August 8, 1901J 



NA TURE 



;6[ 



the method by which the metals are prepared. If they are in a 

 ■chemically active state, as lead is which has been prepared from 

 tartrate of lead, they will, in many cases, take fire in air and 

 burn at the ordinary temperature. This lead burns readily when 

 shaken in air. If this mass of uranium, for which I am indebted 

 to M. Moissan, is filed in air, the detached particles will ignite, 

 lletallic iron which has been reduced by hydrogen from its 

 oxide at a temperature below 700° C. will also lake fire and 

 burn in air at the ordinary temperature, a point of extraordinary 

 interest in relation to the allotropy of iron (Osmond and Cartaud, 

 Ann. des Mines, August 1900). Metals in this chemically active 

 state are said to be " pyrophoric." 



So far as I am aware, metals in this chemically active state 

 have not been used as fuels. Neither am I aware that any use 

 has been made of the allotropy of metals as enabling them to 

 be used as fuel, but Prof. Graham once told me that pyrophoric 

 iron had been suggested for warming ladies' muffs, the intention 

 being to place the iron in a small receptacle and to admit air 

 gradually as warmth was needed. Sir Henry Trueman Wood 

 also remembers the suggestion, but tells me that he can find no 

 record of it in \he Journals of the Society of Arts. I may just 

 mention that the burning of metallic antimony plays a very 

 important part in roasting silver ores, and the behaviour of the 

 metal is so peculiar while burning that I must pause to show it 

 you. {A. melted globule of antimony, if thrown on to a tray of 

 paper, darts about and cannons from the sides, leaving a track 

 of dark oxide on the paper.] 



The metal I am going to employ as fuel is aluminium, the 

 o.Kygen for its combustion being supplied by metallic oxides, 

 which readily part with their ogxyen to aluminium if it be 

 raised to certain definite temperatures. This question of the 

 transference of oxygen from one metal to another, which results 

 in the liberation of the metal attacked, is of special interest to 

 us at the Royal Institution, for it undoubtedly originated within 

 these walls and is due to Sir Humphry Davy. He discovered 

 potassium in 1S07, and in 1S09 attempted to remove the oxygen 

 from alumina by heating it with metallic potassium. He says 

 (Phil. Trans., part i. iSio, p. 60), " if I had succeeded in iso- 

 lating the metal I should have called it «/««/«/«." His success 

 was iiTiperfect, but he certainly did obtain, by the intervention 

 of metallic potassium, an alloy of aluminium and iron. It 

 retrained for Wohler to prepare pure metallic aluminium from 

 its chloride in 1827, and for Henri Saint Claire Deville, who 

 began to work in 1854, to establish the metallurgy of aluminiuni 

 on an industrial scale. As regards the reduction of metals from 

 their chlorides, Wohler [Ann. dcr Cheinie, vol. cvi. p. iiS) 

 obtained crystalline compounds of chromium and aluminium 

 and Michel [ibid., vol. cxv. p. 102 ; ibid., vol. cxiii. p. 248) 

 compounds of aluminium with manganese, iron, nickel, tung- 

 sten, molybdenum and titanium. Levy (Coinples rendtis, 

 vol. cvi. p. 66) obtained an alloy of titanium and aluminium, 

 Beketoff {.-i««. der Cheniie, vol. ex. p. 374) an alloy of barium 

 with aluminium from the chloride of barium mixed with baryta. 

 Dr. Goldschmidt (ibid.. May 1S98) has given references to 

 these authorities in a recent valuable paper. In 1856, Charles 

 and Alexandre Tissier (Comftes rendus, vol. xliii. 1856, p. 

 11S7) observed the fact which is the starting point of the ex- 

 periments I have to show you. They found that aluminium 

 decomposes the oxides of lead and of copper, miich heat being 

 evolved by the reaction. 



They do not appear to have used aluminium in a finely 

 divided state, and therefore failed to reduce certain metals 

 from their oxides which are now known to be perfectly easy to 

 reduce. It was not until comparatively recently that the use of 

 aluminium for separating other metals from their oxides assumed 

 serious proportions. Claude Vautin showed on June 13, 1894, 

 at a soiree of the Royal Society, a few metals, and among them 

 carbon-free chromium and manganese, which he had prepared, 

 and as he undoubtedly gave the impulse that started much of 

 the subsequent work in this direction, it may be well to give the 

 description which was appended to the specimens he showed. 

 It runs as follows : 



Specimens of Metallic Chromiitnt, Manganese, Tungsten Iron, 



Ss'c, free from Carbon, also fused Alumina, obtained 



durin« reduction of the metallic samples. 



The specimens of metallic chromium, m.inganese, &c., have been reduced 



from their o.vldes by means of metallic aluminium. The oxide of the metal 



to be reduced is intimately mixed with finely divided aluminium, and heated 



in magnesia-lined crucibles. The heat produced by the oxidation of 



aluminium daring the operation is sufficient to fuse alumina, a specimen 



of which is exhibited. 



NO. 1658, VOL. 64] 



The subject is, however, in a sense your own, for, so far as I 

 know, the lecture on "The Rarer Metals and their Alloys" 

 (Nature, May 2 and 9, 1895), «'hich I delivered here 

 in 1895, was the first occasion on which the reducing action 

 of aluminium was demonstrated on a comparatively large 

 scale, and covered an extended series of metallic oxides. 

 Since that time great progress has been made, the most 

 noteworthy advance being in the direction of the use of 

 aluminium for the sake of the heat afforded by its combustion as 

 a true fuel, the oxygen being derived, not from the air, but from 

 a metallic oxide. In order that I may be clear, let me repeat 

 that when coal is burnt the oxygen is derived from the air. When 

 aluminium is used as a fuel the oxygen is derived from a metallic 

 oxide, the metals change places, the aluminium is oxidised, 



and the other metal set free from its oxide. This part of the 

 subject must be carefully approached, and the question at once 

 arises as to what extent the aluminium must be heated before it 

 will begin to abstract oxygen from air or from an oxide. It is 

 well known that the metal aluminium will not oxidise sensibly 

 in the air at the ordinary temperature, but the presence of a 

 little mercury enables it to oxidise readily. Le Bon (Comptes 

 rcndiis, October 29, 1900, p. 707) has .shown how minute the 

 quantity of mercury may be. This wire of aluminium to which 

 a thermo-couple is attached will, if a mere trace of mercury 

 be rubbed on its surface, become rapidly heated by oxidation, 

 the temperature rising to 102° C, while at the same time a 

 fungoid-like growth of alumina forms on its surface (see Fig. I). 

 Aluminium foil will burn readily in oxygen if its combustion 



V> O O O (/ 



1 o o o r; 

 \i o o 00? 



110 o o on 



Fig. 2.— Crucible in which the reduction of metallic oxides is effected 

 -A., diagrammatic section of the perforated sheet-iron crucible, b, lined 

 with magnesia ; c is the mixture of aluminium and the metall 

 to be reduced to metal ; « is a piece of magnesium ribbon pla< 

 mixture, 6, of aluminium and some readily reducible oxide. 



ide 



be started by a glowing fraginent of charcoal. The temperature 

 at which aluminium will abstract oxygen from a metallic oxide 

 will depend on the oxide submitted to its action. Three cases 

 may be taken : (i) Lead oxide and granulated aluminium may 

 be ignited by a match, as may also silver oxide (Ag.,0), for it 

 parts with its oxygen very readily. (2) Chromium oxide 

 (Cr^Oj) and granulated aluminium burns slowly and requires 

 rather a high temperature to start the reaction. O.^ide of iron 

 (Fe.103) and granulated aluminium also requires the presence of 

 a readily reducible oxide to start the reaction. On the other 

 hand, (3) a mixture of sodium peroxide, carbide of calcium and 

 granulatedaluminium may hz started by a drop of water by 

 the mere inflammation of the acetylene. In all these cases, 

 or in any other case, the products are solid, for if any of the 



