STOP Early Journal Content on JSTOR, Free to Anyone in the World This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR. Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries. We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes. Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content . JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org. C 253 3 XX. On the Alloys of steel. By J. Stodart, Esq, F.R. S. and Mr. M. Faraday, Chemical Assistant in the Royal Institu- tion. Communicated by J. Stodart, Esq* F.R.S. Read March 21, 1822. 1 he alloys of steel made on a small scale in the laboratory of the Royal Institution proving to be good, and the experi- ments having excited a very considerable degree of interest both at home and abroad, gave encouragement to attempt the work on a more extended scale, and we have now the pleasure of stating, that alloys similar to those made in the Royal Institution, have been made for the purpose of manu- facture ; and that they prove to be, in point of excellence, in every respect equal, if not superior, to the smaller produc- tions of the laboratory. Previous however, to extending the work, the former experiments were carefully repeated, and to the results were added some new combinations, namely, steel with palladium, steel with iridium, and osmium, and latterly, steel with chromium. In this last series of experi- ments we were particularly fortunate, having by practice acquired considerable address in the management of the fur- naces, and succeeded in procuring the best fuel for the pur- pose. Notwithstanding the many advantages met with in the laboratory of the Royal Institution, the experiments were frequently rendered tedious from causes often unexpected, and sometimes difficult to overcome ; among these, the fai- lure of crucibles was perhaps the most perplexing. We have 254 Mr. Stodart and Mr. Faraday never yet found a crucible capable of bearing the high degree of temperature required to produce the perfect reduction of titanium ; indeed we are rather disposed to question whether this metal has ever been so reduced : our furnaces are equal* (if any are ) to produce this effect, but hitherto we have failed in procuring a crucible. The metals that form the most valuable alloys with steel are silver, platina, rhodium, iridium and osmium, and palla- dium ; all of these have now been made in the large way, except indeed the last named. Palladium has, for very obvious reasons, been used but sparingly ; four pounds of steel with T ~ part of palladium, has however been fused at once, and the compound is truly valuable, more especially for making instruments that require perfect smoothness of edge. We are happy to acknowledge the obligations due from us to Dr. Wollaston, whose assistance we experienced in every stage of our progress, and by whom we were furnished with all the scarce and valuable metals; and that with a liberality which enabled us to transfer our operations from the labora- tory of the chemist, to the furnace of the maker of cast steel. In making the alloys on a large scale, we were under the necessity of removing our operations from London to a steel furnace at Sheffield ; and being prevented by other avocations from giving personal attendance, the superintendence of the work was consequently intrusted to an intelligent and confi- dential agent. To him the steel, together with the alloying metals in the exact proportion, and in the most favourable state for the purpose, was forwarded, with instructions to see • We have succeeded in fusing in these furnaces rhodium, and also, though im- perfectly, platinum in crucibles. on the alloys of steel. 255 the whole of the metals, and nothing else, packed into the crucible, and placed in the furnace, to attend to it while there, and to suffer it to remain for some considerable time in a state of thin fusion, previous to its being poured out into the mould. The cast ingot was next, under the same super- intendence, taken to the tilting mill, where it was forged into bars of a convenient size, at a temperature not higher than just to render the metal sufficiently malleable under the tilt hammer. When returned to jus* it -was subjected to exami- nation both mechanical and chemical, as well as compared with the similar products of the laboratory From the exter- nal appearance, as well as from the texture of the part when broken by the blow of the hammer, we were able to form a tolerably correct judgment as to its general merits ; the hard- ness, toughness, and other properties, were farther proved by severe trials, after being fashioned into some instrument, or tool, and properly hardened and tempered. It would prove tedious to enter into a detail of experiments made in the Royal Institution ; a brief notice of them will at present be sufficient. After making imitations of various specimens of meteoric iron by fusing together pure iron and nickel, in proportions of 3 to 10 per cent, we attempted mak- ing an alloy of steel with silver, but failed, owing to a super- abundance of the latter metal ; it was found, after very many trials, that only the -^~ part of silver would combine with steel, and when more was used a part of the silver was found in the form of metallic dew lining the top and sides of the crucible ; the fused button itself was a mere mechanical mix- ture of the two metals, globules of silver being pressed out of the mass by contraction in cooling, and more of these glo- bules being forced out by the hammer in forging ; and far- 256 Mr. Stodart and Mr. Faraday ther, when the forged piece was examined, by dissecting it with diluted sulphuric acid, threads or fibres of silver were seen mixed with the steel, having something of the appear- ance of steel and platina when united by welding: but when the proportion of silver was only T ~ part, neither dew, glo- bules, nor fibres appeared, the metals being in a state of per- fect chemical combination, and the silver could only be detected by a delicate chemical test. With platina and rhodium, steel combines in every pro- portion ; and this appears also to be the case with iridium and osmium : from 1 to 80 per cent, of platina was perfectly com- bined with steel, in buttons of from 500 to 2000 grains. With rhodium, from 1 to 50 per cent, was successfully used. Equal parts by weight of steel and rhodium gave a button, which, when polished, exhibited a surface of the most exqui- site beauty : the colour of this specimen is the finest imagi- nable for a metallic mirror, nor does it tarnish by long exposure to the atmosphere : the specific gravity of this beau- tiful compound is 9.176. The same proportion of steel and platina gave a good button, but a surface highly crystalline renders it altogether unfit for a mirror. In the laboratory we ascertained that, with the exception of silver, the best proportion of the alloying metal, when the object in view was the improvement of edge tools, was about -~ part, and in this proportion they have been used in the large way. It may be right to notice, that in fusing the metals in the labo- ratory no flux whatever was used, nor did the use of any ever appear to be required. Silver being comparatively of little value with some of the alloying metals, we were disposed to make trial with it as the first experiment in the large way 81bs. of very good on the alloys of steel. 25^ Indian steel was sent to our agent, and with it ^~ part of pure silver : a part of this was lost owing to a defect in the mould ; a sufficient quantity was however saved, to satisfy us as to the success of the experiment. This, when returned, had the most favourable appearance both as to surface and frac- ture ; it was harder than the best cast steel, or even than the Indian wootz, with no disposition whatever to crack, either under the hammer, or in hardening. Some articles, for va- rious uses, have been made from this alloy ; they prove to be of a very superior quality ; its application will probably be extended not only to the manufacture of cutlery, but also to various descriptions of tools ; the trifling addition of price can- not operate against its very general introduction. The silver alloy may be advantageously used for almost every purpose for which good steel is required. Our next experiment made in the large way, was with steel and platina. lolbs. of the same steel, with T ~ g part of pla- tina, the latter in the state produced by heating the ammonia muriate in a crucible to redness, was forwarded to our agent, with instructions to treat this in the same way as the last named metals. The whole of this was returned in bars re- markable for smoothness of surface and beauty of fracture. Our own observation, as well as that of the workmen em- ployed to make from it various articles of cutlery, was, that this alloy, though not so hard as„the former, had considerably more toughness; this property will render it valuable for every purpose where tenacity, as well as hardness, is re- quired ; neither will the expense of platina exclude it from a pretty general application in the arts ; its excellence will much more than repay the extra cost, mdcccxxii. LI 358 Mr. Stodart and Mr. Faraday The alloys of steel with rhodium have also been made in the large way, and are perhaps the most valuable of all ; but these, however desirable, can never, owing to the scarcity of the metal, be brought into very general use. The compound of steel, iridium and osmium, made in the large way, is also of great value ; but the same cause, namely, the scarcity and difficulty of procuring the metals, will operate against its very general introduction. A sufficient quantity of these metals may, perhaps, be obtained to combine with steel for the purpose of making some delicate instruments, and also as an article of luxury, when manufactured into razors. In the mean time, we have been enabled, repeatedly, to make all these alloys (that with palladium excepted) in masses of from 8 to solbs. each ; with such liberality were we furnished with the metals from the source already named. A point of great importance in experiments of this kind was, to ascertain whether the products obtained were exactly such as we wished to produce. For this purpose, a part of each product was analysed, and in some cases the quantity ascertained ; but it was not considered necessary in every case to verify the quantity by analysis, because, in all the experi- ments made in the laboratory, the button produced after fu- sion was weighed, and if it fell short of the weight of both metals put into the crucible, it was rejected as imperfect, and put aside. When the button gave the weight, and on analysis gave proofs of containing the metal put in to form the alloy, and also on being forged into a bar and acted on by acids, presented an uniform surface, we considered the evidence of its composition as sufficiently satisfactory. The processes of analysis, though simple, we shall briefly state ; the informa- on the alloys of steel. 259 tion may be desirable to others who may be engaged on similar experiments ; and farther, may enable every one to detect any attempt at imposition. It would be very desirable at present, to possess a test as simple, by which we could dis- tinguish the wootz, or steel of India, from that of Europe ; but this, unfortunately, requires a much more difficult process of analysis. To ascertain if platina is in combination with steel, a small portion of the metal, or some filings taken from the bar, is to be put into dilute sulphuric acid ; there will be rapid action ; the iron will be dissolved, and a black sediment left, which will contain carbon, hydrogen, iron and platina ; the carbon and hydrogen are to be burnt off, the small portion of iron separated by muriatic acid, and the residuum dissolved in a drop or two of nitro-muriatic acid. If a piece of glass be moistened with this solution, and then heated by a spirit lamp and the blow pipe, the platina is reduced, and forms a me- tallic coating on the glass. In analysing the alloy of steel and silver, it is to be acted on by dilute sulphuric acid, and the powder boiled in the acid ; the silver will remain in such a minute state of division, that it will require some time to deposit. The powder is then to be boiled in a small portion of strong muriatic acid ;* this will dissolve the iron and silver, and the latter will fall down as a chloride of silver on dilution with water ; or the powder may be dissolved in pure nitric acid, and tested by muriatic acid and ammonia. * Although it is a generally received opinion that muriatic acid does not act on silver, yet that is not the case ; pure muriatic acid dissolves a small portion of silver very readily. 260 Mr. Stodart and Mr. Faraday The alloy of steel and palladium, acted on by dilute sul- phuric acid, and boiled in that acid, left a powder which, when the charcoal was burnt from it, and the iron partly separated by cold muriatic acid, gave on solution in hot muriatic acid, or in nitro-muriatic acid, a muriate of palladium ; the solution, when precipitated by prussiate of mercury, gave prussiate of palladium ; and a glass plate moistened with it and heated to redness, became coated with metallic palladium. The residuum of the rhodium alloy obtained by boiling in diluted sulphuric acid, had the combustible matter burnt off, and the powder digested in hot muriatic acid : this removed the iron ; and by long digestion in nitro-muriatic acid, a mu- riate of rhodium was formed, distinguishable by its colour, and by the triple salt it formed with muriate of soda. To analyse the compound of steel with iridium and osmium, the alloy should be acted on by dilute sulphuric acid, and the residuum boiled in the acid ; the powder left is to be collected and heated with caustic soda in a silver crucible to dull red- ness for a quarter of an hour, the whole to be mixed with water, and having had excess of sulphuric acid added, it is to be distilled, and that which passes over condensed in a flask : it will be a solution of oxide of osmium, will have the peculiar smell belonging to that substance, and will give a blue precipitate with tincture of galls. The portion in the retort being then poured out, the insoluble part is to be washed in repeated portions of water, and then being first slightly acted on by muriatic acid to remove the iron, is to be treated with nitro-muriatic acid, which will give a muriate of iridium. In these analyses, an experienced eye will frequently on the alloys of steel. 261 perceive, on the first action of the acid, the presence of the alloying metal. When this is platina, gold, or silver, a film of the metal is quickly formed on the surface of the acid. Of alloys of platina, palladium, rhodium, and iridium and osmium, a ready test is offered when the point is not to ascer- tain what the metal is, but merely whether it be present or not. For this purpose we have only to compare the action of the same acid on the alloy and on a piece of steel ; the increased action on the alloy immediately indicates the pre- sence of the metal ; and by the difference of action, which on experience is found to be produced with the different metals, a judgment may be formed even of the particular one present. The order in which the different alloys stand with regard to this action, is as follows : steel, cromium alloy, silver alloy, gold alloy, nickel alloy, rhodium alloy, iridium and osmium alloy, palladium alloy, platina alloy. With similar acid the action on the pure steel was scarcely perceptible ; the silver alloy gave very little gas, nor was" the gold much acted on. All the others gave gas copiously, but the platina alloy in most abundance. In connection with the analysis of these alloys, there are some very interesting facts to be observed during the action of acids on them, and perhaps none of these are more strik- ing than those last referred to. When the alloys are im- mersed in diluted acid, the peculiar properties which some of them exhibit, not only mark and distinguish them from com- mon steel, and from each other, but also give rise to some considerations on the state of particles of matter of different 262 Mr. Stodart and Mr. Faraday kinds when in intimate mixture or in combination, which may lead to clearer and more perfect ideas on this subject. If two pieces, one of steel, and one steel alloyed with pla- tina, be immersed in weak sulphuric acid , the alloy will be immediately acted on with great rapidity and the evolution of much gas, and will shortly be dissolved, whilst the steel will be scarcely at all affected. In this case, it is hardly pos- sible to compare the strength of the two actions. If the gas be collected from the alloy and from the steel for equal in- tervals of time, the first portions will surpass the second some hundreds of times. A very small quantity of platina alloyed with steel confers this property on it : ? ~ increased the action considerably ; with ^Q and T £ 3 it was powerful ; with 10 per cent, of pla- tina it acted, but not with much power ; with 50 per cent, the action was not more than with steel alone ; and an alloy of 90 platina with 20 steel Was not affected by the acid. The action of other acids on these alloys is similar to that of sulphuric acid, and is such as would be anticipated : dilute muriatic acid, phosphoric acid, and even oxalic acid, acted on the platina alloy with the liberation of more gas than from zinc ; and tartaric acid and acetic acid rapidly dissolved it. In this way chalybeate solutions, containing small portions of protoxide of iron, may be readily obtained. The cause of the increased action of acids on this and similar alloys, is, as the President of this Society suggested to us, probably electrical. It may be considered as occasioned by the alloying metal existing in such a state in the mass, that its particles form voltaic combinations with the particles of on the alloys of steel. 263 steel, either directly, or by producing a definite alloy, which is diffused through the rest of the steel ; in which case the whole mass would be a series of such voltaic combinations : or it may be occasioned by the liberation, on the first action of the acid, of particles which, if not pure platina, contain, as has been shown, a very large proportion of that metal, and which, being in close contact with the rest of the mass, form voltaic combinations with it in a very active state : or, in the third place, it may result from the iron being mechanically divided by the platina, so that its particles are more readily attacked by the acid, analogous to the case of proto-sulphuret of iron. Although we have not been able to prove by such experi- ments, as may be considered strictly decisive, to which of these causes the action is owing, or how much is due to any of them, yet we do not hesitate to consider the second as al- most entirely, if not quite, the one that is active. The reasons which induce us to suppose this to be the true cause of the action, rather than any peculiar and previous arrangement of the panicles of steel and platina, or than the state of division of the steel, are, that the two metals combine in every propor- tion we have tried, and do not, in any case, exhibit evidences of a separation between them, like those, for instance, which steel and silver exhibit ; that when, instead of an acid, weaker agents are used, the alloy does not seem to act with them as if it was a series of infinitely minute voltaic combinations of steel and platina, but exactly as steel alone would do; that the mass does not render platina wire more negative than steel, as it probably in the third case would do ; that it does not rust more rapidly in a damp atmosphere ; and that when 2 6% Mr. Stod art and Mr. Faraday placed in saline solutions, as muriate of soda,&c, there is no action takes place between them. In such cases it acts just like steel ; and no agent that we have as yet tried, has pro- duced voltaic action that was not first able to set a portion of the platina free by dissolving out the iron. Other interesting phenomena exhibited by the action of acid on these steels, are the differences produced when they are hard and when soft. Mr. Daniel, in his interesting paper on the mechanical structure of iron, published in the Journal of Science, has remarked, that pieces of hard and soft steel being placed in muriatic acid, the first required five fold the time of the latter to saturate the acid ; and that when its sur- face was examined, it was covered with small cavities like worm-eaten wood, and was compact and not at all striated, and that the latter presented a fibrous and wavy texture. The properties of the platina alloy have enabled us to ob- serve other differences between hard and soft steel equally striking. When two portions of the platina alloy, one hard and one soft, are put into the same diluted sulphuric acid and suffered to remain for a few hours, then taken out and ex- amined, the hard piece presents a covering of a metallic black carbonaceous powder, and the surface is generally slightly fibrous, but the soft piece, on examination, is found to be covered with a thick coat of grey metallic plumbaginous matter, soft to the touch, and which may be cut with a knife, and its quantity seven or eight times that of the powder on the hard piece : it does not appear as if it contained any free charcoal, but considerably resembles the plumbaginous pow- der Mr. Daniel describes as obtained by the action of acid on cast iron. on the alloys of steel. 265 The same difference is observed if pure steel be used, but it is not so striking ; because, being much less rapidly attacked by the acid, it has to remain longer in it, and the powder pro- duced is still farther acted on. The powder procured from the soft steel or alloy in these experiments, when it has not remained long in the acid, ex- actly resembles finely divided plumbago, and appears to be a carburet of iron, and probably of the alloying metal also. It is not acted on by water, but in the air the iron oxidates and discolours the substance. When it remains long in the acid, or is boiled in it, it is reduced to the same state as the powder from the hard steel or alloy. When any of these residua are boiled in diluted sulphuric "or muriatic acid, protoxide of iron is dissolved, and a black powder remains unalterable by the farther action of the acid ; it is apparently in greater quantity from the alloys than from pure steel, and when washed, dried, and heated to 300" or 400 in the air, burns like pyrophorus,with much fume ; or if light- ed, burns like bitumen, and with a bright flame ; the residuum is protoxide of iron, and the alloying metal. Hence, during the action of the acid on the steel, a portion of hydrogen enters into combination with part of the metal and the charcoal, and forms an inflammable compound not acted upon by the acid. Some striking effects are produced by the action of nitric acid on these powders. If that from pure steel be taken, it is entirely dissolved ; and such is also the case if the powder be taken from an alloy, the metal of which is soluble in nitric acid ; but if the powder is from an alloy, the metal of which is not soluble in nitric acid, then a black residuum is left not mdcccxxh. M m 266 Mr. Stodart and Mr. Faraday touched by the acid ; and which, when washed and carefully dried, is found, when heated, to be deflagrating; and with some of the metals, when carefully prepared, strongly ex- plosive. The fulminating preparation obtained from the platitia alloy, when dissolved in nitro-muriatic acid, gave a solution containing much platina, and very little iron. When a little of it was wrapped in foil and heated, it exploded with much force, tearing open the foil, and evolving a faint light. When dropped on the surface of heated mercury, it exploded readily at 400 of Fahrenheit, but with difficulty at 370 . When its temperature was raised slowly, it did not explode, but was decomposed quietly. When detonated in the bottom of a hot glass tube, much water and fume were given off, and the residuum collected was metallic platina with a very little iron and charcoal. We are uncertain how far this preparation resembles the fulminating platina of Mr. Edmund Davy. In these alloys of steel the differences of specific gravity are not great, and may probably be in part referred to the denser state of the metals from more or less hammering ; at the same time it may be observed, that they are nearly in the order of the specific gravities of the respective alloying metals. The alloys of steel with gold, tin, copper, and chromium, we have not attempted in the large way. In the laboratory, steel and gold were combined in various proportions ; none of the results were so promising as the alloys already named, nor did either tin or copper, as far as we could judge, at all improve steel. With titanium we failed, owing to the imper- fection of crucibles. In one instance, in which the fused on tlie alloys of steel. 267 button gave a fine damask surface, we were disposed to attri- bute the appearance to the presence of titanium; but in this we were mistaken ; the fact was, we had unintentionally made wootz. The button, by analysis, gave a little silex and alumine, but not an atom of titanium ; menachanite, in a par- ticular state of preparation, was used : this might possibly contain the earths or their basis, or they may have formed a part of the crucible. M. Berthier, who first made the alloy of steel and chro- mium,* speaks very favourably of it. We have made only two experiments. 1600 grains of steel, with 16 of pure chrome, were packed into one of the best crucibles, and placed, in an excellent blast furnace ; the metals were fused, and kept in that state for some time. The fused button proved good and forged well : although hard, it showed no disposition to crack. The surface being brightened, and slightly acted on by dilute sulphuric acid, exhibited a crystalline appearance; the crystals, being elongated by forging, and the surface again polished gave, by dilute acid, a very beautiful damask. Again, 1600 grains of steel with 48 of pure chrome were fused : this gave a button considerably harder than the former. This too was as malleable as pure iron, and also gave a very fine damask. Here a phenomenon rather curious was ob- served : the damask was removed by polishing, and restored by heat without the use of any acid. The damasked surface, now coloured by oxidation, had a very novel appearance : the beauty was heightened by heating the metal in a way to ex- hibit all the colours caused by oxidation, from pale straw to blue, or from about 430 to 6oo° of Fahrenheit. The blade * Annales de Chimie, XVII. 55. a68 ! Mr. Stod art and Mr. Faraday of a sabre, or some such instrument, made from this alloy, and treated in this way, would assuredly be beautiful, whatever its other properties might be; for of the value of the chrome alloy for edge tools we are not prepared to speak, not having made trial of its cutting powers. The sabre blade, thus co- loured, would amount to a proof of its being well tempered ; the blue back would indicate the temper of a watch spring, while the straw colour towards the edge would announce the requisite degree of hardness. It is confessed, that the opera- tion of tempering any blade of considerable length in this way, would be attended with some difficulty. In the account now given of the different alloys, only one triple compound is noticed, namely, steel, iridium and os- mium ; but this part of the subject certainly merits farther investigation, offering a wide and interesting field of research. Some attempts to form other combinations of this description proved encouraging, but we were prevented, at the time, by various other avocations, from bestowing on them that atten- tion and labour they seemed so well to deserve.* It is a curious fact, that when pure iron is substituted for steel, the alloys so formed are much less subject to oxidation. 3 per cent, of iridium and osmium fused with some pure iron, gave a button, which when forged and polished was exposed, with many other pieces of iron, steel, and alloys, to a moist atmosphere : it was the last of all showing any rust. The co- lour of this compound was distinctly blue ; it had the property of becoming harder when heated to redness and quenched in a cold fluid. On observing this steel-like character, we sus- * It is our intention to continue these experiments at every opportunity, but they are laborious, and require much time and patience. on the alloys of steel. st6£ pected the presence of carbon ; none however was found, al- though carefully looked for. It is not improbable that there may be other bodies, besides charcoal, capable of giving to iron the properties of steel ; and though we cannot agree with M. Boussingault,* when he would replace carbon in steel by silica or its base, we think his experiments very interesting on this point, which is worthy farther examination. We are not informed as to what extent these alloys, or any of them, have been made at home, or to what uses they have been applied ; their more general introduction in the manu- facture of cutlery would assuredly add to the value, and con- sequently to the extension of that branch of trade. There are various other important uses to which the alloys of steel may advantageously be applied. If our information be cor- rect, the alloy of silver, as well as that of platina, has been to some considerable extent in use at His Majesty's Mint. We do know, that several of the alloys have been diligently and successfully made on the Continent ; very good specimens of some of them having been handed to us ; and we are proud of these testimonies of the utility of our endeavours. To succeed in making and extending the application of these new compounds, a considerable degree of faithful and diligent attention will be required on the part of the operators. The purity of the metals intended to form the compound is essential ; the perfect and complete fusion of both must, in every case, be ascertained : it is farther requisite, that the metals be kept for some considerable time in the state of thin fusion ; after casting, the forging is with equal care to be attended to ; the metal must on no account be overheated ; * Annates de Chimie, XVI. I. 270 Mr. Stodart and Mr. Faraday on the alloys of steel. and this is more particularly to be attended to when the al- loying metal is fusible at a low temperature, as silver. The same care is to be observed in hardening: the article is to be brought to a cherry-red colour, and then instantly quenched in the cold fluid. In tempering, which is best performed in a metallic bath properly constructed, the bath will require to be heated for the respective alloys, from about 70 to 100 of Fahrenheit above the point of temperature required for the best cast steel. We would farther recommend, that this act of tem- pering be performed twice ; that is, at the usual time before grinding, and again just before the last polish is given to the blade. This second tempering may perhaps appear super- fluous, but upon trial its utility will be readily admitted. We were led to adopt the practice by analogy, when considering the process of making and tempering watch springs.