has arisen as to the importance or otherwise of such " temper hrittleness " on the ground that it can only be detected by a partic- ular form of test the " notched bar impact test." Hatfield has repeatedly suggested that this test measures a property of no prac- tical importance and that, therefore, temper brittleness is only " apparent " and not " real." This view, however, has not been generally accepted, since a number of investigators (Rosenhain, Greaves and others) have found a very distinct correlation between a low value under the notched bar impact test and cases of actual failure in service.
For certain purposes, where simple nickel steels are not found adequate, a chrome-vanadium steel has been extensively used. The fact that vanadium plays a really important r6Ie in such steels has come to be recognized, but there are still difficulties in its metal- lurgical applications. For quite a different range of purposes a steel containing relatively large amounts of chromium (about 13 %, Brearley) has found wide applications. Its best-known application is to the production of " stainless " cutlery which is proving extremely successful in practice and, in spite of the relatively high cost, is coming into extensive use as a labour-saving device. Some difficulty was at first experienced in hardening this material satisfactorily ; it requires a higher quenching temperature than cutlery-hardeners have been accustomed to employ (Hatfield), and, as a consequence, the earlier products were insufficiently hardened and gave rise to the complaint that such cutlery would not keep its edge. Recent practice has overcome this defect, although the fact still remains that " stainless " knives, not being automatically sharpened by daily polishing, require careful sharpening from time to time. The appli- cations of this steel to other purposes where resistance to corrosion is important are being found almost daily. Incidentally, it was found that this steel is capable of resisting the severe conditions which occur in the service of an exhaust-valve in aeroplane engines, al- though a high-tungsten steel (18% tungsten) has been found to be even slightly better (Aitchison).
High-speed Steels. In the metallurgical progress of the period under review, the development of high-speed cutting steels occupies an important place, their possibilities having been steadily exploited to an increasing extent, their development being naturally accom- panied by an evolution of machine-tools capable of utilizing the high- cutting powers of the new steels. Under war conditions, the great demand for tool-steels of this kind created a relatively enormous demand for tungsten, and considerable developments in the mining and production of tungsten took place. Efforts to replace tungsten by other metals were also made, and very great claims were advanced for a " new " high-speed steel in which molybdenum in combination with vanadium was used in place of tungsten (Arnold). The use of molybdenum in place of tungsten, however, was by no means new, and the merits of the new " discovery " therefore depended entirely upon the extent to which the known disadvantages of molybdenum in this connexion had been overcome. Uncertainty of quality and a strong tendency to crack while in the ingot form were the chief of these. Indeed, at the present time, there seems to be no tendency for the normal tungsten high-speed steel to be discarded in favour of the molybdenum-vanadium alloy.
In connexion with high-speed steels, mention may here be made of two alloys which have been used, with considerable success, in place of such steel. These are known as " stellite " and "cooperite" respectively. Stellite consists, according to one reliable analysis, mainly of cobalt 56%, chromium 34%, tungsten 9%, carbon I %. Actual cutting tests with this material have shown that it is capable of cutting rather faster than the best tungsten steel, provided that the cut is smooth and regular, but that for roughing cuts, where the tool is subjected to sudden shocks and jars, the alloy is unsuited as it is too brittle and the tool frequently breaks off. In the form as ordinarily marketed, moreover, stellite cannot be softened, so that it must be used in the cast condition and the tool has to be ground to shape while hard. Recently, a form of stellite which can be softened has been announced, but no data in regard to it are yet available.
Cooperite is an alloy of nickel, tungsten and zirconium, described as containing 80 % of nickel, 14 % of tungsten and 6 % of zirconium.
Cobalt Steel. Returning to alloy steels, mention must be made of the use of cobalt. The great development of cobalt production at Sudbury in Canada has made this metal available and has attracted interest to its possible uses. A high-speed tool-steel containing cobalt has been produced in Sheffield, and has not only been found to be successful in general use, but to have the remarkable property that it does not undergo any distortion during the hardening process. Should this property be established in practical service the steel is likely to prove of very great importance in the production of shaped cutting tools in which great accuracy is required. War conditions brought with them painful experience of the great difficulty which then existed in the production of accurate parts such as those of shells and fuzes which were required to be strictly interchangeable. This difficulty extended back to the gauges and master-gauges used for the checking of such parts and, ultimately, in many cases to the cutting-tools used in their production. In other cases, the steel of which the gauges themselves were made gave much trouble owing to distortion during hardening, requiring considerable adjustment by " lapping " of the hardened article.
A cobalt-steel has also made its appearance as an important inno- vation in another field that of steels for permanent magnets. Made in the first place of hardened carbon steels, the requirements particularly of the magnetos used for ignition purposes in air-craft engines led to the use of special steels containing about 6 % of tung- sten. A Japanese invention, based upon extensive researches carried out in that country (Honda), has produced a cobalt magnet steel which, in its best examples, gives very surprising results, combining an exceptionally high coercive force with a relatively large rema- nence. By the use of this steel a much smaller and lighter magnet would suffice for a magneto of given power. The steel is extremely expensive, and when supplied in quantity appears to vary in quality, while there is also some difficulty in its workshop manipulation! These, however, are probably difficulties arising from the novelty of the product and are likely to disappear as it becomes better under- stood by the makers.
Welding. One of the most remarkable developments (metal- lurgical in the wider sense) during the period under review has been that of autogenous welding, both by the oxy-acetylene flame and by the electric arc. Both these processes afford a relatively very cheap and simple means of making joints in metal, particularly in iron and steel, and as the joint consists of " the same metal " as that which is being joined, there is a specious suggestion that the joint is " perfect " in the sense of being as good as the unjointed portions of metal. Although it is. quite possible to obtain welded test-pieces which break, under a tensile test, away from the actual joint, the joint itself can never be regarded as equal in strength and toughness to the unjointed steel. There are several reasons for this conclusion. In the first place, the material in the weld itself has simply solidified from fusion and is at best equal in properties to the same steel in the cast condition, while the rest of the plate itself has been immensely improved in quality by forging and rolling and possibly by heat- treatment. Further, adjacent to every such weld there is a region of steel which has either been severely overheated or a little farther away which has been heated to a temperature just below the critical range. In both these regions the steel is seriously weakened and it is in the latter that test-pieces generally break. The most serious difficulty, however, is that of being sure that any autogenous weld is truly sound. Examination of many such welds has shown that soundness is very difficult to secure and that it is the exception rather than the rule, even in careful practice; while it is not possible to ascertain by any external examination of a weld whether it is sound or not. Examination by the aid of a powerful X-ray installation can sometimes be employed to assure the soundness of an important weld, but as a rule this is not feasible. It would seem, therefore and the best-informed opinion is steadily coming to take this view that welded joints cannot be relied upon to carry severe working stresses, and that they should not be employed in vital parts unless an exceptionally heavy factor of safety can be allowed. Actual ex- perience in aeroplane construction has borne out this view, and a construction in which steel tubes are joined together by pinned and soft-soldered joints has been found more reliable than autogenous welding, provided that the working stress on the solder is kept to a low value. On the other hand, it must be borne in mind that the welding processes afford a ready means of making joints and effectin repairs where no other process could be used, and under war con ' tions particularly rapid repairs were frequently executed with gn success. None the less, many such cases involved very decide " war risks." At the same time, welded joints should not be con pared with the unjointed material but rather with joints made : other ways, such as riveting; and there the comparison is much me favourable except for the serious element of uncertainty. The app cation of welding, and particularly of electric arc welding to sue purposes as ship construction, has, however, found considerab' acceptance (see WELDING).
NON-FERROUS METALS. The outstanding feature in tl progress of non-ferrous metallurgy is undoubtedly summed in the one word, " flotation." In spite of long-continued litig tion, which has resulted in clouding the whole matter in obscurity from which published scientific research (Sulma Langmuir) has not yet fully extricated it, this process has ma enormous progress and has to a considerable extent revolution ized the entire practice of the extraction of many non-ferrou metals, particularly those occurring in the form of sulphid minerals, such as galena, zinc-blende and the various pyrit copper ore's. Its effects have been direct in superseding mo gravitational methods of separation, and indirect in view of i fact that flotation deals primarily with very finely divide material, including the " slimes " which were the greatest dif culty of the pre-flotation metallurgist. Not only has this affecte ore-grinding and handling practice, but it has brought about great change in smelting practice also. Thus the treatment copper concentrates is being carried out to a rapidly increasir extent in the reverberatory furnace to the steady exclusion the blast furnace.
