for distribution. He should never advocate polyphase currents Mr. Esson. for working tramway motors, because direct currents were for this purpose so much more simple; but in the particular case of the Sheba Mine it was thought better to use the induction motor as being far more suitable for the purpose than the continuous

current motor.

Correspondence.

Mr. C. F. HEATHCOTE observed that the Author had mentioned Mr. Heathcote. the difficulty of transport to the mining fields, and the necessity of keeping as few spare parts as possible; but he thought the difficulty of obtaining skilled labour was not sufficiently emphasized, and was seldom taken into account by English manufacturers of mining machinery. It was of greater importance than a slight increase of efficiency that all machinery intended for new or "one-mine" fields should be simple in construction, easily repaired, and suitable for constant running, by having reliable dust-, dirtand water-proof bearings with sight-feed lubrication.

Mr. ALFRED E. HUNT noticed that the Author referred only to Mr. Hunt. copper as an electrical conductor, and desired to point out that aluminium was now, especially in the United States, rapidly finding favour in such application on account of its cost rendering it a cheaper conductor. The specific gravity of copper was 8.93, and its electrical conductivity when pure was 100 in the Matthiessen scale, but as ordinarily used in electrical conductors it was 97.61. Its tensile strength was between 16,500 lbs. per square inch and 65,000 lbs. per square inch in hard-drawn bars, and its cost was 14 cents per lb. in the United States, the equivalent selling price being 130 marks per hektogram in Germany, for wire, bars, and rods such as were used for electrical conductors. Aluminium had a specific gravity of 2.68, an electrical conductivity in commercially pure metal of 63.0, a tensile strength in pure soft wire of 26,000 lbs. per square inch, and in hard-drawn rods of wire of 40,000 lbs. per square inch. The Pittsburgh Reduction Company in the United States were selling rods, bars, plates, and wire drawn down to 0.08 inch in diameter, in large orders for electrical conductors at the rate of 29 cents per lb. at their works. This price was below the regular rate for aluminium, offered for the special purpose of meeting the price of copper for electrical conductors alone, in order to favour the adoption of aluminium for this purpose and to overcome the [THE INST. C.E. VOL. CXXXV.]

H

2.689

ΤΟ

Mr. Hunt. handicap that aluminium had occasioned by its lower conductivity. It was therefore evident that (1) any given volume of copper was 8.93 or 3.332 times heavier than an equal volume of aluminium; (2) the equal price of 14 cents per lb. for copper for any length of any equivalent section of aluminium wire or bar would be 14 cents times the factor 3.332, or 46.65 cents per lb., that was, 1,000 feet of wire of, say, inch diameter would cost as much if bought of copper at 14 cents per lb. or aluminium at 46 65 cents per lb. ; (3) reckoning the copper conductor to have its maximum of 100 per cent. conductivity, and the aluminium to have a conductivity of 60 per cent. (which the Pittsburgh Reduction Company were ready to guarantee for their special pure aluminium metal for electrical conductors), then for an equivalent electrical conductivity a given section of copper that could be placed at 100 should be increased in area to about 160 to give an equal conductivity; (4) the relative specific gravities were such that the weight of the given equal length of the aluminium conductor with 160 sectional area will be only 48 per cent. of the weight of the copper conductor with sectional area of 100; (5) aluminium at 29 cents per lb. was therefore a more economical conductor than copper at 14 cents per lb. Taking as an illustration an aluminium conductor to replace a copper wire of about 1 inch diameter, the aluminium wire of equal, in fact, somewhat superior, electrical conductivity would be slightly over inch diameter. The weight of 1 mile of the copper wire was 162.32 lbs., and its cost at 14 cents per lb. would be $22.72. The weight of 1 mile of the aluminium wire would be 79.46 lbs., and at 29 cents per lb. would cost $23.04. Forty-eight per cent. of the weight of the copper wire, which would give equal electrical conductivity in aluminium wire, would weigh only 77-91 lbs., so that more accurately $22.59 would be the cost of 1 mile of aluminium wire at 29 cents per lb. to replace 1 mile of copper wire at 14 cents per lb. costing $22.72. It appeared from the Paper that the Sheba line consisted of nineteen wires 0.057 inch diameter, and that the outer conductors had also nineteen wires of the same diameter, which were stranded in two segments. This copper wire weighed 9.89 lbs. per 1,000 feet. Aluminium wire of the same diameter would weigh 2.968 lbs. per 1,000 feet. Aluminium wire of equal electrical conductivity with the copper wire would weigh 4.75 lbs. per 1,000 feet. Nineteen wires in copper would therefore weigh 187 91 lbs. per 1,000 feet, which, reckoned at 15 cents per lb. as the cost for copper (and probably the rate paid for the copper delivered in Africa was considerably higher) would give a cost per 1,000 feet

of $28. 19. Aluminium wire of equal electrical conductivity, each Mr. Hunt. strand of which would weigh 4.75 lbs. per 1,000 feet, would weigh 91.25 lbs. per 1,000 feet. At the rate of 29 cents per lb. this would cost $26-16, or a difference of $2.03 per 1,000 feet of the nineteen-strand aluminium wire of equal conductivity to the copper. Added to this sum should be a considerable amount of saving in transport, bearing in mind that a little less than onehalf the weight of the wire in the electrical conductor was required to give equal electrical conductivity in aluminium. Against this figure, however, for a case like that referred to by the Author, was the added cost for insulation and the added weight of the insulation. The actual amount of surface to be insulated, and therefore of materials to furnish this insulation, was almost exactly one-fifth. However, the cost of applying the insulation was not much greater for the larger section than for the smaller, and it had been proved by experience in the United States that about one-sixth of added cost for insulation was occasioned by the increased section of the aluminium over copper of equal electrical conductivity. Aluminium had to meet this very considerable added handicap, which did not occur where naked wires were used for conductors. It, however, only resolved itself into a question of savings and the relative advantages of the two metals for any purpose. In a considerable number of conductors in the United States in the past year the lowered price of aluminium had been cheerfully made in order to overcome this added cost for insulation. (6) The continental requirement in tensile strength for soft copper wire, rods, and bars used as electrical conductors was 22 kilograms per square millimetre, the English requirement being 14 tons per square inch, and the American requirement was about its equivalent of 32,000 lbs. per square inch. Aluminium wire, rods, and bars would be furnished of 60 per cent. electrical conductivity, which would have an equal tensile strength per unit of area with the copper, and therefore with the electrical conductivity equivalent of 48 per cent. of the weight of the copper and sectional area of 160 against the area of the copper section 100; the tensile strength of the aluminium conductors would be as 100 for the copper was to 160 for the aluminium. This would mean, if a square inch of copper conductor was used of, say, 32,000 lbs. per square inch tensile strength, the equal conductivity area of 1.6 inch of aluminium would have a tensile strength of 51,200 lbs. It had been already determined that with aerial lines the snow and ice load was practically as heavy on lengths of small wire as upon

Mr. Hunt. larger sections, so that no objection upon this score could probably be found to the use of the larger sections of aluminium wire. Both on account of having only 48 per cent. of the weight and on account of having about 60 per cent. more strength the aluminium conductor could be used in much longer spans between supports, and the number of expensive poles and insulators could be materially diminished. Properly drawn aluminium wire was as tough and would stand bending as severely without breaking as soft copper wire. The toughness of aluminium wire was, however, greatly modified by the care and skill used in manufacture. If it was drawn too severely through the dies, or was not well annealed at the proper intervals in the drawing operation, it was finished much more brittle than when properly manipulated. Hard-drawn copper wire, especially that in the smaller sections drawn through diamond dies, was furnished with a tensile strength of 65,000 lbs. per square inch. The maximum tensile strength of the best pure hard-drawn aluminium reached under similar favourable conditions for developing the maximum tensile results had not yet been determined, but from experiments already made it could be predicted that at least 50,000 lbs. per square inch and perhaps even higher strength could be obtained. Aluminium hardened with a small percentage of alloying ingredients could be furnished in wire with a tensile strength far in excess of what could be obtained in pure aluminium. Experiments were now being made by the Pittsburgh Reduction Company to determine the alloy that would furnish the maximum tensile strength, together with maximum electrical conductivity. From results already obtained it could surely be predicted that an alloy of aluminium could be furnished which, drawn into wire, would have a tensile strength of at least 100,000 lbs. per square inch and electrical conductivity of about 50 in the Matthiessen scale, this material to rival the silicon-bronze materials which were now in use, where maximum tensile strength, together with good electrical conductivity, was required.

The power of withstanding corrosion was greatly in favour of aluminium as an electrical conductor over copper. Copper did not change in dry air, but in moist air it became covered with a green layer of basic carbonate of copper, which itself had a corroding action and did not coat and protect the underlying copper from further corrosion. Ammonia in contact with copper absorbed oxygen, and the copper dissolved in consequence of the formation of a soluble cupric-oxide and ammonia. This action was especially troublesome where copper wire was used in connecting rail-joints

in the lines of electrical railways where the ground was often M. Hunt. subjected to ammoniacal solutions. Aluminium similarly was not acted upon in dry air, and the corrosion in moist air was of the oxide of aluminium, alumina, a harmless salt which formed an impenetrable coating on the metal and protected it from further corrosion to a considerable extent. Ammonia solutions acted on aluminium only upon the surface, attacking it and leaving behind a more resisting surface-coating of a brown colour containing silicon, which resisted corrosion from dilute mineral acids and dilute solutions of organic acids as well as moist and dry air. Subject to sulphur gases such as locomotive flue-gases, aluminium withstood corrosion better than copper. If kept free from galvanic action with any other metals electro-negative to itself, aluminium was far less easily corroded than copper. The difficulty of soldering or brazing aluminium was the chief drawback to its use as an electrical conductor. Aluminium could be soldered strongly, but it was a more difficult and slow operation at best as compared with copper, and there was much more rapid weakening of the soldered joint due to galvanic action between aluminium and the metals of the solder than with the less electro-positive metal copper. Several forms of joints had been successfully used to avoid the necessity of soldering, the best forms being to use thin aluminium sheets to wrap the joints and to twist or otherwise bind with the aluminium bars or wires to be joined. These wrapped and twisted joints with aluminium sheets could be left smooth on the outside, when desired, could be made much stronger than the body of the conductors, and were really a more serviceable job than soldered or brazed work in many cases with copper. One very practical way of making joints of aluminium-wire was to roll the thin aluminium sheet of about 6 inches width into two cylinders from opposite edges of the sheet. These double cylinders were very cheaply made on mandrils in a lathe. The ends of the wires to be joined were inserted in these cylinders from opposite ends and both the wire and sheet twisted with pliers until a firm joint was secured, much stronger than the body of the wire. The joint could readily be made impervious to the air and moisture. There were certain places where aluminium would be at a disadvantage over the smaller section of equal conductivity of copper, in ducts or conduits, where space was a considerable item. In such cases, the use of aluminium would necessarily be prevented. Aluminium, next to gold, was the most malleable of all the metals and was much more malleable than copper. Aluminium in ductility stood next to copper and was easily drawn into all sections that were