the bearing be made of hard metal, considerable friction is set up and a battle will take place between the axle and the bearing, the softer of the two being worn away, causing the axle to swerve considerably. By the use of a soft metal "liner" the axle is not worn, and adapts itself to the condition of the bearing, and runs with much less friction than in the previous case, and, as stated before, the lining is easily replaced when worn away too much. This tendency of white metal to reduce friction has given rise to the term "anti-friction" metal as a name for such alloys. The points of great importance in a bearing are: (1) that it should not cut the journal, (2) should be durable, (3) not become heated by friction, (4) sufficiently soft to adapt itself to the bearing surface, and (5) the metal should be capable of being readily melted and cast. Red metal bearings are distinguished by great hardness and power of resistance, and used where the speed is high and the pressure great. § 86. Babbit's Anti-friction Metal. This was once largely used, but is now replaced in many cases by other alloys. Mr. Babbit recommended to melt together 4 lbs. copper, 8 lbs. antimony, and 24 lbs. tin. This he called hardening. For every pound of the above, he added 2 lbs. more tin. Since its introduction many different mixtures have been sold under the same name, some of the tin being replaced by zinc and lead. One receipt gives 2 lbs. antimony, 2 lbs. tin, and 20 lbs. lead. The bearing to be lined with Babbit's metal is recommended to be washed with alcohol, and powdered over with sal-ammoniac; and those surfaces which are not to be coated, are to be covered with a clay-wash. It is then to be warmed sufficiently to volatilise a portion of the sal-ammoniac, and then tinned. § 87. A white metal alloy for bearings has recently been introduced, under the fanciful title of "magnolia metal," and appears to have found considerable patronage for marine and railway work. Professor R. H. Smith of Mason's College, Birmingham, has submitted this metal to a variety of tests, and states "that it is much superior to either Babbit's metal or gun-metal. It produces less friction, keeps the bearing temperature lower, requires less lubrication, and possesses greater durability. This characteristic of durability is a most important one. Within the wide limits covered by my tests, it would be true to say that the longer the magnolia metal is used, and the more severe the duty imposed upon it, the better becomes its condition. The elevation of bearing temperature above that of the surrounding air is extremely low. The general conclusion I have arrived at from my experiments is, that magnolia metal is a very excellent metal for bearings; that its specially good qualities appear more particularly when it is subjected to intense pressures, such as could not be borne by other metals without firing or melting, and that under very trying circumstances magnolia metal may be trusted to remain cool—that is, at a temperature that does not interfere with good working.' An analysis of this metal, made by Mr. Dean in the author's laboratory, gave approximately § 88. The following table shows the composition of various white alloys for bearings: MELTING AND CASTING OF BRONZE § 89. Bronze is prepared in crucibles, and in reverberatory furnaces, according to the quantity required for casting. The general rule is to melt the copper first, then add the scrap-metal, and when these have been thoroughly melted together, to add the tin, or tin and zinc, as the case may be, previously heated as much as possible without melting. The addition of the tin cools the copper, and as it is advisable to get the whole charge melted and mixed as quickly as possible, in order to prevent loss by oxidation and volatilisation, the fire should be kept very brisk while the alloying is taking place. Bronze contracts on solidifying, as with other copper alloys, but the amount of contraction depends on the composition and on the temperature at which it is cast, varying from to of its bulk. 1 130 The difficulties attending the casting of bronze are much the same as those already discussed with regard to brass. The tin shows a greater tendency than copper to unite with oxygen, so that in remelting bronze the alloy becomes a little richer in copper, and therefore a slight excess of tin should be added in the first place, to supply the subsequent loss. The metals should be excluded as much as possible from the atmosphere during melting, by covering the surface with charcoal or anthracite powder, or oxides will be formed, which will be dissolved by the molten alloy, and impair its strength and toughness. Moreover, bronze has the property of absorbing air and other gases when in the molten state, and emitting them as the surface solidifies. If the castings are thick, or the cast metal rapidly chilled, the absorbed gases cannot escape from the interior, and therefore produce numerous small hollow spaces or pin-holes. Gases are removed more perfectly from molten metal by mechanical agitation, such as vigorous stirring with a suitable rod, and if the rod be of wood, or carbonaceous material, some of the dissolved oxides are decomposed, the oxygen combining with the carbon and escaping as a gas. The other impurities also more readily rise to the surface through the agitation, and form dross. It is probably for a similar reason that phosphorus, manganese, and other reducing agents, when added in small quantity, are so efficacious in producing sound castings. The loss of metal in melting, as well as the quality of the product, is influenced by the construction and arrangement of the furnace. In a reverberatory-furnace a neutral or reducing atmosphere should be maintained, so as to avoid unnecessary oxidation, but the chief point is to get the |