sitates the employment of the purest copper; electrolytic copper is sometimes used when not too highpriced, but Lake Superior copper is generally found satisfactory enough. Even the purest copper may contain dissolved cuprous oxide or occluded gases, and it is one of the functions of aluminium to reduce these oxides and gases, forming slag, which rises to the surface, and leaving the bronze free from their influences. If tin occurs in the copper it lowers very greatly the ductility and strength of the bronzes, but zinc is not so harmful.

Care should also be taken as to the purity of the aluminium used, though its impurities are not so harmful as they would be if occurring in similar percentage in copper, since so much more copper than aluminium is used in these alloys. Yet the bronzes are so sensitive to the presence of iron that an aluminium with as small a percentage of this metal as possible should be used. The silicon in commercial aluminium is not so harmful as the iron, but it does harden the bronze considerably and increases its tensile strength. The purest aluminium alloyed with the purest copper always produces the highest quality of bronze.

For preparing the bronzes the following directions. are given: Melt the copper in a plumbago crucible

and heat it somewhat hotter than its melting point. When quite fluid and the surface clean, sticks of aluminium of a suitable size are taken in tongs and pushed down under the surface, thus protecting the aluminium from oxidation. The first effect is necessarily to chill the copper more or less in contact with the aluminium, but if the copper was at a good neat to start with, the chilled part is speedily dissolved and the aluminium attacked. The chemical action of the aluminium is then shown by a rise of temperature which may even reach a white heat Considerable commotion may take place at first, but this gradually subsides. When the required aluminium has been introduced, the bronze is let stand for a few minutes and then well stirred, taking care not to rub or scrape the sides of the crucible. By the stirring, the slag, which commences to rise even during the alloying, is brought almost entirely to the surface. The crucible is then taken out of the furnace, the slag removed from the surface with a skimmer, the metal again stirred to bring up what little slag may still remain in it, and is then ready for casting. It is very injurious to leave it longer in the fire than is absolutely necessary. No flux is required, the bronze needing only to be covered with charcoal powder. The particular point to be

attended to in melting these bronzes is to handle as quickly as possible when once melted.

The manufacture of aluminium-bronzes on a large scale, as carried on by the Cowles Electric Smelting and Aluminium Company, is as follows: The furnace used consists of a brick box 1 foot wide, 5 feet long and 15 inches deep. From opposite ends enter two immense electrodes, that are really electric-light carbons, 3 inches in diameter and 30 inches long. These are partly contained in pipes that, in turn, pass through stuffing boxes in the ends, to exclude the air and, at the same time, admit of adjusting the electrodes.

To protect the walls of the furnace from the intense heat, it is lined with finely-powdered charcoal, which, having been first washed in a solution of lime-water, retains its non-conductivity even after the particles have been partially converted into graphite by heat.

The bottom of the furnace is now lined to a depth of two or three inches with this fine, prepared charcoal, and by means of a sheet-iron gauge, the walls of the furnace are covered with charcoal to the thickness of two inches.

The charge, consisting of about 25 lbs. of corundum, 12 lbs. charcoal and carbon, and 50 lbs. of

granulated copper, is placed about the electrodes to within a foot of each end of the furnace. A layer of coarsely-broken charcoal is now spread over the charge, and the sheet-iron gauge withdrawn. The coarse charcoal on top allows the escape of carbonic oxide gas formed during the process. An iron cover, lined with fire-brick, is luted on to prevent the entrance of air.

The charge is now prepared, and the furnace ready to be connected with a large Brush dynamo, capable of producing ninety horse-power of electric energy. In the circuit between the dynamo and furnace is an ammeter, designed to register from 50 to 20,000 ampères of current, which is controlled by a large resistance-box, as the ends of the electrodes may at first be too close together to make it safe to start the dynamo. By watching the ammeter and moving the electrodes, the resistance-box can be taken gradually out of circuit without producing a "short circuit" at the beginning of the operation. In about ten minutes, after the copper about the electrodes has become melted, the latter are slowly moved apart until the current becomes steady. It is now increased to about 1300 ampères and fifty volts. Carbonic oxide begins to escape from the orifices made in the top, and burns in two white

plumes of flame. By regulating the distance between the electrodes, the current is kept constant for about 5 hours, and all parts of the charge are brought into the reducing zone.

The

When the operation is completed, a resistance is placed in the box, and the current is switched into another furnace charged in a similar manner. product is an alloy of copper, containing 15 to 30 per cent. of aluminium, and having a beautiful silver color when broken. The copper performs no part in the reduction, but is employed to absorb the aluminium, which would otherwise be converted into a carbide.

This alloy is now melted in an ordinary crucible furnace and run into ingots, which, after being analyzed, are re-melted, and sufficient copper added to produce the standard bronzes.

Two runs from the furnace described will produce about 100 pounds, containing about 15 per cent. of aluminium.

When a 10 per cent. aluminium bronze is made by simple mixing of ingredients, it is brittle, and does not acquire its best qualities until having been cast several times. After three or four meltings it reaches a maximum, at which point it may be melted several times without sensible change. As it cools