In non-ferrous metals the problems of corrosion, although dif ferent, are equally important. In regard to the alloys of copper, the production of condenser tubes immune from failure by pit ting still offers a problem although much important progress has been made. In regard to aluminium alloys, the corrosion problem stood for some time in the way of their wider application. The difficulty, has, however, been met by the development of protec tive coatings. Aluminium owes its natural resistance to corrosion to a thin transparent coating of oxide which forms on the metal immediately it is exposed to the air and to a great extent prevents further attack. As naturally formed, however, this is neither strong enough nor impervious enough to prevent corrosion under more difficult conditions, such as exposure to sea-water, etc. It has, however, been found that the natural oxide co gating can be strengthened and thickened by making the metal the anode in a bath containing chromic acid or potassium bichromate and grad ually raising the voltage. Aluminium and many of its alloys, when treated in this way, become very resistant to corrosion and this resistance can be further increased by saturating the "anodic oxi dation" film with fatty matter such as lanoline. While the value of this process has been very fully demonstrated and recognized in England it is curious that reports upon it both from American and German sources are less satisfactory, but this is perhaps due to inadequate attention to important details.
Among further recent developments in the production and treat ment of metals, the great expansion of welding operations in the jointing of steel may be mentioned. The two principal processes are oxyacetylene and electric welding. An advanced technique for both methods has been developed and welded joints are finding in creasing application. While it is admitted that the strength of steel—and especially of heat-treated steel—is locally impaired by the high temperatures applied in welding, yet the great con venience of the process and the saving in weight and labour which it allows as compared with rivetting, outweigh these disadvantages wherever satisfactory allowance for the impaired properties of the steel in the neighbourhood of the weld can be made. It must further be borne in mind that the correct basis of comparison in estimating the value of a welded joint is not the strength of the solid metal but of any alternative joint such as a rivetted seam.
The most serious defect of welding is the difficulty of being sure that a given joint is sound. Reliance must be placed on the skill and conscientiousness of the workman. The examination of welds by means of X-rays is slow and costly and not entirely certain of showing up defective places. In England there is still some hesi tation in using welded joints in structures exposed to severe stresses. In France the process has been very widely used even for boiler construction, but a few serious failures have occurred.
Important advances have also been made in regard to materials for use at high temperatures. The demand for these arises from the effort to improve the thermal efficiency of heat engines— stimulated by the struggle, which has been in progress for a num ber of years, between the steam-turbine and the internal combus tion engine. The result is a demand for steels and non-ferrous al loys capable of withstanding stress and oxidation at temperatures sometimes as high as 800° C. Ordinary steel becomes too weak to be used at these temperatures, and even materials which show a reasonably high strength under an ordinary tensile test carried out at the high temperatures have been found to fail under much lower stresses if exposed to them for a sufficiently long time. This has led to the conception of the "limiting creep stress" which is measured by determining the highest stress which the material can bear, at a constantly maintained high temperature, for long pe riods of time without undergoing measurable changes of length. At 800° C some of the best materials yet available commercially show a creep stress limit of little more than two tons per square inch. For so high a temperature, however, this constitutes a marked achievement and has only been attained by a few special "heat-resisting" steels and by one type of non-ferrous alloy. These steels contain relatively large proportions of nickel, chromium and tungsten, the combined presence of all three being essential. The non-ferrous alloys suitable for high temperature use consist mainly of nickel and chromium. For less extreme conditions, such as those encountered in the construction of high pressure steam boilers, turbine blading, etc., where 400° C is not likely to be ex ceeded in the near future, less expensive types of material are available.