So much interest is just now being excited in the direction of cheap-ening the production of this metal, that no excuse is needed for introducing here in the form of an appendix some information on the subject which has appeared in print only since the early portion of this volume was already in type.

In a communication made to the London section of the Society of Chemical Industry in March 1883, Weldon declared that the only manufacturer of aluminium in the world at present is Pechiney, of Salindres, France, and that the process there employed is the same as was worked out by Devilie more than a quarter of a century ago. This process consists of 3 successive operations: (1) The preparation of alumina from bauxite, by furnacing the bauxite in admixture with sodium carbonate, dissolving out the resulting sodium aluminate, and then treating the solution so obtained by C02; (2) the preparation of double aluminium-sodium chloride, by mixing with carbon the alumina obtained in the first operation, drying the resulting mixture of alumina and carbon, and then heating it in a current of chlorine; and (3) the reduction by free sodium of the A12C16 of the double chloride so obtained. These have been already described on pp. 43-4. Weldon directs attention to the relative proportional cost of each operation, which he states as follows:- If the total cost of a unit of aluminium at Salindres be called 100, the cost of the first operation, including that of the bauxite upon which it is performed, is 9 . 67; that of the second operation is 33 . 4; and that of the third operation, including 3 the cost of the sodium used in it, and also that of the cryolite employed as a flux, is 56.93. The first operation is thus comparatively inexpensive; the second costs 2 1/2 times as much as the first; and the third costs 1 1/3 times as much as the first and second put together.

It follows that, to produce aluminium at a price appreciably lower than the present cost of aluminium at Salindres, either one must produce it by a process quite different from Deville's, or one must cheapen the second, or the third, or both operations of Devi lie's process; since to reduce the cost of its first operation even to nothing at all, the second and third operations of it remaining as costly as at present, would cheapen aluminium only by 9.6 per cent.

Weldon then proceeds to show that Webster's invention relates simply to the obtainment of anhydrous alumina from potash alum. If his method of obtaining alumina were 50 per cent, cheaper than the Salindres method, it would thus be capable of reducing the cost of aluminium only by less than 5 per cent. But his method is twice or three times as costly as the Salindres method. Towards the cheapening of aluminium, therefore, Webster has done nothing at all, in Weldon's opinion.

Next, Weldon alludes to the fact that a considerable number of persons, quite independently of each other, are endeavouring to reduce alumina by carbon. It is often rash to say that a proposed thing can never be done; but that alumina can ever be reduced by carbon is surely impossible. When 2 x 27.3 parts of aluminium combine with 3 x 16 parts of oxygen, there is liberated as heat a quantity of energy capable of raising 391,600 parts of water from 0° C. to 1° C. To abstract, by purely chemical action, the oxygen of the resulting alumina, leaving its aluminium free, one must cause to react upon the alumina a body by whose combination with oxygen more than 391,600 calories are liberated; and carbon is certainly not such a body. For the reaction

2A12O3+3C=2A1 + 3C02 to be possible, the heat of combination of C+O, must be at least

2x391600/3 = 261066 while for

A1203+3C=2A1+3C0 to be possible, the heat of combination of C + O must be at least

391600/3 = 130533.

But the heat of formation of C02 is only 97,000, and that of CO is only 28,800. The often proposed reaction-

A12C16+6H = 2AI+6HC1 is similarly hopeless, since the heat of combination of Al2Cl6is 321,800; while 6 times that of gaseous HC1 is only 6x22,000 = 132,000.

To arguments of this kind it is often objected that carbon reduces sodium from Na2O, notwithstanding that the heat of formation of Na2O is very much greater than the heat of formation of CO. But carbon does not really "reduce " Na2O. When a mixture of carbon and Na20, or, what comes eventually to the same thing, a mixture of carbon and sodium carbonate, is intensely heated, the Na2O volatilizes, and some of its vapour becomes dissociated into free oxygen and vapour of free sodium; and it is not until dissociation has thus taken place that any of the carbon combines with any of the oxygen of the Na20. The carbon acts only by combining, after dissociation has taken place, with some of the oxygen of that part of the Na20 which has been dissociated by heat, and so preventing" re-association on cooling. Unlike Na20, however, Al2O3 does not undergo dissociation at any temperature attainable industrially.

If, then, we cannot hope to reduce A1203 by carbon, or A12C16 by hydrogen, in what direction can we look for the cheapening of aluminium ? It seems to Weldon that there are 4 conceivable resources:

1. The cheapening of Al2Cl6, 2NaCl.;

2. The substitution for Al2Cl6,2NaCl of some other anhydrous compound of Al, not containing oxygen;

3. The substitution for sodium of some cheaper reducing agent; and

4. The cheapening of sodium itself.

Pechiney .has already effected important economies in the production of Al2Cl6,2NaCl, and will probably effect others. The production of a cheaper haloid salt of aluminium than Na2Al2Cl6, would at first blush seem difficult, but is not, Weldon thinks, impossible; and a cheaper agent than Sodium, capable of reducing certain salts of aluminium, he believes he could now indicate, if he were free to do so.

As for cheaper sodium, sodium no doubt will be cheapened. To manufacture a ton of sodium at present costs, roughly, between 300/. and 350/. The quantity of sodium carbonate containing a ton of sodium at present costs in the market about 16/. The manufacturer of sodium, however, obtains only about a third of the sodium contained in his raw material: He effects only a partial dissociation of his Na20, and to that partial dissociation there succeeds partial re-association. Still, 3 times 16/. for raw material leaves 250/. to 300/. for the cost of extracting 1 ton of sodium from about 7 tons of Na2CO,.