A certain amount of preliminary experimental work was perforce entailed, both for ensuring that the band, when cast, should exhibit an estimated degree of shrinkage (for the greater the shrinkage the sounder the casting, a shrinkage of 1/4 inch being usually accepted as a minimum), and that the metal should be capable of withstanding certain specified Government tests, the one condition, of course, being contingent on the other. Tests which were actually made proved wholly satisfactory, the average results being an elastic limit of 7·5 tons per square inch, an elongation of 45 per cent. on two inches, and finally a breaking stress of 14·5 tons. De-oxidisation, that is to say the process of removing oxygen from the metal for the purpose of obtaining castings that were sound, free of blow-holes and of oxide of copper, was effected by mixing a small percentage of phosphor-copper with the molten copper in the crucible. Boron-copper was also tried as a de-oxidiser, but no real advantage was noticeable. Comparative tests, too, were made for the purpose of ascertaining the percentage of loss of copper when melted in crucibles and again in a reverberatory furnace. The former process resulted in a loss not exceeding 3/4 or 75 per cent.; whereas the latter was responsible for a 7 per cent. loss. Any saving which came within the meaning of the word "economy" as completely removed and distinct from that of "parsimony" was a precept not merely preached but extensively practised throughout the locomotive department at Crewe, and as an illustrative instance of this praiseworthy, and therefore patriotic, policy the casting of copper bands may be cited. Although cheaper and possibly less reliable methods of producing copper bands may have conceivably come into being during the final stages of the war, it was obvious to even the least well-informed in such matters that, provided a mixture could be obtained whereby the metal could be relied upon to pass the Government tests, the process of pouring molten copper from a crucible into a sand-cored cast-iron chill was likely to be at any rate cheaper than that involving the employment of presses, rolls, shears, and punches. On comparing the estimated cost of manufacturing copper bands by the pressing and casting processes respectively, a difference of one shilling per finished band was shown in favour of the latter system; and although it may seem a mere bagatelle, a drop in the ocean of squandered millions, to those who not merely are encouraged, but who encourage others, in the art of reckless and profligate extravagance when handling the public purse, this modest shilling per copper band saved represented an aggregate of £1750, a sum not altogether to be sneezed at when we consider that the value of the 35,000 bands cast in Crewe Works was but an infinitesimal fraction of the total munition expenditure during the war.