578

years ago will remember that most people had the idea that the great thing to be aimed at was a very large surface. Mr. Cookson's invention, therefore, seemed most valuable, and he set to work with the energy which is his characteristic, especially when working at something new. Mr. Cookson made up a dozen large cells, which he sent to Mr. Jameson, the inventor of the coke process, who was then working at an incandescent lamp. Some of the lead hair was made up into little cells, and tried in various ways by Professor Herschell, of Newcastle. Professor Herschell published some account of his work in Nature. Mr. Cookson also made arrangements with the writer to go into the question of secondary batteries. The theory that large surfaces of lead was necessary gave way when the question of local action was more fully studied, but after all it seems to be near the truth.

Some reason may be demanded for bringing before this society the account of experiments which were made so long ago. Many of them have been made again since, and many are of no use at all; but, on the other hand, there are a number which are still new, and the next best thing to knowing what to do is knowing what not to do. We made altogether more than 300 different cells; but to prevent this paper running to too great a length, only a few of them are described.

Cells with Hair Lead.

The first were Mr. Cookson's large cells. These had four plates; each plate was about 8 inches by 10 inches by 1 inch, and consisted of a sort of cage of sheet lead with holes in each side. The lead was joined by burning, no solder being used. The inside of the cage was filled with lead hair, a lead wire being arranged zigzag through it, and burnt to the cage. Each plate was sewn in flannel. A battery of 12 of these cells was sent to Mr. Jameson to be charged. They came back to the lead works without any particulars as to what had been done with them. We therefore coupled them up with a dynamo, and charged them in what appeared to be the opposite direction. They had 12 ampères through them for about seven hours a day for no less than five months, and even after three months gas was coming off the reduced plate only. The cells seem to have had enormous capacity, and must originally have been charged in the other direction. After three months the reduced plate gave off gas, but the peroxide plates went on charging for the whole five months. We did not try running them down to see what charge they really had, as other experiments which were going on at the same time led us to suppose that the local action would render them useless. Probably the surface of these plates was so large that the local action formed enough sulphate of lead during the night to utilise the current during the whole of the next day in oxidising it. At the end of the five months the flannel round the plates was not spoilt. This seems to show that it is not the acid that spoils the flannel in the Faure cells, but the oxidation due to the contact of peroxide of lead. Some experiments on organic substances with oxidising agents will be described presently.

The first difficulty to be overcome with the Cookson lead hair was making good electrical connection with the support. Various methods of soldering were tried, but the little wires of lead were melted off whenever it was tried to burn them to the back. Solder with a flux of caustic soda ran along the fibres too much. Various methods of connecting the fibres by pressure were tried. The material was woven into loose mats, and these were passed under large rollers. Every degree of density could be obtained in this way. Though the metal was clean, and appeared to stick together at first, it did not adhere as strongly as might have been expected. The chief fault of the lead hair seemed to be the vice of local action. In those days local action was not so well understood as now. The theory seems to be this: If lead and peroxide of lead are electrically connected, and any of the surface of the lead is exposed to the acid, that part of the surface is attacked, and a thin film of lead sulphate is formed which protects the surface from further corrosion. The lead is then in contact with the peroxide at some points, and with lead sulphate at other points, but is not touched by the acid anywhere. If the cell is charged further, the film of lead sulphate is oxidised into peroxide, so that the plate again consists of lead and peroxide of lead only. As the peroxide formed from a given quantity of lead sulphate occupies much less room, a small portion of the surface of the plate is again exposed, but is immediately coated with a film of peroxide; the result is that the whole surface of the plate is protected. The least movement seems to expose some of the surface of the lead to the acid. Mr. Crompton told me this summer that he found that cells were practically indestructible if not discharged too much, and this seems to be the grand secret of successful work with secondary batteries. Messrs. Drake and Gorham mentioned the same thing in their paper read at the British Association this autumn. Either the peroxide of lead is so porous that it never quite protects the plate, or it is disturbed by small bubbles formed in the coating; for if a current is passed continually in one direction a lead plate is eaten into, as may be seen in a water resistance with lead electrodes-the oxidised plate seemingly does go on forming. In the lead hair cells the surface exposed was so enormous that a very slight formation was enough to demand a very large supply of electrical energy, so that the cells seemed to run down as fast as they were charged. No doubt the fault was that the fibres of lead used were rather too fine; if we had used much thicker lead wire we should have probably made a very good battery. Several plates were tried with hair lead, red lead, litharge, &c., but our ideas as to the validity of Faure's patent were not quite what they are now, so comparatively little was done in that direction,

[DECEMBER 10, 1886.

Secondary cells are, however, tedious things to experiment upon; it may take a year or so to see what a cell is going to do. The best method is to keep several sets of experiments going on at the same time.

For the sake of conciseness some of each class will be given, arranged in groups.

Quick-forming.

Cells were made with lead hair in a solution of sodium chloride: the anode was quickly attacked, but the coating was not adherent, and came off in clouds; black clouds, presumably of finely-divided lead, came off the other plate.

Lead hair and potassium bichromate: seemed to oxidise well. Lead hair and potassium sulphate: no very marked result. Amalgamated lead in dilute sulphuric acid: the peroxide formed had little stiction or power of adherence, and kept peeling off and exposing the bright amalgamated surface; this happened especially on reducing.

The object aimed at was to get an electrolyte which partly dissolved the plate; the idea being that an action like that of which advantage is taken in the manufacture of white lead might be used in the quick formation of battery plates.

Lead in solution of sodium-thiosulphate: thick black mud of lead sulphide formed.

Amalgams of lead and 10 per eent. and 5 per cent. of mercury : coating had no stiction.

Alloy of lead and zinc: this was not so easily attacked as one might expect.

Lead hair in solution sodium hydrate and sulphide: the solution got yellow, and deposited sulphur on the oxidised plate.

Lead in solution of ammonium chloride: was no better than the similar cell with sodium chloride. No doubt a little nitric acid was formed, but the cell was not a success.

Cells were tried with solutions of potassium fluoride the plates were not formed well.

Lead in potassium sulphides: not much attacked; the action of the sulphide seemed uncertain, and is partly dependent on the amount of hydrate present.

Lead plates in solution of salt, with felt between : after reversal this gave good spongy lead, but a great deal of the peroxide was loose and detached.

Lead with various mixtures of sulphides and carbonates and nitrates: none of these gave deposits which would reduce wellthey were all too loose.

Lead in sulphuric acid: just diluted enough to make it conduct; the anode was not much attacked.

Lead in a mixture of sulphuric and nitric acids: more quickly formed.

Lead in nitric acid alone: eaten through very quickly. The action of nitric acid on a lead plate is very peculiar. Most of the solutions already mentioned attacked the plate evenly and oxidised it gradually through, but nitric acid makes pits and round holes. As nitric acid seemed by far the best means of quick formation, several experiments were made with it. Plates were eaten wholly through with acid which was too strong to dissolve much nitrate; the acid was then siphoned off and replaced by a solution of sodium sulphate and reversed. There was difficulty in reducing the lead sulphate; it formed cakes which did not make good contact with the plate. The great objection to the various methods of forming above described is that the coatings produced have too little stiction to reduce well. When lead was the anode in nitric acid no peroxide was produced. Of course no peroxide is formed when the plate is in contact with sulphuric and hydrochloric acids, as peroxide of lead liberates chlorine in an acid solution, while chlorine precipitates lead peroxide from an alkaline solntion of the lead salt.

The best solution we tried for forming plates was a mixture of dilute sulphuric and acetic acids: this solution made a fine and adherent coating which reduced well. One of the most obvious disadvantages of such solutions as dilute nitric acid is the difficulty of getting rid of the least traces of acid; of course, traces of such things as nitric or hydrochloric acid in a cell may be very destructive, or they may be quite harmless. Acetic acid can easily be driven off as vapour, and nitric acid might perhaps be reduced to ammonia and then volatilised.

Mr. Brush heats his plates to accelerate their formation; we did not try this.

Local Action.

To reduce the local action in the peroxide element we tried amalgamating the plates: this did not act. We tried to gild lead plates, but did not succeed; a gold sleeve-link seemed to act well as a support for a little peroxide plate, and it was hoped that gilding might preserve lead. We tried to protect lead by various sulphides. These experiments seem absurd; in fact, many of our experiments were rather wild; but the supposed method of protecting plates has recently been patented by no less able a chemist than the late Mr. Tribe. Lead was fused with various proportions of galena, and the resulting substances were tried as anodes. Galena itself is at once coated with lead sulphate when used as anode in dilute sulphuric acid.

Lead was also combined with arsenic by heating in glass tubes: this was very troublesome work, as the arsenic was apt to volatilise, and sometimes the tubes broke, and the laboratory had to be abandoned until the smell was gone.

Double sulphides of antimony and sulphur bases were tried, but were of no use. In our provisional protection applied for in 1883 we mention coating the plates with lead sulphide by means

DECEMBER 10, 1886.]

of thiocarbamide. This idea was taken from a receipt seen in a copy of the Plumbers' Times and Paperhangers' Gazette, or some such paper, which was sent me with an intimation that I had no business not to take in my trade journal. Thiocarbamide could not be bought, and was too troublesome to make, so this plan was not tried. The process is said to produce a beautiful polished adherent coating of lead sulphide on lead, or even on organic substances such as netting, and is of very considerable interest, and perhaps of commercial value.

Silver was tried as peroxide plate in various solutions. It behaved very much like lead, though it is not so easily attacked by sulphuric acid; peroxide of lead sticks to it well, but of course it is too expensive for commercial use. Silver might be used for connections. Most metals exposed to the action of the spray from secondary batteries are attacked, and make bad contacts, but silver may be laid across the top of a cell without tarnishing. Probably electro-plated copper would do as well. Very few experiments were tried with platinum because of the expense. Mr. Brush has patented plates made of an alloy of lead and platinum. Most likely it would take a very large proportion of platinum to protect a plate, as the alloys of platinum are easily oxidisable. No doubt Mr. Brush's invention is only waiting for the discovery of a platinum mine.

Local action may be a very serious evil in reduced plates, especially with strong acid. In some cases reduced plates in dilute sulphuric acid give off little bubbles of gas continuously for a couple of weeks. Spongy lead seems to have very nearly strong enough affinity for the electro-negative radicle of sulphuric acid to combine by evolving hydrogen, more especially if the acid is strong. It seems likely that traces of some metal even slightly negative to lead may start local action. Platinum and lead, for instance, decompose dilute acid at once; but perhaps traces of arsenic or antimony in the oxides from which the reduced coating has been made, or in the acid, may sometimes produce the same result. Peroxide of lead, on the other hand, is more stable, though when heated with strong sulphuric acid it seems to evolve some oxygen and turn lighter in colour.

Electro-gilded copper was attacked at once by sulphuric acid.

Carbon for Peroxide Plates.

The practice of making peroxide plates of carbon does not seem to be quite abandoned even yet. Carbon is more easily oxidised than is popularly supposed; in fact, Brodie's process for disintegrating graphite depends on the action of an oxidising agent. We tried some Faure plates with carbon in dilute sulphuric acid and in a solution of zinc sulphate. In both cases the carbon was attacked. After standing for some days the peroxide plates seemed quite good, and the peroxide appeared to stick well to the carbon and to form good coatings. On examination, however, it was found that the carbon under the peroxide was converted into a soft black pulp. Carbon was tried as anode in various solutions, such as sodium carbonate, sodium phosphate, phosphoric acid, dilute sulphuric acid, potassium fluoride, potassium sulphide, hydrochloric acid, potassium chromate, ferri-cyanide, ferro-cyanide, permanganate, borax, potassium tungstate, potassium hydrate, sodium chloride, &c. In all cases where oxygen would otherwise have to be evolved the carbon was attacked, but when a current could pass by evolving chlorine the carbon was not attacked. As peroxide of lead acts on a chloride in an acid solution, carbon is evidently useless for peroxide plates. It may do admirably for bleaching or evolving chlorine or making hypochlorites. The same conclusions were arrived at by Bartoli. Biscuit, or unglazed porcelain, impregnated with carbon deposited from hydrocarbon by heat, was tried: it split into small pieces. It is odd that none of the sufferers from primary battery disease have used this material for porous pots, as many of them have the idea that a carbon porous pot somehow reduces the resistance of the cell.

As no metals were available as indestructible supports for the oxidised plate in dilute sulphuric acid, and as carbon also appeared useless, the next thing was to try to make the supports of peroxide itself. Peroxide made by oxidising coatings made originally from litharge or red lead was not hard enough to make good supports; but the peroxide deposited by electrolysis of an alkaline solution of a lead salt was very much harder. The best way seemed to be to make a paste of litharge and caustic soda, and to attach it to the lead plate and make it the anode in a solution of caustic soda. When the whole of the litharge was oxidised the lead plate was removed. The resulting peroxide was very hard, and seemed quite impervious; when a bit was used as a plate in a cell, and the cell discharged, it ran down at once, the peroxide being covered by a very thin film of lead sulphate. The surface seemed to be the only part in contact with the acid. It would be a matter of great practical difficulty to make cells with peroxide supports commercially. It would be interesting to know how this form of peroxide compares with Mr. Fitzgerald's lithanode in point of hardness and durability. The peroxide from a solution of nitrate and tartrate of lead looks very dense and hard, but it comes in the form of scales resembling iodine. Before leaving the subject of lead plates some experiments on lead sulphate must be described. It was supposed at one time that lead sulphate could not be reduced in dilute acid. Sir William Thompson had made experiments with it, and had found it would not reduce. We made a cell with platinum plates and precipitated lead sulphate in dilute sulphuric acid; it was charged for five weeks, and at the end of that time the sulphate on one was completely oxidised, but on the other it was only reduced here and there in spots. A similar cell was then started

579

with a little litharge mixed with the sulphate: the sulphate then reduced perfectly. This is a result of some importance. The reduction of sulphate seems to be purely a question of good contact. In many cells which gave very small outputs the reduced plates were found to be in fault. The slightest film of thoroughly-formed sulphate acts as a perfect insulator. It is a good plan to keep a small cell always ready with both plates fully charged; then, when a cell under test runs down, each plate is separately tried with a proof plate, and it is then found which of the plates has failed. The film of sulphate formed on a piece of lead reduces all right, but as sulphate of lead is very bulky its formation seems to cause buckling about and resulting bad contacts. To see how great the swelling due to the formation of sulphate of lead is, the specific gravities were taken, and the following are the resulting figures :

100 volumes of lead form

Peroxide of lead
Red Lead
Litharge

Lead sulphate

...

...

...

This shows that lead converted into sulphate swells into three times its original bulk, and that peroxide on becoming sulphate is doubled in volume; this at once shows why coatings tumble off plates. To see how much of the active material in a cell is really used, several plates were made up with weighed quantities of lead oxides. One of these was a little Sellon plate, made of a piece of a large plate filled with minium or litharge; others were Faure plates. After charging, the plates were short-circuited through copper voltameters. The maximum output was from 5 to 7 per cent. of that calculated.

Many attempts were made to make coatings that would stick well. Collodion was of no use; litharge and sugar did not make good coating; sodium silicate mixed with the litharge before applying was useless. Litharge and glycerine form a very useful cement known as engineers' cement. Though the coating was very easy to manage, there seemed to be no permanent advantage in using glycerine. This coating has been patented. In those days it was usual to wrap plates in flannel to keep the coating on. The flannel generally went to pieces very soon; it appeared, however, that this is not due to the action of the acid, but to oxidation through contact with the peroxide. Asbestos paper is too weak to be of any use. Cotton or linen is soon destroyed by acid, but resist strong alkali longer; while flannel resists acid, but is attacked by alkali. Pyroxylin is easily attacked by alkali, not easily by acid. Tribe proposed to use pyroxylin for supporting plates. Xylonite has also been tried for me more recently by Mr. Cathcart, then my assistant. It is much the same as celluloid, being made of pyroxylin and nitro-benzole, camphor, or same such solvent. This will stand being boiled in moderately strong sulphuric acid for a long time, and should be very useful in battery work. Flannel was soaked in collodion, with the idea that the fibres might thus be protected, but they were not. Flannel was also soaked in sodium silicate and attacked with sulphuric acid : it succumbed.

Strength of Acid.

It was usual to use acid of a strength of 1 to 10. This probably arises from that being the strength in vogue for Daniell batteries. The strength of acid was also not considered, because the action of secondary batteries was for some unaccountable reason supposed to be due to the formation of litharge on both plates. I am afraid Dr. Lodge was much to blame for perpetuating this error. Another absurd theory was that the action is due to occluded hydrogen in the reduced plate. It has been said that the study of secondary batteries belongs to chemists, not to electricians; but it is strange that when chemists have to deal with electrolysis they seem to forget their chemistry, or to imagine that the laws of chemistry no longer hold good. As the action of the cell depends on the formation of lead sulphate, it is clearly an advantage to have as much acid as possible available in the coatings. If the coatings have only weak acid, it may be all absorbed when the cell is discharged a little, and the electromotive force will fall, and there will be a tendency to the formation of basic salts in the coatings; but, on the other hand, if the acid is made stronger, it may be decomposed without electrolysis, as when peroxide of lead evolves oxygen and forms lead sulphate, or as when spongy lead evolves hydrogen and forms lead sulphate. In the case of spongy lead there may or may not be electrolysis. If the acid is so strong that the lead decomposes it directly, there is no local action; but if there be some more electro-negative metal, such as, say, antimony, in contact with it to evolve the hydrogen, there is electrolysis. The best strength of acid was not determined by us, and it no doubt depends on the purity of the lead, and of all the compounds forming the coatings, and of the acid itself. It also depends on the way the cells are to be used. If the acid is strong, both oxygen and hydrogen will be given off slowly, even during several weeks. As the idea of using secondary batteries as magazines for storing energy for long periods is being given up, this slow decomposition is of little importance; but there is another point-that is, that sulphuric acid which is not very dilute separates into two strata of different specific gravities. Before leaving the subject of lead cells, it may be as well to mention that the troublesome spray that comes during charging may be prevented by floating paraffin on the top of the acid. The spray not only tarnishes, rusts, or destroys everything near, but is the cause of all sorts of ground leaks. The paraffin prevents the acid from creeping; in fact, it creeps itself, and acts as an insulator.

580

Alkaline Cells.

It is well known that iron is very electro-negative in an alkaline solution; but such solutions as that of caustic soda or potash are not available, because nearly all salts of lead are soluble in them. It was therefore attempted to make cells with iron supports in an alkaline solution of some salt whose acid radicle would not attack iron, and would precipitate lead salts. Iron in solutions of potassium cyanide, ferro-cyanide, and ferri-cyanide was attacked, the ferro-cyanide or yellow prussiate of potash plate was least eaten. Even such salts as sodium sulphate dissolve some lead sulphate. This property is important in connection with ordinary secondary batteries. A little sodium sulphate mixed with the acid may tend to make the sulphate of lead more easily reduced; as, when all the sulphate is reduced that is in contact with the support, hydrogen is evolved, and caustic soda is formed in the immediate neighbourhood of the plate, and can dissolve the sulphate, the lead being then precipitated as spongy lead. The spongy lead can thus make contact with the sulphate. Magnesium sulphate has been recommended as a solution, but it is difficult to see why. Zinc sulphate might be good.

Alkaline sulphides of course attack iron.

Iron plates with coatings of white lead in solutions of sodium carbonate; these oxidised and reduced very well, but the iron seemed to be slightly attacked. On testing the peroxide, traces of iron were found. Thinking the corrosion of the plate was perhaps due to impurities in the carbonate, such as sulphate, experiments were tried with the purified salts. The iron seemed to be in a sort of unstable state. Sometimes it would remain quite clear and bright in contact with the peroxide, and with bubbles of oxygen evolving from its surface; but suddenly a green spot would appear, due no doubt to the formation of carbonate of iron, and this spot would extend and cover the whole plate. The best results were got with bicarbonate of soda.

Nickel was tried in the same solutions: it behaved in the same way, though it was not quite so easily attacked. The carbonate was lighter in colour, being apple green.

Iron plates with borax solution: green discoloration. Iron plate with caustic, potash, and tannin: iron eaten. Some cells were tried with iron plates and ammonia; but as it is probable that nitric acid would soon be formed in them they were abandoned. The alkaline cells gave 17 volt. The measurements were taken roughly. A calibrated tangent galvanometer was used as a voltmeter; for measuring currents a low resistance was made up consisting of a lot of wires put in parallel, their resistance having been measured in series.

Ammonia sometimes gave a brown deposit with iron plates. As it seemed hopeless to get iron to stand without being attacked, various methods of protecting it were tried. Iron gauze was coated with peroxide by making it the anode in a solution of litharge in caustic. When this solution is electrolised clouds of litharge are evolved close to the plate. This is very curious, for if the solution is plumbite of lead the electro-negative radicle would be Pb Og, not Pb 0: such a salt being presumably Na, Pb O2. Peroxide of lead deposited from solution of litharge in caustic is too dense to be used as an active coating; it was proposed to use it to protect the iron, and to have then applied a coat of porous peroxide such as is used in acid cells. The peroxide was never satisfactory. To make up for the imperviousness of the peroxide, red lead and water were rubbed into the pile of velvet; the velvet was placed with its face against the iron plate, and the coating was oxidised in an alkaline solution. The velvet was then dissolved out by strong alkali, so as to leave a coating with a very large surface. This plate, however, had very small capacity. Electro-gilding the iron was then resorted to. It is a little difficult to gild iron, as the gold does not stick well. After each coat of gold the iron was heated to make the gold sink in. The gilded plates did not answer well. The Bower-Barff process for protecting iron from rust was tried, by getting an ornamental railing top and making it the anode in a cell. The black oxide did not prove to be a good enough protection. The strange thing about the corrosion of iron in alkaline solutions is that it does not take place if there is no coating. A piece of sheet iron may be used as anode in the solution of carbonate or bicarbonate of soda without being the least corroded; but if a coating of, say, white lead is applied, green spots appear at once, but only under the white lead; the back of the plate remains quite clear. Oddly enough, sheet tin is not easily attacked by an alkaline solution. Sometimes it turns black, or the surface, which consists of an alloy composed chiefly of tin, peels off in black scales. Sheet tin behaved much as iron and nickel; that is to say, it behaved very well till a coating was put on, and then it was corroded. There is a very curious alloy of iron and tin. If iron and tin are melted together, the resulting alloy separates into two portions; one is approximately Fe Sn, and the other consists chiefly of tin, and has a low melting point. The alloy Fe Sn solidifies at a red heat. It was too unmanageable to be of much use for batteries, and it was no better than sheet tin; but it was so hard that a straw-hardened cold chisel even would not cut it, but broke in pieces, while a softer chisel turned up at the edge. This alloy might be useful in commerce. The writer has since tried to make more of this alloy for some experiments on persistent magnetism, but has never succeeded in making it so hard again. Iron without a coating is not attacked in such solutions as potassium bichromate, or permanganate. At first sight this seems strange, but it must be remembered that such salts owe their corrosive properties, especially when acting on organic bodies, not to the strength of their acid radicles, but to the electro-positiveness of chromium and manganese, which makes

[DECEMBER 10, 1886.

them combine energetically with an acid radicle to form chromium or manganese salts. Thus potassium bichromate or chromic acid in the presence of sulphuric acid is a very powerful oxidising agent, but if there is no acid present to form a chromium salt it is no longer energetic. Iron in alkaline solutions is very convenient for adjustable water resistances.

We worked at iron cells with some tenacity, because we seemed to be very near sucesss; and if we could make a peroxide plate with iron without local action, the negative might be of zinc, as this combination gives a high electromotive force.

Cells with Zinc.

Evidently cells with zinc could not be used in most acid solutions, as so many of its salts are soluble. Secondary batteries with zinc plates have often been brought out; but as the zinc dissolved on discharging is not deposited on the same part of the plate on charging, these cells cannot well be used very often. M. Reynier has lately worked at cells with zinc in sulphuric acid, and Planté peroxide plates. He does not say how he proposes to get over the difficulty of the zinc being deposited in the wrong place. He says that the deposited zinc is so pure that there is very little local action. His cells give 2:37 volts.

We tried plates with reduced zinc coatings in alkaline solutions that do not dissolve zinc, such as sodium carbonate and sodium bicarbonate. The coatings were made of oxide or carbonate of zinc. The chief trouble was the proper reduction of the zinc. These salts of zinc behave very much like lead sulphate-that is to say, they are very bulky-and the coatings form hard white cakes which are not in good contact with the support. When the zinc is eventually reduced it is very soft and loose. Very likely we made a mistake in reducing coatings of oxide or carbonate of zinc, instead of forming the coating out of the solid plate by attacking it when used as anode. The object aimed at was the use of iron for the support of lead peroxide in the other element; but as we did not get iron to work satisfactorily, we did not do very much with the zinc.

Primary Cells.

A few experiments were made, but they are not worth recording. Lately, however, the writer saw a note in the Electrician to the effect that a French savant had used agar-agar jelly in a Leclanché cell. It seemed possible that the use of this might avoid the necessity for a porous pot in a two-fluid cell, the idea being that to make up a cell one might take a zinc plate, place on it a wad of zinc sulphate jelly, place on that a wad of copper sulphate jelly, and cover it with a copper plate. Zinc and copper jellies are easily made, but the drawback is the rapid diffusion of the copper salt. Some cells were made with zinc filings suspended in the zinc jelly, so that any wandering copper salt might be caught before it reached the zinc plate. As rapid diffusion is always the result of a low-resistance porous cell, it seems useless to try to make a low-resistance two-fluid cell unless some colloïd oxidising agent can be found. Some of the solutions of iron were found by Graham to be colloïdal; perhaps something might be done with them.

A bichromate jelly cell was tried, but it was also a failure. The same thing was noticed as in the case of flannel and cotton. Bichromate jelly with no acid keeps for any length of time; and sulphuric acid jelly can also be made; but the jelly dissolves at once if both acid and bichromate are present. The same thing happens with potassium permanganate.

Agar-agar is a fucus or sea-weed. It has, if the writer is not mistaken, the same composition as starch. It has a very high gelatinising power-about ten times that of gelatine. It is probably used for sizing paper or calico, or for cooking. Mr. Stanford, of Glasgow, is an authority on the uses of sea-weeds.

Gelatine was tried, but it was liable to go bad, and was at once attacked by oxidising agents.

Probably gelatinous silica may do, but the writer has not yet made enough experiments to say. A low-resistance bichromate cell in which the bichromate had not access to the zinc would be very useful where a large power is occasionally required, as in the laboratory.

Miscellaneous.

While various cells were being charged or discharged there were often odd quarters of hours to spare. These were spent in trying different experiments not connected directly with secondary cell.

White Lead.

It was attempted to make white lead electrolytically. The present method of making white lead, which is felicitously termed the "Dutch process," is slow and barbarous, and exceedingly unhealthy. The women who are continually among the white lead absorb the lead through their skins, or perhaps swallow it little by little, and are apt to suffer from lead poisoning. Messrs. Cookson & Co. have probably the best sanitary regulations and arrangements in the world; but even then the Dutch process is very dangerous.

To make white lead electrolytically, plates of lead were made anodes in solutions of carbonate or bicarbonate of soda, with very small currents. In many cases white deposits were obtained, but in no case was any quantity obtained. If the current was increased the plates were soon coated with a brown compound. To imitate the action of the basic acetate a little acetic acid was added to some of the solutions. In some cases a very curious phenomenon was seen. Long white columns like small candle drippings started about the middle of the plate, and grew down to the

bottom of the beaker. The plates were about three inches from the bottom.

Experiments were tried with lead plates and solutions of salt with a view of making white oxychloride of lead. The plates were of course easily attacked, but if the current was very small the coating formed very slowly, and appeared crystalline, and if the current was increased the coating turned brown.

Sodium Chlorate.

Potassium chlorate is very largely used for making English matches, and in dyeing; it is probably used instead of sodium chlorate because it can easily be separated from chloride by crystallisation. To make sodium chlorate, solution of salt was electrolysed. This gives off chlorine at the anode, and various salts, such as sodium hypochlorite, are formed. If sodium hypochlorite is boiled it gives chlorate and chloride. We prepared a mixture of sodium chloride and chlorate, but the percentage of chlorate was too small for the process to be valuable.

Aniline Dyes.

There may be a large field for the application of electrolysis in the aniline industry.

A cell was tried with aniline, but it would not conduct.

A cell was tried with phenol, but it would not conduct. Potassium hydrate was therefore dissolved in the phenol to make it conduct. A black compound was formed at the anode.

These last experiments are not brought forward as of any value, but they are instances of applications of the dynamo to very different industries. Now that we have the dynamo as a commercial machine, we seem to be resting content with one very small use for it. We use it almost entirely for the electric light. Though it is quite well understood that power can easily be transmitted electrically, the use of motors can hardly be said to be pushed at all. Engineers have only to be shown motors at work to see what beautiful, smooth, and silent-running things they are, and they will be applied in thousands of instances. The dynamo is now being introduced for the treatment of argentiferous copper; but this is a comparatively small industry. The electro-plating fraternity are a very peculiar people. Though constantly using it they seem to know as little about electricity as medical men. Before electricians can meet them on common ground they must give up talking of the strength of the quantity of current from seventeen Smee cells.

But above all these uses the electrical engineer can supply the chemist with a most efficient and convenient oxidising or reducing agent; unfortunately, however, the electrical engineer does not know what the chemist wants, and the chemist does not know what the electrician can give him.

Every week more young men begin studying to qualify as electrical engineers. Like pins, it is unknown what becomes of them. Would it not pay some of those who are sharp and inventive to take up the study of electro-metallurgy systematically, and to explore that almost untrodden field of the application of the dynamo?

The PRESIDENT said it was too late to commence the discussion on Mr. Swinburne's paper that evening. The next being the annual meeting, would take place on December 9th. The meeting then adjourned.

CORRESPONDENCE.

The Management of Accumulators.

I feel as grateful to you for pointing out the failings of my pamphlet as for your praise. It is right for me to point out one or two instances of error into which you have fallen, some, it is true, from no fault of your own. The copy sent for your perusal was an early one, and without the preface, which counteracts the bluntness of the "Introduction." The primitive diagram will eventually give place to a proper one, and a photograph of the accumulator-house described, when the price will be one shilling. These are not yet ready, and publication would have been delayed, but for some special reasons with which I need not trouble

you.

The words used, "formed lead plates," are generally understood the right way as "unpasted ones." I assumed that there would be two classes of readers, one well acquainted with cells, and one who only attend them, so to make every point clear to both would entail a large volume, which would rarely be referred to. This accounts for the reason why the manipulative details only are most explicit. You ask why do I not give the cause for bad sulphating? It seems to me that it is already there, by stating that the cells should never be very much discharged. In future copies an appendix is being added, and should the public honour me by demanding a second edition, much new matter will be added, for each day new

facts are found, and other chemical theories will be given. You appear also to have fallen into a misunderstanding regarding what is stated about the possibility of an indefinitely prolonged charging. If the current is continually made smaller as the material on the plates is reduced, the series is an infinite one, and therefore saturation is never reached. This is only looking at the question from a mathematical point of view. I know well many errors have been made which in time I hope to correct, yet the main object in view remains intact, namely, my firm belief that the rules given will much assist to keeping an accumulator in good order over many years.

December 6th, 1886.

David Salomons.

[In reference to the first portion of Sir David. Salomons' letter, we have only to say that the pamphlet which he now states to have been incomplete came to us marked "For review if desired," in Sir David's own handwriting.-EDS. ELEC. REV.]

Clark's Standard Cell.

In your promised article on Clark's Standard Cell, can you kindly give the proper proportions for the ingredients of the "paste" which receives the zinc ?

There are no quantities given in any of the published descriptions I have met with for the zinc sulphate and mercurous sulphate; and it is, perhaps, reasonable to assume that the E.M.F. of the cell may depend to some extent on the respective proportions, just as the s.g. of the solutions for the Daniell cell influences the E.M.F. Dr. Fleming, in his lectures to electrical artisans, prescribes the addition of zinc carbonate; but, like other writers, gives no proportion for any of the ingredients.

Ajax.

[Probably the writer of the article will, if necessary, reply to "Ajax" in our next.-EDS. ELEC. REV.]

Tramway Gradients.

Some time since you asked for information as to places where tramways are worked over gradients so great as 1 in 17. I recently had an opportunity of examining the ordnance survey maps of Plymouth and Devonport, and find that on the tramways between these towns the well-known Devonport Hill has a gradient of 1 in 14, the hill being at the start 1 in 50, and in the middle 1 in 20. Four horses are habitually used to take the cars up this hill.

On another line laid in Plymouth and intended to be worked by steam, but now dormant, the company being in a state of transition, has two hills-one varies from 1 in 38 to 1 in 16, and the other from 1 in 31 to 1 in 9.4. Your last week's article reminds me that this may now be interesting.

December 7th, 1886.

F. Tremain, A.S.T.E.

"A Wild Scheme."

In your issue of the December 3rd you reproduce an article from the Financial News on the subject of the proposed Atlantic cable via Lisbon and the Azores. In a foot-note appended to this article you justly point out that the difficulties connected with this scheme arise from a financial point of view, and not through physical impediments.

In support of your argument I would call attention to the cables laid to and between the islands forming the Canary, and the Cape Verde groups. All these islands are of the most pronounced volcanic character, yet the cables referred to have not only been successfully laid, but have lasted remarkably well. I cite these islands as instances, for they are directly comparable with the formation of the Azores, and with the characteristics of the surrounding ocean bed.

The author of the article in question draws a terrific picture of the bottom in the neighbourhood of the

Azores precipices from hundreds to thousands of fathoms in depth, series of huge sharp-pointed rocks, submarine valleys and ridges, volcanic mountains, and submarine mud and other volcanoes, do not suggest the best bottom for a cable to rest on; but these difficulties have been encountered before, and have been successfully overcome. In the case in point they have been grossly exaggerated.

The author appears to suggest that the cable must be laid so as to hang suspended from ridge to ridge (as if the mountain tops were to be made use of as submarine telegraph poles), and his remark that "scientific men and experienced naval officers" have come to the conclusion that no cable can be manufactured strong enough to bear the strain of such suspension, should be of considerable interest to submarine cable engineers, introducing a novel feature in their work, for their aim has been hitherto, I believe, to carefully avoid any such operation as is here suggested.

A very great interest in submarine telegraphy must be my excuse for encroaching upon your space to such a length.

X. X.

Administration and Working of the British Telegraph

System.

I am induced by your mention of the fact that 1887 is the jubilee year of the invention of the practical electric telegraph, to ask if the time has not arrived for a searching enquiry into the failure of the postal telegraph department, both financially and from an administrative point of view.

As to the financial question I have only to draw attention to the long refusal to the public of the sixpenny telegram on the false assumption that it would not pay, and to the eccentricities of finance by which the expenses of the postal and telegraph branches have been so proportional as to make the latter appear a paying concern, whilst the former is in a supposed state of bankruptcy.

In the matter of administration, the chronic state of discontent mentioned in the resolution of the clerks at Liverpool would be enough, even without the glaring want of efficiency as to errors and delays, which you have frequently drawn attention to, as regards the metropolis. The administrative failure is self-evident to the public, but not more so than the financial juggling, if they care to examine the estimates. For many years it was manifest from these that the proportion of expense charged to the telegraph department, as compared with the postal branch (mostly two-third telegraphs one-third postal) was unjust, as the staff and salaries opposite these proportions would at once show. This in fact is admitted, for in recent years the proportions have assumed a more varied character, such as eleven-sixteenths and five-sixteenths, and so on. A matter not so well known to the public is the manner in which that pet idea, the parcels post, is nursed, all parcels sent by post for the telegraph. branch being stamped, and therefore paid for; whilst telegrams for the parcels post are transmitted free of charge.

It is easily understood that part of postmasters' salaries should be debited to the telegraph account; but why their chief clerks who in the presence of a telegraph superintendent do no telegraph work should be partly paid from that department's funds is somewhat more mysterious.

Again, where this proportionate arrangement obtains and mails are landed or dispatched, the postal officer, who does the duty, receives half a guinea or a guinea allowance, and a due proportion of this is religiously charged to the unfortunate telegraph account.

Even more ludicrous instances have been published of this convenient proportionate finance.

Another thoroughly unpractical thing was that admitted failure, the amalgamation of the staffs in large offices, and this administrative blunder has a good deal of inefficiency to answer for.

It is not necessary to remark on this matter of

[DECEMBER 10, 1886.

inefficiency further than to indicate what in the writer's opinion is the cause thereof. It is believed to be largely due to the administration of the telegraphs by postal officials, devoid of all telegraphic knowledge and the inventors of unworkable rules. These, too, have been administered with Spartan severity, until the staff are drilled into a rigid adherence to the letters of them, instead of observing their spirit; and are thus encouraged to make no use of their intelligence.

To the employment at branch offices of clerks, who are led to consider their telegraph work of quite minor importance to their postal duties, forgetting that they are, in telegraphing, admitted as it were into the merchant's office, whilst in dealing with sealed communications, they may be looked upon as in his messenger's lobby.

To the supervision being performed by old and valued servants, many of whom, however, whilst well adapted to clerical work, are unfitted for superintending the work on sounders they cannot read, and apparatus beyond their comprehension. They are men of the old school, and have done good and faithful service for 20 to 30 years, but are not judiciously employed.

I have said that the use of one's intelligence is not encouraged. I say more, that knowledge is absolutely at a discount, for if you know a little, you soon find that in the opinion of your superiors you know too much. I have been for many years a student of technology, have the highest honours in my own subjects, and some of the associated ones, and have no complaint myself to make of want of recognition, but I do complain, and believe it to be a serious defect, that in these days of technology little or no encouragement is given to its study in the telegraph service of this country. No work is well done, or system perfected, unless those engaged in its evolution are enthusiasts and diligent. I therefore submit that for the telegraph department to be a success financially, and to be administered in the best interests of the public, those who control it must be telegraph men with an enthusiasm for their vocation and possessed of a keen desire to engage in friendly rivalry with those who control the kindred departments. Such an one should be the secretary for telegraphs, and for his immediate staff he should have those who could qualify for the duties from his own department. Men intimate with every detail of telegraph routine, although not perhaps the rapid manipulators they once were, or acquainted with the elaborations of modern science. Allowances should be made for testing duties, as an encouragement to those who acquire technical skill. Two or three hundred allowances of, say five shillings a week, would have a wonderfully energising effect and raise the intelligence of the staff. The latter's status should be elevated as much as possible, and an esprit d'corps, and sense of personal honour encouraged in the accuracy, speed, and general success of their department.

At present absolutely the only encouragement to scientific study and self-culture in the telegraph department, is the prospect of appointment in the engineering branch, and to the credit of the universally respected heads of that department, be it said that they appear to be possessed with only one desire, and that to select the best from the almost bewildering number of applicants they get for a very few appointments, perhaps on an average two or three a year.

Let such men control the telegraph system of this country, surrounding themselves with telegraph men, and working on commercial principles, and I, as one having studied the question from inside the department for 10 years, have no fear for the result, either as to the efficiency of services rendered the public, or the contentment and good work of the employees.

Since writing the above I hear that evidence is to be submitted to the Royal Commission on the Civil Service in writing only and through heads of departments. Such a method will lead to a fiasco, as one of the pretty ways of departmental chiefs is the indulgence in reprisals.

Telegraphist.