chance to set under natural condition, being relieved from a too rapid freezing action with consequent formation of undue amounts of combined carbon. The metal, therefore becomes softer, is machined more easily, is freed from gas and pin holes, undue casting strains and has less internal shrinkage than where the metal suffers from more or less oxidation through imperfect melting practice.

Tests Verify Other Investigations

The above tests, made in the foundry of the writer, were undertaken to supplement a long series of daily tests at the Chicago Hardware Foundry Co., where a uniformly good increase in strength, machinability and soundness of castings was found as the result of cerium additions. The results at the two foundries are about the same. The interesting feature of the analyses made is that with additions of as much as half a per cent of cerium none could be found in the castings. Evidently the avidity of cerium for oxygen is so strong that after a portion has been used up in the molten metal, the balance must have been oxidized by continued contact with the air over the ladle. It is one of the noticeable features of the use of this alloy that much slag is taken from the molten metal, partly through increasing the fluidity and the balance from the oxidation products of the alloy itself. The meta! cerium and its concomitant elements lanthanum, etc., seem to be particularly powerful deoxidizing agents, whereas other deoxidizers seems to have this property limited up to a certain point, any excess remaining in the casting.

Further work will be undertaken to get more light on the properties of this new ferroalloy and its effect on the various classes of cast iron. It is a well understood axiom in foundry work that the purer the metal as delivered from cupola or furnace, the better the chances for good castings.. Hence anything that makes for purity, in addition to taking every precaution to get good stock, to melt it right and to gate the molds properly, is to be welcomed.

Discussion-Cerium in Cast Iron

THE CHAIRMAN, MR. W. A. JANSSEN.-Can cerium be present as a ferrocerium alloy?

DR. RICHARD MOLDENKE.-Yes, it has to be, because cerium alone is so easily oxidized that it has to be carried with a large percentage of iron. This is then added to the ladle of molten metal. In adding all these ferroalloys to the ladle, it should be in powdered form and the best way is to sprinkle it on the stream of molten iron as it issues from the taphole of the cupola.

MR. E. F. CONE.-I want to ask if Dr. Moldenke has given the composition of ferrocerium and if there are any other elements in it of importance, like titanium?

DR. RICHARD MOLDENKE.-As far as I know, only pure iron is added to the misch metal, as they call it, and consequently the composition would be iron, cerium and several other metals of this group.

Considerations Affecting Brass Melting in the Gray Iron Shop

By R. R. CLARKE, Seattle

Melting stands high among the particulars of brass practice. Representative experience will scarcely dissent from the opinion that inferior castings more frequently result from incorrect mixing and melting than from all other causes combined. Molders make their own scrap; furnaces can make everybody's.

All reputable brass foundries appreciate the significance of melting and equip along the lines and policies dictated by experience. The iron shop making brass occasionally is not so favorably situated. Lack of time, knowledge or inclination sponsors neglect of the importance of melting and often forces the inevitable consequence in the castings. Discussing the question from the iron viewpoint, we can scarcely realize that exhaustion of detail so absolute to exclusive brass experience. We shall, however, aim to cover the general features, hoping that those interested may be able of themselves to fill in the more important particulars.

From a melting standpoint, nonferrous is by no means ferrous; neither is an alloy of either a combination of the other, nor is any alloy of any constituency identical with any alloy of any other composition or to any one metal element entering into the composition. Every metal element has its melting peculiarities in both single and combined state and these broad facts of difference are fundamentally requisite to results. With them iron men must strive to do as brass men do, recognize and reckon with these differences.

The subject may best be discussed under its separate heads. First let us consider metal selection and mixing practice. Common practice among iron men leans toward loose brass scrap of unknown composition and indefinite antecedents. This tendency is a bad one in that all the constituents in the different nonferrous alloys are not congenial to each other when thrown together in a conglomerated mass. Thus a quantity of loose scrap containing some yellow brass, some red, some Tobin

bronze, manganese bronze, Muntz metal, some phosphor bronze and some aluminum bronze for instance-and in indiscriminate scrap it is easy to find all these types of alloy-would make a bad metal to handle besides yielding physical properties in the casting entirely unfit for any practical purpose in general. Experienced brass men can fairly well judge brass scrap by color, fracture, etc., though the best brass foundries go farther than that through analysis in their laboratories. To the average iron foundry these advantages are denied so the better policy is to purchase and use selected scrap approximating known composition. It would perhaps be better still to purchase scrap brass ingot under analysis; this comes but slightly higher than loose scrap. In so purchasing, it is well to patronize reputable sources, since not all the firms making scrap ingot turn out a high grade product. Scrap ingot usually may be divided between that containing zinc or that high in zinc and that including no zinc or low in zinc.

Zinc is highly determinative of certain alloy qualities. It controls color, density, toughness, malleability and cleanliness but loses out in hardness and in antifrictional qualities. Thus we want zinc in golden color metals, in most pressure metals and in the softer, tougher metals. But we cannot have much of it in the harder metals or in the bearing metals. On tin and lead we must rely for these respective requisites.

If the iron foundries doing occasional brass work would keep two kinds of ingots on hand, the one inclusive and the other exclusive of zinc, and at the same time carry a limited stock of virgin copper, tin, lead and zinc they could approximate at reasonable cost a high grade alloy of most any desired constituency. They could do this by adding one or more new metals at the expense of others in the ingot as the case might require. Out of an ingot approximating copper 80, tin 10 and lead 10, which is an excellent bearing metal, they could by adding 90 pounds of copper and 10 pounds of zinc to 100 pounds of the ingot realize a high grade red brass applicable to the average purpose. This metal would approximate very closely the formula copper 85, tin 5, lead 5 and zinc 5, which is quite common and reputable in brass foundry work.

When loose and indiscriminate scrap must be resorted to, some effort at least should be made to pick it over and get the best for the more particular cases. The better grades of brass will usually be found in such castings as valve bodies, stems, disks, bonnets, glands, plugs or keys; and also in the better class of plumbing goods, in locomotive steam castings, and in most cases where the casting is known to have rendered some particular red brass service.

In making up alloys from new metals altogether, the order of adding the metals is important. The general and safe rule is to melt the copper to a fair liquid state, add the tin and lead and finally the zinc, stirring the bath well during the entire process of adding all metals. The zinc should be added in small rather than large pieces and thoroughly stirred into the bath. Though somewhat foreign to the subject, a passing glance at the more common metals used in brass alloys and their effects on each other might be worth while. These metals are copper, tin, lead and zinc. An idea of the function of each will appear in the following table: