September 15, 1892J 



NATURE 



487 



lathe) to press upon and traverse a revolving cylindrical surface 

 on which the depasit is taking place, and while it is immersed 

 in the copper solution. The rcjult is that it is kept smooth and 

 bright to the end of the process. 



But the use of a burnisher is not the only means available for 

 the production of a smooth deposit. It was observed in the 

 early days of electro-plating how great a change was effected in 

 the character of the metal deposited by the presence of a very 

 small quantity of certain impurities. It was found, for example, 

 that an exceedingly minute dose of bisulphide of carbon, if put 

 into a bath from which the silver was being deposited, caused the 

 deposit to change from dull to bright. 



I have lately had experience of a similar kind with nickel and 

 with copper. I was working with a hot solution of nickel, and 

 up to a certain point the deposit had the usual dead-grey appear- 

 ance. Suddenly, and without doing anything more than putting 

 in a new cathode. I found the character of the deposit com- 

 pletely changed. Instead of the grey, tough, adherent deposit, 

 there was produced a brittle, specular deposit, which scaled off 

 in brilliantly shining flakes of metal. I sought for the cause of 

 this extraordinary change, and traced it to the accidental intro- 

 duction into the solution of a minute quantity of glue. 



By adding gelatine to a fresh nickel solution I obtained the 

 same peculiar bright and brittle deposit that had resulted from 

 the accident. I then made a similar addition to a solution of 

 copper, and when I hit the right quantity — an exceedingly 

 minute one— bright copper, instead of dull or crystalline, was 

 deposited. Here are some specimens. These were deposited on 

 a bright surface, and they are bright on both sides. 



Not only is the copper made bright, under the conditions I 

 have described, but, if the proportion of the gelatine be carried 

 to the utmost that is consistent with the production of a bright 

 deposit, it becomes exceedingly hard and brittle. Beyond this 

 point the deposit is partly bright and partly dead, the arrange- 

 ment of the patches of dead and bright being in some cases very 

 peculiar, and suggestive of a strong conflict of opposing forces. 



Before I leave the subject of copper deposition, I may mention 

 that I have found the range of current density within which it is 

 possible to obtain a deposit of reguline metal, far wider than is 

 commonly supposed. 



The rate of deposition in copper-refining is usually very slow, 

 and it is one of the drawbacks of the process, since slow de- 

 position necessitates large plant. But rapid deposition necessi- 

 tates a larger consumption of power, and larger cost on that 

 account, and therefore, there is a point beyond which it is not 

 good economy to go, in the direction! of more rapid deposition. 

 Still there are cases, where, if we had the power to deposit more 

 rapidly, it might be found useful to exercise it. The subject of 

 more rapid deposition is also interesting from a scientific point 

 of view, I therefore mention an unusual result I have arrived at 

 in this direction. 



Taking as one extreme, the slow rate of deposit, of one ampere 

 per square foot of cathode — a rate not infrequent in copper- 

 refining, I have found that the limit in the other direction is not 

 reached by a rate of deposit one thousand times faster. I have 

 produced, and I hope to be able to produce before you, a per- 

 fectly good deposit of copper, with a current density of looo 

 amperes per square foot of cathode. 



This cell contains a solution of copper nitrate with a small 

 proportion of ammonium chloride. The plate on which I am 

 going to produce a deposit of copper has an exposed surface of 

 21 square inches. Opposite, at a distance of one inch, is a plate 

 of copper. When I close the circuit, a current of 140 amperes 

 is passing through the solution. I continue this for just one 

 minute. Now I wash it and remove the outer edge so as to 

 detach the deposit, and as you see, I have a sheet of good copper 

 — an electrotype. 



To have produced a deposit of this thickness at the ordinary 

 rate used in electrotyping operations would have occupied more 

 than an hour. 



In this experiment an extreme degree of rapidity of deposition 

 has been shown. I do not intend to suggest such a rate of 

 practical value. But it is at least interesting, as showing that 

 the characteristic properties of copper are not less perfectly 

 developed when the atoms of metal have been piled up one on 

 the other at this extremely rapid rate than when there is slower 

 aggregation. 



I think it probable that a rate of deposit intermediate between 

 this rate and the usual one of about 10 amperes per square foot 

 may frequently be useful, for no doubt the slowness of the rate 



NO. I I 94, VOL. 46] 



of deposit has often prevented electrotyi)e from being made use 

 of where, if the rate could have been increased ten times, it 

 might have been used with advantage. 



Here are some thick plates, deposited at the rate of icx> 

 amperes per square foot. They are as solid and as free from 

 flaw as plates deposited ten times more slowly. 



I said that electrolytic copper-refining owed its existence to 

 the discovery and improvement of the dynamo, and that other 

 electro-metallurgic industries hadoriginated from the same cause. 

 One of these industries is the electrolytic production of aluminium. 

 When Deville produced aluminium by the action of sodium on 

 aluminium chloride, exaggerated expectations were entertained 

 of the great part it was about to play in metallurgy. It was very 

 soon found that aluminium had not all the virtues that its too 

 sanguine friends had claimed for it, but that it had a great many 

 most valuable properties, and, given a certain degree of cheap- 

 ness, a number of useful applications could be found for it. Some 

 of these are suggested and shown by the various articles made of 

 aluminium, kindly lent by the Metal Reduction Syndicate, and 

 metallurgical research is rapidly extending our knowledge of its 

 importance in connection with the improvement of steel castings, 

 and the production of bronzes and other alloys of extraordinary 

 strength. The cost of aluminium produced by Deville's process 

 was too great to permit of its use on any large scale for these 

 purposes. 



After Davy demonstrated, by the electrolytic extraction of 

 potassium and sodium, the power of the electric current to 

 break down the strong combination existing between the alkaline 

 metals and oxygen, it seemed natural to expect that aluminium 

 would also be reduced by the same means. But Davy did not 

 succeed in producing any appreciable quantity of aluminium by 

 the electrolytic method. Deville and Bunsen were more suc- 

 cessful, but they did not possess the modern dynamo : that has 

 made all the difference between the small experimental results 

 they achieved and the industrial production of to-day, a produc- 

 tion now so large that I suppose every day it amounts to at 

 least one ton, and has resulted in a very great reduction of the 

 price of the metal. 



There are two electrolytic processes at work. One is the 

 Hall process — ^^employed at Pittsburg, and at Patricroft, Man- 

 chester — and now in experimental operation here. The other, 

 the Herault process, worked at Neuhausen, is not greatly 

 different from the Hall process — the shape of the furnace or 

 crucible is different, and the composition of the bath yielding the 

 aluminium may be different, but in all essentials these two pro- 

 cesses are one and the same. They depend on the electrolysis 

 of a fused bath, composed of cryolite, aluminium fluoride, fluor- 

 spar, and alumina. In the Hall process this mixture is con- 

 tained in a carbon-lined iron crucible — the cathode in an 

 electric circuit ; and between which and the anode — a stick of 

 carbon immersed in the fused bath — a difference of potential of 

 10 volts is maintained. In carrying out the process on a manu- 

 facturing scale, there are many of these sticks of carbon to each- 

 bath. Here, in our experimental furnace, there is only one. 



The heat developed by the passing of so large a current as we 

 are using (180 amperes) through an electrolyte of but a few 

 inches area in cross section, is sufficient to melt and keep red- 

 hot the'fluorides in which the alumina is dissolved. 



The electrolytic action results in the separation of aluminium^ 

 from oxygen. The metal settles to the bottom of the pot, and 

 is tapped or ladled out from time to time as it accumulates. 

 The oxygen goes to the carbon cylinder, and bums it away at 

 about the same rate as that at which aluminium is produced. It 

 is only necessary to keep up the supply of alumina to enable 

 the operation to be continued for a long time. I mean, of 

 course, in addition to the keeping up of the current and the 

 supply of carbon at the anode. 



By far the greater part of the cost of aluminium obtained by 

 electrolysis is the cost of motive power : 20 horse-power hours 

 are expended to produce i pound of aluminium. Therefore it 

 is essential for the cheap production of aluminium to have cheap 

 motive power. 



There is one feature about the Neuhausen production of 

 aluminium which is very striking, and that is the generation of 

 the electric current by means of water power derived from a 

 portion of the falls of the Rhine at Schaffhausen. 



The motive for making use of water power is economy. But, 

 apart from that, it is interesting to see water replacing coal, 

 not only in the production of power, but also in the production. 

 of the heat required in a smelting furnace. 



