over Besseiner has already been noticed(p. 33). From him we learn '^ that Martien's cxf>eri-ments leading to his patent of September 15, 1855,had been carried out at the Ebbw Vale Works inSouth Wales, where he engaged in "jx'rfectins; theRenton process." "' Martien's own process consistedin passing air through metal as it was run in a troughfrom the furnace and before it passed into the puddlingfurnace.It is known that Martien's patent was in the handsof the Ebbw \'ale Iron Works by March 1857.^" Thisfact nnist be added to our knowledge that Mushet'spatent of September 22. 1856 was drawn up with a specific reference to the application of his "triple " Ihtd., p. 823, "Bessemer, op. cit. (footnote 7), p. 169.5' Mining Journal, 1856, vol. 26, p. 631. '"James Rcnton's process (U. S. patent 8613. December 23,1851) had been developed at Newark, Now Jersey, in 1854,It wa? a modification of the puddlini; furnace, in which theore and carbon were heated in tubs, utilizing the waste heatof the revcrbcratory furnace (sec the Mechanics' .\taga-ine, vol. 62, p. 246, 1855). Ronton died at Newark in September1856 (Mechanics' Maga-int, 1856, vol. 65, p. 422). " Mining Journal, 1857, vol. 27, p. 193, J-T compound" to "iron . . . puriiied by the action of air, in the manner invented l)y Joseph GilbertMartien," '^ and that this and his other manganesepatents were imder the elVective control of Eblnv \'ale.It seems a reasonable deduction from these circum-stances that Brown's ofler to buy out Bessemer and hissubsequent threat were the consequences of a deter-mination by Ebbw \'ale to attack Bi'sscnier by meansof patent infringement suits.Some aspects of the Ebbw \ ale siuiaiion are notyet explained. Martien came to South Wales fromNewark, New Jersey, where he had been manager ofRenton's Patent Semi-Bituminous Cloal Furnace,owned by James Quimby, and where he had some-thing to do with the installati(jn of Renton's firstfurnace in 1854. The first furnace was unsuccessful.^'Martien next appears in Britain, at the Ebbw \'aleIron Works. No information is available as to whetherMartien's own furnace was actually installed at Ebbw\'ale, although as noted above, I)a\id Mushet claimsto have been in\iied to see it there.Martien secured an .American patent for his processin 1857 and to file his application appears to havegone to the United States, where he remained at least until October 1858.''" He seems to have taken theopportunity to apply for another patent for a furnacesimilar to that of James Renton. This led to inter-ferences proceedings in which Martien showed thathe had worked on this furnace at Bridgend, Glamor-ganshire (one of the Ebbw Vale plants), improvingRenton's design by increasing the number of "de-oxydizing tubes." This variation in Renton's designwas held not patentable, and in any case Renton'sfirm was able to show that they had successfullyinstalled the furnace at Newark in 1852-1853, whileMartien could not satisfy the Commissioner that hisinstallation had been made before September 1854.Priority was therefore awarded to Quimby. Brown,Renton, and Creswell.'" ?? BritLsh patent 2219, September 22, 1856. "Joseph P. Lesley, I'he iron manufacturer's guide. New York,1859, p. 34. Martien's name is spelled Martcen. A descrip-tion of the furnace is given in Scientific American of February 11,1854, (vol. 9, p. 169). In the patent interference proceedingsreferred to below, it was stated that the furnace was in success-ful operation in 1854. *" U. S. patent 16690, February 22, 1857. A correspondentof the Mining Journal (1858, vol. 28, p. 713) states that Martienhad not returned to England by October 1858. *' U. S. Patent Office, Decision of Commissioner of Patents,dated May 26, 1859 in the matter of interference between theapplication ofJames M. Quimby and others . . . and of JosephMartien. Since Renton had not patented his furnace inGreat Britain, Martien's use of his earlier knowledgeof Renton's work and of his experience at Bridgendin an attempt to upset Renton's priority is a curiousand at present unexplainable episode. Perhaps theearly records of the Ebbw Vale Iron Works, if they exist, will show whether this episode was in someway linked to the firm's optimistic combination ofthe British patents of Martien and Mushet.That Ebbw Vale exerted every effort to find analternative to Bessemer's process is suggested, also,iiy their purchase in 1856 of the British rights to theUchatius process, invented by an .Austrian Army officer. The provisional patent specifications, datedOctolaer 1. 1855, showed that Uchatius proposed toinake cast steel directly froin pig-iron by meltinggranulated pig-iron in a crucible with puKerized "sparry iron" (siderite) and fine clay or with grayoxide of manganese, which would determine theamount of carbon combining with the iron. Thisprocess, which was to prove commercially successfulin Great Britain and in Sweden but was not u.sed in .America,*" appeared to Ebbw \'ale to be somethingfrom which, "we can ha\e steel produced at the [jriceproposed by Mr. Bessemer, notwithstanding thefailure of his process to fulfil the promise." *^So far as is known only one direct attempt wasmade, presumably instigated by Ebbw \'ale, toenforce their patents against Bessemer, who records "a visit by Mushet's agent some two or three monthsbefore a renewal fee on Mushet's basic manganesepatents became payable in 1859. Bessemer "entirelyrepudiated" Mushet's patents and ofTered to performhis operations in the presence of Mushet's lawyersand witnes.ses at the Sheffield \\'orks so that a prose-cution for infringement "would be a very simplematter." That, he says, was the last heard from theagent or from Mushet on the subject.*' The renewalfee was not paid and the patents were thereforeabandoned by Kl)bw \'aie and their associates, a - J. S.Jeans, op.cil. (footnote 5), p. 108. The process is notmentioned by James M. Swank, History oj the manujaclure ofiron in all ages, Philadelphia. American Iron and .Steel Asso-ciation, 1892. " Mining Journal, 1856, vol. 26, p. 707. *' Bessemer, op. cil. (footnote 7), p. 290. '^ The American Iron and Steel Institute's "Steel centen-nial (1957) press information" (see footnote 2), includes apamphlet, "Kelly lighted the fireworks . . ." by VaughnShclton (New York, 1956), which asserts (p. 12) that Bessemerpaid the renewal fee and became the owner of Mushet's "vital" patent.36 BULLETIN 218: CONTRIBUTIONS FROM THE MUSEUM ()! HIS!<)R^ ANU TECHNOLOGY fact which did not come lo Mushet's knowledge uniil1861. when he himself declared that the patent ''wasnever in my hands at all [so] that I could not enforceit." '"Further support for the thesis that Ehbw Vale'spolicy was in part dictated ljy a desire to make Bes-semer "see the matter difl'erently"" is lo be found inthe climatic episode. Work on Marlien's patentshad not been abandoned and in 1861 certain patentswere taken out by George Parry, Kbliw Vale'sfurnace manager. These, represented as improve-ments oi Martien's designs, were regarded by Bes-semer as clear infringements of his own patents."When it came to Bes.semer's knowledge that Ehbw\'ale was proposing to "go to the public" for addi-tional capital with which to finance, in part, a largescale working of Parry's process, he threatened thefinancial promoter with injunctions and succeededin opening negotiations for a settlement. All thepatents "which had been for years suspended" overBessemer were turned over to him for ?30,000.Ebbw Vale, thereupon, issued their prospectus **with the significant statement that the directors "have agreed for a license for the manufacture of steel by the Bessemer process which, from the peculiarresources they possess, they will be enabled to producein very large quantities. . . ." So Bessemer becametheownerof the Martien and Parry patents. Mushet'sbasic patents no longer existed.Mushct and BessemerThat Mushet was "used" by Ebbw \'ale againstBessemer is, perhaps, only an assumption; but that hewas badly treated by Ebbw Vale is subject to nodoubt. Mushet's business capacity was small but itis difficult to believe that he could have been so foolishas to assign an interest in his patents to Ebbw \'alewithout in some way insuring his right of consultationabout their disposition. He claims that even in thedrafting of his specifications he was obliged to followthe demands of Ebbw \'ale, which firm, believing, "on the advice of Mr. Hindmarsh, the most eminentpatent counsel of the day," " that Martien's patentoutranked Bessemer's, insisted that Mushet link his process to Martien's. This, as late as 1861, Mushetbelieved to be in effective oiieration.'" His laterrepudiation of the process as an absurd and impracti-cable patent process "possessing neither value nor utilit\ " ^' may more truly represent his opinion, espe-cialK as, when he wrote his 1861 comment, he stilldid not know of the disappearance of his patents.Mushet's boast '-' that he had never been into anironworks other than his own in Cloleford is a clue tothe interpretation of his behavior in general and alsoof his frecjuent presumptuous claims. \Vhen, for in-stance, the development of the L'chatius process waspublicized, he gave his opinion ^' that the process wasa useless one and had been patented before L'chatius "understood its nature"; yet later ^' he could claimthat the jirocess was "in fact, m\' own invention and I had made and sold the steel thus produced for someyears previously to the date of Captain L'chatius'patent." Moreover, he claims to have instructedL^chatius' agents in its operation! He may, at thislater date, have recalled his challenge (the first ofmany such) in which he offered Uchatius' agent inEngland to pay a monetary penalty if he could notshow a superior method of fjroducing "sound service-able cast steel from British coke pig-iron, on lite slomicplan and without any mi.xture of clay, oxide of man-ganese or any of these pot destroying ingredients." ^^It was David Mushet (or Robert, using hisbrother's name)''" who accused Bessemer, or rather hispatent agent, Carpmael, of sharp practice in connec-tion with Martien's specification, an allegation latersupported by Martien's first patent agent, Avery."The story was that for the drafting of his final specifica-tion, Martien, presumably with the advice of the EbbwVale Iron Works, consulted the same Carpmael, as "the leading man" in the field. The latter advisedthat the provisional specification restricted Martiento the application of his method to iron flowing in achannel or gutter from the blast furnace, and soprevented him from applying his aeration principlein any kind of receptacle. In effect, Carpmael was *' Robert Mushct, //? Bessemer-Mushcl process, Chcllcnhaiii,1883, p. 24; The Engineer, \?,b\, vol. 12, pp. 177 and 189. " The Engineer, 1862, vol. 14, p. 3. Bessemer, op. cil. (foot-note?), p. 296. *' Mining Journal, 1864, vol. 34, p. 478. " The Engineer, 1861, vol. 12, p. 189. =0 Ibid., p. 78. " Mushet, op. cit. (footnote 46), p. 9. " Ibid., p. 25. " .Mining Journal, 1857, vol. 27, p. 755. '* Mushet, op. eil. (footnote 46), p. 28. The Uchatius processbecame the "You-cheat-us" process to Mushet {.Mining Journal,1858, vol. 28, p. 34). *' .Mining Journal, 1857, vol. 27, p. 755 (italics supplied). " See footnote 22. " Mining Journal, 1856, vol. 26, pp. 583, 631. P.APER 3: BEGINNINGS OF CHE.AP STEEL 37 actinij unprorcssionally by giving Bessemer the |5iiorclaim to the use of a receptacle. According toMusliet, Marticn had in fact "actually and pviblich-proved" his process in a receptacle aiui not in agutter, so that his claim to priority could be main-tained on the basis of the provisional specification.This, like other Mushet allegations, was ignored byBessemer, and probably with good reason. At any rate, Martien's American patent is in terms similarto those of the British specification; he or his advisersseem to have attached no significance to the dis-tinction between a gutter and a receptacle.Mushet's claim to have afforded Bessemer themeans of makine: his own process useful is still sub-ject to debate. Unfortunately, documentation of thecase is almost wholly one sided, since his biggestpublicizer was Mushet himself. An occasional edi-torial in the technical press and a few replies toMushet's "lucubrations" are all the material which exists, apart from Bessemer's own story.Mushet and at least five other men patented theuse of manganese in steel making in 1856; his ownprovisional specification was filed within a month ofthe pui)licatioii of Bessemer's British .Association ad-dress in August 1856. So it is strange that RobertMushet did not until more than a year later join inthe controversy which followed that address. '*''* Inone of his early letters he claims to have made of "his"' steel a bridge rail of 750 pounds weight; al-though his brother insists that he saw the same railin the Ebbw Vale offices in London in the spring of1857, when it was presented as a specimen of Uchatiussteel!" Robert Mushet's indignant "advertisement"'of January 5, 1858,"" reiterating his parentage of thissample, also claimed a double-headed steel rail "made by me under another of my patent processes,"and sent to Derby to i)e laid down ihcrc lo be "sub-jcted to intense vertricular triturations." Mushet's de-scription of the preparation of this ingot "' shows thatit was derived from "Bessemer .scrap" made by Ebbw^Vale in the first unsuccessful attempts of that firmto simulate the Bessemer process. This .scrap Mushethad remelted in pots with spiegel in the proportionsof 44 pounds of scrap to ?> of melted spiegel. It was his claim that the rail was rolled direct from theingot, soniethinu; Bessemer himself could not do atthat time.This was the beginning of a series of claims byMushet as to his e.s.sential contributions to Bessemer'sinvention. The silence of the latter during this jieriodis impressive, for according to Bessemer's own ac-count " his British Association address was premature,and although the sale of licenses actually provided himwith working funds, the impatience of those experi-menting with the process and the flood of competing "inventions" all embarrassed him at the most criticalstage of this development of the process: "It was,however, no use for me to argue the matter in thepress. All that I could say would be mere talk and Ifelt that action was necessary, and not words." ''Action took the form of continued experiments and,by the end of 1857, a decision to build his own plantat Sheffield."* An important collateral developmentresulted from the visit to London in May 1857 ofG. F. Goran.s.son of Gefle. .Sweden. L sing Bessemerequipment, Goransson began trials of the process inNovember 1857 and by October 1858 was able toreport: "Our firm has now entirely given up themanufacture of bar iron, and our blast furnaces and tilt mills are now wholly employed in makinu; steel bythe Bessemer process, which may, therefore, be nowconsidered an accomplished commercial fact." '^Goransson was later to claim considerable improve-ments i;n the method of introducing the blast, and, inconsequence, the first effective demonstration of theBessemer method ""?this at a time when Bessemerwas still remelting the product of his con\erter incrucibles, after granulating the steel in water. IfMushet is to be believed, this success of Goransson'swas w holly due to his ore being "totally free from phos-phorous and sulphur." '" However, Bessemer's ownprogress was substantial, for his Sheffield works wererejjorted as being in acti\'e operation in .April 1859,and a price for his engineers' tool and spindle steel was '* October 17, 1857, writing as "Sidcros" {Mining Journal,1857, vol. 27, p. 723). ^' Mining Journal, 1857, vol. 27, p. 871, and 1858, vol. 28,p. 12. "/Airf. (1858), p. 34. " Mushet, op. cil. (footnote 46), p. 12. The phrase quoted istypical of Mushet's style. '- Bessemer, op. cil. (footnote 7), pp. 161 ff. and 256 ff. ?'/*/(/., p. 171. ''* This enterprise, started in conjunction with Galloway's ofManchester, one of the firms licensed by Bessemer to make hisequipment, was under way by .April 1858 (see Mining Journal,1858, vol. 28, p. 259). "' Mining Journal, 1858, vol. 28, p. 696. Mushet commented(p. 713) that he ha' off the carbonic gas formed in decarbonizing the iron." B is the port through which the charge of fluid iron is re-ceived, C and C are the tuyeres, and D is the tap hole forletting out the refined metal.42 BULLETIN 218: CONTRIBUTIONS FROM THE MUSEUM OF HISI'ORV .\ND TECHNOLOGV Muslut. by this time, had apparently decided loi^encralize the ap[)Hcation of his compound instead ofciting its use in conjunction with Marticn's process, or, as he put it, he had been obliged to do for hisEnglish specification by the Ebbw Vale Iron W'orks.The discussion in the Scientific American, which wasmostly concerned with Martien's claim to priority,soon evoked a letter from William Kelly. Writingunder date of September 30, 1856, from the SuwaneeIron Works, Eddyville, Kentucky, he claimed to havestarted "a series of experiments" in November 1851which had been witnessed by hundreds of persons and "discussed amongst the ironmasters, etc., of this sec-tion, all of whom are perfectly familiar with the wholeprinciple ... as discovered by me nearly five yearsago." A number of English puddlers had visited himto see his new process. "Several of them have sincereturned to England and may have spoken of myinvention there." Kelly expected "shortly to havethe in\ention perfected and bring it before thepublic." ?'Bessemer's application for an American patent wasgranted during the week ending November 18, 1856,and Kelly began his interference proceedings some-time before January 1857.'*Kelly's witnesses were almost wholly from the ranksof employees or former employees. The only excep-tion was Dr. Alfred H. Champion, a physician ofEddyville. Dr. Champion describes a meeting in thefall of 1851 with "two or three practical Ironmastersand others" at which Kelly described his process andinvited all present to see it in operation. He stated:The company present all differed in opinion from Mr.Kelly and appealed to me as a chemist in confirmationof their doubts. I at once decided that Mr. Kelly wascorrect in his Theory and then went on to explain thereceived opinion of chemists a century ago on this subject,and the present received opinion which was in directconfirmation of the novel theory of Mr. Kelly. I alsomentioned the analogy of said Kelly's process in decarbon-ising iron to the process of decarbonising blood in the humanlungs. " Scientific American, 1856, vol. 12, p. 43, Kelly's suggestion ofpiracy of his ideas was later enlarged upon by his biographer .John Newton Boucher, WiUuim Kelly: A true history oj llie so-called Bessemer process, Greensburg, Pennsylvania, 1924. ?' Ibid., p. 82. Kelly's notice of his intention to take testimonywas addressed to Bessemer on January 12, 1857. See paperson "Interference, William Kelly vs. Henry Bessemer DecisionApril 13, 1857." U. S. Patent Office Records. Quotationsbelow arc from this file, which is now permanently preservedin the library of the U. S. Patent Office. I'APER 3: BEGINNINGS OF CHEAP STEEL The Doctor does not say, specifically, if he or anyof the "company" went to see the process in operation.Kelly obtained aHidaviis from another seventeenwitnesses. Ten of the.se recorded their recollections ofexperiments conducted in 1847. Five described the1851 work. Two knew of or had seen both. One ofthe last group was John B. Evans who became forgemanager of Kelly's Union Forge, a few miles fromSuwanee. This evidence is of interest since a man inhis position should have been in a position to tell some-thing about the results of Kelly's operations in termsof usable metal. Unfortunately, he limits hiiiLself loa comment on the metal which had chilled around atuyere which had been sent back to the Forge ("it waspartly malleable and partly refined pig-iron") and toan account of a conversation with others who hadworked some of Kelly's "good wrought iron" madeby the new process.Only one of the witnesses (William Soden) makesa reference to the phenomenon which is an accom-paniment of the blowing of a converter: the prolongedand violent emission of sparks and flames whichstartled Bessemer in his first use of the process ^ andwhich still provides an exciting, if not awe-inspiring,interlude in a visit to a steel mill. Soden refers,without much excitement, to a boiling commotion,but the results of Kelly's "air-boiling'' were, evidently,not such as to impress the rest of those who claimedto have seen his furnace in operation. Only five ofthe total of eighteen of the witnesses say that theywitnessed the operations. .Soden, incidentally, knewof seven different "air-boiling" furnaces, some withfour and some with eight tuyeres, but he also neglectedlo report on the use of the metal.As is well known, Kelly satisfied the Acting Com-missioner that he had "made this invention andshowed it by drawings and experiment as early as1847," and he was awarded priority by the ActingCommissioner's decision of April 13, 1857, and U. S.Patent 17628 was granted him as of June 23, 1857.The Scientific American sympathized with Bessemer'srealization that his American patent was "of no morevalue to him than so much waste paper" but took theopportunity of chastising Kelly for his negligence innot securing a patent at a much earlier date and ccm-plained of a patent system which did not require aninventor to luake known his discovery promptly. Thejournal advocated a "certain fixed time" after whichsuch an inventor "should not be allowed to subvert w Bessemer, op cit. (footnote 7), p. 144. 43 a patent granted to another who has taken propermeasures to put the piihlic in possession of the in- vention."**Little authentic is known about Kelly's activitiesfollowing the grant of his patent. His biographer ""does not document his statements, many of whichappear to be based on the recollections of members ofKelly's family, and it is diHicult to reconcile some ofthem with what few facts are available. Kelly'sown account of his invention,'"^ itself undated, assertsthat he could "refine fifteen hundredweight of metalin from five to ten minutes," his furnace "supplying acheap method of making run-out metal" so that "after trying it a few days we entirely dispensed withthe old and troublesome run-out fires." '"^ Thisstatement suggests that Kelly's method was intendedto do just this; and it is not without interest to notethat several of his witnesses in the Interference pro-ceedings, refer to bringing the metal "to nature," aterm often used in connection with the finery furnace.If this is so, his assumption that he had anticipatedBessemer was based on a misapprehension of what thelatter was intending to do, that is, to make steel.This statement leaves the reader under the impres-sion that the process was in successful use. It is to becontrasted with the statement quoted above (page43), dated September 1856, when the process had,clearly, not been perfected. In this connection, itshould be noted that in the report on the SuwaneeIron Works, included in The iron manufacturersguide,^"* it is stated that "It is at this furnace that Mr.Kelly's process for refining iron in the hearth hasbeen most fully experimented upon." "? Scifnlific American, 1857, vol. 12, p. 341. '?' Boucher, op. cit. (footnote 97). '?- U. S. Bureau of the Census, Report on tlie manu/arturets of llieUnited Slates at the lent/i census {June 7, 1SS0) . . ., Manufactureof iron and steel, report prepared by James M. Swank, specialagent, Washington, 1883, p. 124. Mr. Swank was secretaryof the American Iron and Steel Association. This materialwas included in his History oj the rnanujaclure oj iron in atl ages,Philadelphia, 1892, p. 397. "" Ibtd., p. 125. The run-out fire (or "finery" fire) was acharcoal fire "into which pig-iion, having been melted andpartially refined in one fire, was run and further refined toconvert it to wrought iron by the Lancashire hearth process,"according to .^. K. Osborn, An encyclopaedia oj the iron and steelindustry. New York, 1956. '"? J. P. Lesley, op. cit. (footnote 39), p. 129. The preface isdated April 6, 1859. The data was largely collected by JosephLesley of Philadelphia, brother of the author, during a tourof several monllis. .Since Suwanee production is given for 44weeks only of 1857 (i.e., through November 4 or 5, 1857) it isconcluded that Lesley's visit was in the last few weeks of 1857. A major financial crisis affected United Statesbusiness in the fall of 1857. It began in the first weekof October and by October 31 the Economist (London)reported that the banks of the United States had "almost universally suspended specie payment." '"*Kelly was involved in this crisis and his plant wasclosed down. According to Swank,'"* some experi-ments were made to adapt Kelly's process to need ofrolling mills at the Cambria Iron Works in 1857 and1858, Kelly himself being at Johnstown, at least inJune 1858. That the experiments were not particu-larly successful is suggested by the lack of any Ameri-can contributions to the correspondence in the Englishtechnical journals. Kelly was not mentioned ashaving done more than interfere with Besscmer's firstpatent application. The success of the latter inobtaining patents'"" in the United States in November1856, covering "the conversion of molten crude iron . . . into steel or malleable iron, without the use offuel . . ." also escaped the attention of l)oth Entjlishand .American WTiters. It was not until 1861 that the question arose as towhat happened to Kelly's process. The occasion wasthe publication of an account of Bessemer's paper atthe Sheffield meeting of the (British) Society ofMechanical Engineers on August 1, 1861. Acceptingthe evidence of "the complete industrial success" ofBessemer's process, the Scientific American '"* asked: "W'ould not some of our enterprising manufacturersmake a good operation by getting hold of the [Kelly]patent and starting the manufacture of steel in thiscountry?There was no response to this rhetorical question,but a further inquiry as to whether the Kelly patent "could be bought" '"'' elicited a response from Kelly.Writing from Hammondsville. Ohio, Kelly "" said, inpart: I would say that the New England states and New Yorkwould be sold at a fair rale ... I removed from Kentuckv ^"^ Economist (London), 1857, vol. 15, pp. 1129, 1209. '"* Swank, op. cit. (footnote 42), p. 125. John Fritz, in hisAutobiography (New York, 1912, p. 162), refers to experimentsduring his time at Johnstown, i.e., between June 1854 andJuly 1860. The iron manufaiturer^s guide (see footnote 104) alsorefers to Kelly's process as having "just been tried wiih greatsuccess" at Cambria. "> U. S. patents 16082, dated November 11, 1856, and 16083,dated November 18, 1856. Bessemer's unsuccessful applicationcorresponded with his British patent 2321, of 1855 (see foot-note 98). "" Scientific .American, 1861, new sen, vol. 5, pp. 148153. '? Ihid., p. 310. ""/AiW., p. 343.44 BULLETIN 218: CONTRIBUTIONS FROM THE MUSEUM OF HISTORY AND TECHNOLOGY about tlircf years ago. and now reside at New Salisburyabout three miles from Haminondsville and sixty miles froinPittsburg. Accept my thanks for your kind efforts inendeavoring to draw the attention of the community to theadvantages of my process.This letter suggests that the Kelly jMocess had beendormant since 1858. Whether or not as a result ofthe publication of this letter, interest was resumed inKelly's experiments. Captain Eber Brock Ward ofDetroit and Z. .S. Diirfee of New Bedford, Massa-chusetts, obtained control of Kelly's patent. Durfeehimself went to England in the fall of 1861 in anattempt to secure a license from Bessemer. Hereturned to the United .States in the early fall of 1862.assuming that he was the only "citizen of the UnitedStates'' who had even seen the Bessemer apparatus.'"In June, 1862, W^ F. Durfee, a cousin of Z. S.Durfee, was asked by Ward to report on Kelly'sprocess. The report "- was unfavorable. "Thedescription of [the apparatus] used by Mr. Kelly athis abandoned works in Kentucky satisfied me that itwas not suited for an experiment on so large a scaleas was contemplated at Wyandotte [Detroit]."Since it was "confidently expected that Z. S. Durfeewould be successful in his efforts to purchase [Besse-mer's patents], it was thought only to be anticipatingthe acquisition of property rights ... to use such ofhis inventions as best suited the purpose in view."Thus the first "Bessemer" plant in the UnitedStates came into being without benefit of a licenseand supported only by a patent "'not suited" for alarge experiment. Kelly seems to have had no partin these developments. They took some time tocome to formation. Although the converter wasready by September 1862, the blowing engine wasnot completed until the spring of 1864 and the first "blow" successfully made in 1864. It may be nomore than a coincidence that the start of productionseems to have been impossible before the arrival in '" His claim is somewhat doubtful. Alexander LymanHolley, who was later to be responsible for the design of most ofthe first Bessemer plants in the United States had been inEngland in 18.S9, 1860, and 1862. In view of his interest inordnance and armor, it is unlikely that Bessemer could haveescaped his alert obseivation. His first visit specifically inconnection with the Bessemer process appears to have been in1863, but he is said to have begun to interest financiers and iron-masters in the Bessemer procc-ss after his visit in 1862 {Engineer-ing, 1882, vol. 33, p. 115.11- W. F. Durfee: "An account of the experimental steelworks at Wyandotte, Michigan," Transactions of the AmericanSociety of Stechanical Engineers, 1884, vol. 6, p. 40 fl". this country of a \oung man, L. M. Hart, who hadbeen trained in Bessemer operations at the plant ofthe Jackson Brothers at St. Seurin (near Bordeaux)France. The Jacksons had become Bcsscmcr's part-ners in respect of the French rights; and the rcciniit-ment of Hart suggests the possibility that it was fromthis French source that Z. S. Durfee obtained hisinitial technical data on the operation of the Bessemerprocess.'"During the organization of the plant at Wyandolte,Kell\- was called back to Cambria, probably byDaniel J. Morrell, who, later, became a partner withWard and Z. S. Durfee in the formation of the KelKPneumatic Process Company."* We learn from JohnE. Fry,"^ the iron moulder who was assigned to helpKelly, that ? in 1862 Mr. Kelly returned to Johnstown for a crucial,and as it turned out, a final series of experiments by himwith a rotative [Bessemer converter] made abroad and importedfor his purpose. This converter embodied in its materials andconstruction several of Mr. Bessemer's patented factors, ofwhich, up to the close of Mr. Kelly's experiments abovenoted, he seemed to have no knowledge or conception. .Andit was as late as on the occasion of his return in 1862. tooperate the experimental Bessemer converter, that he firstrecognized, by its adoption, the necessity for or the impor-tance of any after treatment of, or additions required by theblown metal to convert it into steel.Fry later asserted "* that Kelly's experiments in1862 were simply attempts to copy Bessemer's ineth-ods. (The possibility is under investigation that theso-called "pioneer converter" now on loan to theU. S. National Museum from the Bethlehem SteelCompany, is the converter referred to by Fry.)William Kelly, in eflfect, disappeared frotti therecord until 1871 when he applied for an extensionof his patent of June 23, 1857. The application wasopposed (by whom, the record does not state) on thegrounds that the invention was not novel when it wasoriginally issued, and that it would be against thepublic interest to extend its term. The Coinmissioner 1" Research in the French sources continues. The arrivalof L. M. Hart at Boston is recorded as of April 1, 1864, hisship being the SS A/riea out of Liverpool, England (.\rchivcsof the United States, card index of passenger arrivals 1849-1891 list No. 39).11* Swank, op. cit. (footnote 42), p. 409. '^^ Johnstown Daily Democrat, souvenir edition, autumn1894 (italics supplied). Mr. Fry was at the Cambria IronWorks from 1858 until after 1882.11' Engineering, 1896, vol. 61, p. 615. PAPER ?>: BLGINNINGS OF CHEAP STEEL 45 ruled that,"" on the first qucsiion. it was settled prac-tice of the Patent Ollice not to reconsider former de-cisions on questions of fact; the novehy of Kelly'sinvention iiad been re-examined when the patentwas reissued in November 1857. Testimony showedthat the patent was very valuable; and that Kelly "had been untirinij in his efforts to introduce it intouse but the opposition of iron manufacturers and theamount of capital required prevented him from re-ceiving anythine; from his patent until within veryfew years past." Kelly's expenditures were shown tohave amounted to $11,500, whereas he had receivedonly $2,400. Since no evidence was filed in supportof the public interest aspect of the case, the Commis-sioner found no substantial reason for denying theextension; indeed "very few patentees are able topresent so strong grounds for extension as the appli-cant in the ca.se."In a similar application in the previous year,Bessemer had failed to win an extension of his U. S.patent 16082, of November 11, 1856, for the solereason that his Briti.sh patent with which it had beenmade co-terminal had duly expired at the end of itsfourteen years of life, and it would have been in-equitable to give Bessemer protection in the UnitedStates while British iron-masters were not undersimilar restraint. But if it had not been for thisconsideration, Bessemer "would be justly entitled towhat he asks on this occasion." The Commissioner '"* ob.served". "It may be questioned whether [Be.s.semer]was first to discover the principle upon which his proc- ess was founded. But we owe its reduction to prac-tice to his untiring industry and perseverance, hissuperior skill and .science and his great outlay." Concl usions Martien was probably never a serious contender forthe honor of discovering the atmospheric process ofmaking steel. In the ])resent state of the record, itis not an unreasonable a.ssumption that his patentwas never seriously exploited and that the RbbwVale Iron Works hoped to use it, in conjunction withthe Mushet patents, to upset Bessemer's patents.The position of Mushet is not so clear, and it ishoped that further research can eventually throw a clearer light on his relationship with the I".l)i)w ValeIron Works. It may well be that the "opinion ofmetallurgists in later years" "'' is .sound, and liiat iioihMushci and Bessemer had successfulK worked at thesame proijlem. The study of Mushei's letters to thetechnical press and of the attitude of the editors ofthose papers to Mushet suggests the possibility that he,too, was used by Ebbw Vale for the pin-poses of theirattacks on Bessemer. Mushet admits that he was nota free agent in respect of these patents, and the failureof Ebbw Vale to ensure their full life under Englishpatent law indicates clearly enough that by 1859 thefirm had realized that their position was not strongenough to warrant a legal suit for infringementagainst Bessemer. Their purchase of the Uchatiusprocess and their final attempt to develop Martien'sideas through the Parry patents, which exposed themto a very real risk of a suit by Bessemer, are also indi-cations of the politics in the case. Mushet seems tohave been a willing enough victim of Ebbw \'ale'sscheming. His letters show an almost presumptuousassumption of the mantle of his father; while hissometimes absurd claims to priority of invention (anddemonstration) of practically every new idea in themanufacturing of iron and steel progressively reducedthe respect for his name. Bessemer claiiTis an impres-sive array of precedents for the use of manganese in steel making and, given his attitude to patents and hisreliance on professional advice in this respect, heshould perhaps, be given the benefit of the doubt.A dispassionate judgment would be that Bessemerowed more to the development work of his Swedishlicensees than to Mushet.Kelly's right to be adjudged the joint insentor ofwhat is now often called the Kelly-Bes.semer process isquestionable.'^" Admittedly, he experimented in thetreatment of molten metal with air blasts, but it is byno means clear, on the evidence, that he got beyondthe experimental stage. It is certain that he neverhad the objective of making steel, which was Besse-mer's primary aim. Nor is there evidence that hisprocess was taken beyond the experimental stage bythe Cambria Works. The rejection of his "apparatus"by W. F. Durfee must have been based, to some extent at least, upon the Johnstown trials. There are strong '" Sec U. S. Patent Office, Decision of Commissioner ofPatcnU, dated June 15, 1871. '" U. S. Patent OfTice, Decision of Commissioner of Patentsdated February 12, 1870. '" William 1'. Jeans, The creators of the age of steel, London,1884. '^'' Bessemer dealt with Kelly's claim to priority in a Utter toEngineering, 1896, vol. 61, p. 367.46 BULLETIN 218: CONTRIBUTIONS FROM THE MUSEUM OF HISTORY .AND TECHNOLOGY grounds then, for agreeing with one Iiistorian '^' whoconchides:The fact that Kelly was an American is evidently theprincipal reason vvh\- certain popular writers have mademuch of an invention that, had not Bessemer developedhis process, would never have attracted notice. Kelly'spatent proved very useful to industrial interests in thiscountry as a bargaining weapon in negotiations withthe Bessemer group for the exchange of patent rights. '-' Louis C. Hunter, "The heavy industries since 1860," inH. F. Williamson (editor), Thf growth of the American economy.New York, 1944, p. 469. Kelly's suggestion '^^ that some British puddlefs mayhave communicated his secret to Bessemer can, prob-ably, never be verified. All that can be said is thatBessemer was not an ironman; his contacts with theiron trade were, so far as can be ascertained, non- existent until he himself invaded SheHield. So it is unlikely that such a secret would have been taken tohim, even if he were a well-known inventor. '-- Later developed into a dramatic story by Boucher, op. cil.(footnote 97). PAPER .3: BEGINNINGS OF CHEAP STEEL 47