obtained, by testing its malleability, colour, and fracture, will generally afford sufficient evidence of the required amount to be added.

In making experimental tests, a small melting furnace, such as that used in a metallurgical laboratory, a strong pair of hand rolls, and an anvil would be very useful adjuncts to every casting shop. The quantity of metal operated upon need not exceed one pound in weight, and as this could be cast in a long strip, its suitability for stamping or rolling could be readily tested. Such test-pieces, if carefully labelled and preserved, would be most valuable for future reference, and there can be no doubt that both employers and employed would thus gain a vast amount of information, which would prove of great benefit both as a standard of workmanship and of economy of production. It is a common experience to find, after a quantity of metal has been mixed and the castings made, that the alloy is unsuitable for the work required of it, either from unsuitable constituents, improper mixing, or impure materials; which annoyance could be avoided by a few preliminary trials on the small scale. The casting of such trial tests could be made in an iron or sand mould, and the time of cooling made to approximate to that of a large mass by judicious treatment. Another advantage of such an experimental plant would be that new combinations could be readily tried, and the effect of certain impurities on well-known alloys ascertained, by purposely adding these bodies in definite amounts to a weighed quantity of the alloy.

The purposes for which alloys are required are endless. Some are required to possess great malleability, for others hardness is the chief requisite; others, again, must possess a high degree of elasticity, while some are useful on account of their low melting point, etc. These different demands can only be satisfied by uniting suitable metals in different proportions.

The metals most often used for alloying at the present time are those which have been known the longest, such as

copper, zinc, lead, tin, gold, and silver; and although combinations of these metals have been known and employed for many centuries, it is only during the latter half of the present century that their intimate properties have been closely studied. Indeed, at the present day our information concerning the nature and properties of alloys is perhaps less than in any other branch of chemical science, and although chemical investigation may do much to enlighten our knowledge, such information will be destitute of great commercial value, unless accompanied with practical knowledge of the working, from observation of the physical properties, when alloys are worked in large quantities, by the manufacturers themselves. The number of simple metals is very limited, but they may be united in various proportions, forming an endless variety of modifications; and since every alloy may be looked upon as a new metal, from the fact of its properties differing from those of its constituents, we have at command the necessary material for producing metals suitable for every requirement for which metallic matter is desirable. The action of metals upon each other is widely divergent ; sometimes one metal may be added to another in quantity without seriously altering its working properties; in other cases a minute quantity of a second metal will altogether change the character of the first metal; so that in alloying it by no means follows, because one metal may be freely added, that another, even of a similar nature, may be as liberally introduced. The man who aspires to the formation of new alloys, or who wishes to produce metals suitable for different requirements as circumstances arise, must be well acquainted with the nature and properties of the simple metals in order to successfully accomplish his object; and although a knowledge of the components is not sufficient of itself, it is of immense advantage in assisting the operator who combines practical experience in mixing metals with this theoretical knowledge. It is for these reasons that a brief account of the elementary metals is included in this work.

In chemical combinations it is a well-known fact that elements always combine with other elements in definite proportions by weight, termed atomic weight, producing compounds of fixed and decided properties, so that the same compounds can be always relied upon to contain the same elements, united in the same proportions. This same law applies to the union of two metals, when such metals are chemically combined, and the same alloy will always have properties identically the same, however it may be tested. Several experimenters have directed their attention to the mixing of metals according to their atomic weights, so as to obtain alloys of determined characteristic properties, but up to the present time the number of such combinations of a useful character is very limited. They are by no means the ones most suited to the wants and requirements of industry. There is always one indispensable item from the manufacturer's point of view which the chemist is not concerned with—that is, the cost of production—and however nicely atomic proportions would suit the requirements of a given alloy, such an alloy would in most cases be useless unless the cost was consistent with the market value. The question then of cost must have consideration, and the proportions must, if possible, be made to fit in with commercial necessities. With regard to copper alloys, such as brass and bronze, the combinations which best exhibit the characters of chemical compounds are hard and brittle, and as copper alloys are much more widely used than any other, there is little inducement to encourage metallurgists to endeavour to alloy copper and zinc, or copper and tin in atomic proportions, since malleability and tenacity are the properties most desired in these alloys. Again, colour is the chief desideratum in many alloys, and this cannot be always obtained by mixing in atomic proportions, especially as it often happens that a very small addition of one of the constituents will alter the shade of colour, so as to produce what is required.

When it is desirable to add a non-metallic element to a

metal or alloy, for the purpose of bringing about a certain result, very much greater care is generally required in apportioning the quantity to be added than with a metal, as non-metals combine much more actively with metals than the metals do with each other, and a very small quantity of a non-metal will suffice to alter the properties of a metal or alloy. It is very surprising to note how, in some instances, a mere trace of another element will alter the properties of a metal. For example, of carbon added to iron will convert it into mild steel; 100 of phosphorus makes copper hot - short; 2000 part of tellurium in bismuth makes it minutely crystalline; 1000 part of antimony in copper renders it exceedingly bad in quality for certain purposes.

207 11.45

=

Lothar Meyer has shown that a remarkable relation exists between the "atomic volumes of the elements." The relative atomic volumes of the elements are found by dividing their atomic weights by their specific gravities. The atomic weight of lead is 207, and its specific gravity 11·45; = 18, the atomic volume of lead. It would appear that the power of an element to produce weakness in a metal, when added in small quantity, is dependent on the atomic volume of the impurity.1 Roberts-Austen tried the effect of various elements on pure gold, and found that when the body added had an atomic volume equal to, or less than that of gold, the strength was little affected, and in some cases, as copper for example, was increased; but when the element added had an atomic volume much greater than that of gold the strength, with two exceptions, was greatly diminished. His results are embodied in the following table :

1 Journal of Soc. of Arts, 19th October 1888.

Alloys generally have properties differing from their constituents, and some have these differences very strongly marked, thus: if a very small quantity of arsenic be added to tin, the resulting alloy will have a crystalline fracture closely resembling zinc. Sometimes metals combine with evolution of heat and sometimes with an absorption of heat. The following metals, according to Roberts-Austen, evolve heat when they are united :--aluminium and copper, platinum and tin, arsenic and antimony, bismuth and lead, gold and just melted tin. On the other hand, lead and tin when they unite absorb heat.

When lead, tin, and bismuth in equivalent proportions, and very finely divided, are mixed with eight equivalents of mercury, at the ordinary temperature, the temperature will fall from 17° C. to -10° C. If the vessel containing the mixture stand on a wet board the water underneath will be frozen. This combination then will form a freezing mixture.

The method of producing alloys by strongly compressing