Sec. 122. Other Alloys.

The following tables, taken from the work of Feussner and St. Lindeck, Zeitschrift fuer Instrumenten Kunde, 1889, vol. ix. p. 233, together with the following notes, will suffice.

Sec. 123. Nickelin.

This is only German silver with a little less zinc, a little more nickel, and traces of cobalt and manganese. It behaves like German silver, but is an improvement on the latter in that all the faults of German silver appear upon a reduced scale in nickelin.

Sec. 124. Patent Nickel.

Practically a copper nickel alloy, used to some extent by Siemens and Halske. It stands pretty well in the same relation to nickelin as the latter does to German silver. After annealing as for manganin it can be made into serviceable standards which do not change more than a few thousandths per cent. I have not come across a statement of its thermo-voltage against copper.

Sec. 125. Constantin.

Another nickel copper alloy containing 50 per cent of each constituent. It appears to be a serviceable substance, having a temperature coefficient of 0.003 per cent per degree only, but an exceedingly high thermo-voltage, viz. 40 micro-volts per degree against copper.

1 2 3 4 5 6 7 8 German Nickelin made Rheo- Patent Nickel Manga- Nickel Silver by Obermaier tane nese Manga- Dia- Dia- Dia- Dia- Copper nese meter meter meter meter Copper 1.0mm 0.1mm 0.6mm 1.0mm

Copper 60.16 61.63 54.57 53.28 74.41 74.71 70 73

Zinc 25.37 19.67 20.44 16.89 0.23 0.52 ... ...

Tin ... ... ... ... trace ... ...

Nickel 14.03 18.46 24.48 25.31 25.10 24.14 ... 3

Iron 0.30 0.24 0.64 4.46 0.42 0.70 ... ...

Cobalt trace 0.19 ... ... trace trace ... ...

Mang- trace 0.18 0.27 0.37 0.13 0.17 30 24 anese.

99.86 100.37 100.40 100.31 100.24 100.24 ... ...

Specific resistance 30.0 33.2 44.8 52.5 34.2 32.8 100.6 47.7

Temperature coefficient 0.00036 0.00030 0.00033 0.00041 0.00019 0.00021 0.00004 0.00003

The specific resistance is in microhms, i.e. 10-6 ohms per cubic centimetre, and the temperature coefficient in degrees centigrade.

126. Nickel Manganese Copper.

I can find no other reference with regard to this alloy mentioned by Lindeck. Nicholls, however (Silliman's Journal [3], 39, 171, 1890), gives some particulars of alloys of copper and ferromanganese. The following table is taken from Wiedemann's Beiblatter (abstract of Nicholl's paper, 1890, p. 811). All these alloys appear to require annealing at a red heat before their resistances are anything like constant.

Let x be percentage of copper, then 100—x is percentage of "ferromanganese."

Values of x. 100 99.26 91 .88 86.98 80.4 70.65

Specific resistance with respect to copper (? pure) 1 1.19 11.28 20.4 27.5 45.1

Temperature coefficient per degree x 10^6(hard) 3202 2167 138 16 22 -24

Ditto (soft) ... ... 184 80 66 21

If nickel is added, alloys of much the same character are obtained, some with negative temperature coefficients—for instance, one containing 52.51 per cent copper, 31.27 per cent ferromanganese, and 16.22 nickel.

A detailed account of several alloys will be found in a paper by Griffiths (Phil. Trans. 1894, p. 390), but as the constants were determined to a higher order of accuracy than the composition of the material—or, at all events, to a higher degree of accuracy than that to which the materials can be reproduced—there is no advantage in quoting them here.

CHAPTER IV

ELECTROPLATING AND ALLIED ARTS

Sec. 127. Electroplating.

This is an art which is usually deemed worthy of a treatise to itself, but for ordinary laboratory purposes it is a very simple matter—so simple, indeed, that the multiplicity of receipts as given in treatises are rather a source of embarrassment than otherwise.

The fundamental principles of the art are:-

(1) Dirty work cannot be electroplated.

(2) Electroplated surfaces may be rougher, but will not be smoother than the original unplated surface.

(3) The art of electroplating being in advance of the science, it is necessary to be careful as to carrying out instructions in detail. This particularly applies to the conditions which determine whether a metallic deposit shall come down in a reguline or in a crystalline manner.

Sec. 128. The Dipping Bath.

An acid dipping bath is one of the most useful adjuncts to the laboratory, not only for cleansing metals for electroplating, but for cleaning up apparatus made out of bits of brass tube and sheet, and particularly for quickly cleaning binding screws, etc, where it is necessary to ensure good electrical contact.

The cheapest and most satisfactory way in the end is to make up two or three rather large baths to begin with. The glass boxes of storage batteries do very nicely for the purpose, and being generally ground pretty flat at the top, they may be covered by sheets of patent plate glass, and thus preserved from the action of the air.

First Bath. A 30 or 40 per cent solution of commercial caustic soda. Objects may be cleansed from grease in this bath by heating them as hot as is consistent with individual circumstances, and plunging them into it.

It is a considerable advantage to begin by removing grease from articles subsequently to be dipped in an acid bath, both because it saves time and acid, and because more uniform results are obtainable when this is done than when it is omitted. It is a great advantage to have the caustic soda solution hot. This is always done in factories where nickel-plating is carried on, but it is inconvenient in the laboratory. The articles after dipping in the alkali are swilled with water, and may even be scrubbed with a brush, so as to remove greasy matters that have been softened but not entirely removed.

Acid Bath. A convenient bath for laboratory purposes is made by mixing two volumes of strong commercial nitric acid with one of strong sulphuric acid in a cell measuring, say, 12 X 10 X 15 inches.

Copper or brass articles are dipped in this bath for a few seconds, then rinsed with water, then dipped again for a second or two, or until they appear equally white all over, and then withdrawn as rapidly as possible and plunged into a large quantity of clean water. Care must be taken to transfer the articles from the bath to the water as quickly as possible, for if time be allowed for gas to be evolved, the surfaces become mat instead of bright.

In order to save acid it is advisable to make up a third bath, using those odds and ends of acids which gradually accumulate in the laboratory. Sulphuric acid from the balance cases, for instance, mixed with its own volume of commercial nitric acid, does very well.

The objects to be dipped receive a preliminary cleansing by a dip in this bath, the strong bath being reserved for the final dip. Sheet brass and drawn tube, as it comes from the makers, possesses a really fine surface, though this is generally obscured by grease and oxide. Work executed in these materials, cleaned in alkali, and dipped in really strong acid, will be found to present a much better appearance than work which has been filed, unless the latter be afterwards elaborately polished.

On no account must paraffin be allowed to get into any of the baths. When the final bath gets weak it must be relegated to a subordinate position and a new bath set up. A weak acid bath leaves an ugly mottled surface on brass work.

Sec. 129. A metallic surface which it is intended to electroplate must, as has been mentioned, be scrupulously clean. If the metal is not too valuable or delicate, cleaning by dipping is easy and effectual. The following notes will be found to apply to special cases which often occur.

(1) Silver Surfaces intended to be gilt. These are first washed clean with soap and hot water, and polished with whitening. They are then dipped for a moment in a boiling solution of potassium cyanide. A 20 per cent solution of common commercial cyanide does well, but the exact strength is quite immaterial. The cyanide is washed away in a large volume of soft water, and the articles are kept under water till they are scratch-brushed.

Mat surfaces are readily produced on standard silver by dipping in hot strong sulphuric acid. The appearance of new silver coins, which is familiar to everybody, is obtained by this process.

(2) Finely turned and finished Brass Work. If it is intended to nickel-plate such work, and if it is desirable to obtain brightly polished nickel surfaces, the work must be perfectly polished to begin with. Full details as to polishing may be found in workshop books or treatises on watch-making. It will suffice here to say that the brass work is first smoothed by the application of successive grades of emery and oil, or by very fine "dead" smooth files covered with chalk. Polishing is carried out by means of rotten stone and oil applied on leather.

In polishing turned work care must be taken to move the file, emery, or rotten stone to and fro over the work with great regularity, or the surface will end by looking scratchy and irregular. The first process of cleaning is, of course, to remove grease, and this is accomplished best by dipping in a bath of strong hot caustic soda solution, and less perfectly by heating the work and dipping it in the cold caustic soda bath.

During this process a certain amount of chemical action often occurs leading to the brass surface exhibiting some discoloration. The best way of remedying this is to dip the brass into a hot bath of cyanide of potassium solution. If it is inconvenient to employ hot baths or to heat the brass work, good results may be obtained by rubbing the articles over with a large rough cork plentifully lubricated with a strong solution of an alkali.

If the surfaces are very soiled or dirty, a paste of alkali and fine slaked lime may be applied on a cork rubber, and this in my experience has always been most effective and satisfactory in every way, except that it is difficult to get into crevices. If the alkali stains the work, a little cyanide of potassium may be rubbed over the surface in a similar manner.

Brass work treated by either of these methods is to be washed in clean water till the alkali is entirely removed, and may then be nickel-plated without any preliminary scratch-brushing. The treatment in hot baths of alkali and cyanide is the method generally employed in American factories as a preliminary to the nickelling of small brass work for sewing machines, etc.

(3) Copper either for use as the kathode in electrolysis calibration experiments or otherwise is most conveniently prepared by dipping in the acid bath, rinsing quickly in cold water, scratch-brushing under cold water, and transferring at once to the plating bath. In the case where the copper plates require to be weighed they are dipped into very hot distilled water after scratch-brushing, and then dried at once by means of a clean glass cloth.

(4) Aluminium (which, however, does not readily lend itself to plating operations [Footnote: This difficulty has now been overcome. See note, section 138.] ) is best treated by alkali rubbed on with a cork, or by a hot alkaline carbonate where rubbing is inexpedient. The clean aluminium is scratch-brushed under water, and at once transferred to the plating bath.

(5) Iron for Nickel-plating. According to Dr. Gore (Electra-metallurgy, p. 319) the best bath for cleaning iron is made as follows: "One gallon of water and one pound of sulphuric acid are mixed with one or two ounces of zinc (which of course dissolves); to this is added half a pound of nitric acid." The writer has been accustomed to clean iron by mechanical means, to deprive it of grease by caustic alkali, and to finish it off by, means of a hard scratch brush. This process has always worked satisfactorily.

(6) Articles soldered with soft solder containing lead and tin do not readily lend themselves to electrolytic processes, the solder generally becoming black and refusing to be coated with the electro-deposit. Moreover, if soldered articles are boiled for any length of time in caustic alkali during the preliminary cleansing, enough tin will dissolve to form a solution of stannate of potash or soda—strong enough to deposit tin on brass or copper. A method of coppering soldered articles will be described later on.

Sec. 130. Scratch-brushing.

This process is generally indispensable, and to its omission is to be traced most laboratory failures in electroplating. Scratch-brushes may be bought at those interesting shops where "watchmakers' supplies" are sold. It will be well, therefore, to purchase a selection of scratch brushes, for they are made to suit particular kinds of work. They are all made of brass wire, and vary both in hardness and in the fineness of the wire. The simplest kind of scratch brush consists merely of a bundle of wires bound up tightly by another wire, and somewhat "frizzed" out at the ends (Fig. 90). A more useful kind is made just like a rotating brush, and has to be mounted on a lathe (Fig. 91).

Fig. 90. Fig. 91.

The scratch brush is generally, if not always, applied wet; the lubricant generally recommended is stale beer, but this may be replaced by water containing a small quantity of glue, or any other form of gelatine in solution—a mere trace (say .1 per cent) is quite sufficient. Very fair results may be got by using either pure or soapy water. The rotating brushes require to be mounted on a lathe, and may be run at the same speed as would be employed for turning wooden objects of the same dimensions.

Since the brush has to be kept wet by allowing water or its equivalent to drip upon it, it is usual to make a tin trough over which the brush can revolve, and to further protect this by a tin hood to keep the liquid from being thrown all over the room. In many works the brush is arranged to lie partly in the liquid, and this does very well if the hood is effective.

There is a superstition that electro-deposits stick better to scratch-brushed surfaces than to surfaces which have not been so treated, and consequently it is usual to scratch-brush surfaces before electro-deposit. However this may be, there is no doubt that adherence and solidity are promoted by frequent scratch-brushing during the process of depositing metal, especially when the latter tends to come down in a spongy manner.

Gilt surfaces—if the gilding is at all heavy—are generally dull yellow, or even brown, when they come from the bath, and require the scratch brush to cause the gold to brighten, an office which it performs in a quite striking manner. The same remark applies to silvered surfaces, which generally leave the bath a dead white—at all events if the deposit is thick, and if ordinary solutions are employed. In either case the touch of the scratch brush is magical.

Sec. 131. Burnishing.

Burnishers of steel, agate, or bloodstone can be bought at the shops where scratch brushes are sold, and are used to produce the same brightening effect as can be got by scratch-brushing. The same solutions are employed, but rather stronger, and the burnisher is swept over the surface so as to compress the deposited metal. Burnishing is rather an art, but when well done gives a harder and more brilliant (because smoother) surface than the scratch brush. On the whole, steel burnishers are the most convenient if in constant use.

If the burnishing tools have to lie about, steel is apt to rust, unless carefully protected by being plunged in quicklime or thickly smeared with vaseline, and the least speck of rust is fatal to a burnisher. In any case the steel requires to be occasionally repolished by rouge and water on a bit of cloth or felt. The process of burnishing is necessarily somewhat slow and tedious, and as a rule is not worth troubling about except in cases where great permanence is required.

The burnisher is moved over the work somewhat like a pencil with considerable pressure, and care is taken to make the strokes as uniform in direction as possible; otherwise the surface looks non-uniform, and has to be further polished by tripoli, whitening, etc, before it is presentable.

Sec. 132. Silver-plating.

The most convenient solution for general purposes is an 8 to 10 per cent solution of the double cyanide of silver and potassium together with 1 or 2 per cent of "free" potassium cyanide. Great latitude is permissible in the strength of solution and density of current. As commercial cyanide of potassium generally contains an unknown percentage of other salts, which, however, do not interfere with its value for the purpose of silver-plating, the simplest procedure is as follows.

For every 100 c.c. of plating solution about 7 grms. of dry crystallised silver nitrate are required. The equivalent amount of potassium cyanide (if dry and pure) is 5.2 grms, but commercial cyanide may contain from 50 per cent upwards to 96 per cent in the best fused cyanide made from ferrocyanide only. An approximate idea of the cyanide content can be obtained from the dealers when the salt is purchased, and this is all that is required.

A quantity slightly in excess of the computed amount of cyanide is dissolved in distilled water, and this is cautiously added to the solution of the silver nitrate till precipitation is just complete. The supernatant liquors are then drained away, and the precipitate dissolved by adding a sufficiency of the remaining cyanide; this process is assisted by warming and stirring.

An allowance of about one-tenth of the whole cyanide employed may be added to form "free" cyanide, and the solution made up to the strength named. It is advisable to begin with the cyanide in a moderately strong solution, for the sake of ease in dissolving the precipitate.

This solution will deposit silver upon articles of copper or brass immersed in it even without the battery, but the coat will be thin. The solution is used cold, with a current density of about 10 to 20 amperes per square foot. The articles to be silvered are scratch-brushed, washed, and electroplated, till they begin to look undesirably rough. They are then taken out of the bath, rebrushed, and the process continued till a sufficiency of silver is deposited. Four grammes weight of silver (nearly) is deposited per ampere hour. It is best to use a fine silver anode, so that the solution, does not get contaminated by copper.

In most factories it is usual to "quicken" the objects to be silvered before placing them in the electrolysis vats, because the deposit is said to adhere better in consequence of this treatment. I have never found it any improvement for laboratory purposes, but it is easy to do. A dilute (say 2 per cent) solution of cyanide of mercury is required containing a little free cyanide. The objects to be "quickened" are scratch-brushed and dipped into the cyanide of mercury solution till they are uniformly white; it is generally agreed that the less the mercury deposited the better, so long as a perfect coating is obtained. The objects are rinsed after quickening, and put in the depositing bath at once.

The mat surface of silver obtained by electrolysis of the cyanide is very beautiful—one of the most beautiful things in nature—shining with incomparable crystalline whiteness. So delicate is it, however, for so great is the surface it exposes, that it is generally rapidly deteriorated by exposure to the air. It may be protected to some extent by lacquering with pale lacquer, but it loses some of its brilliancy and purity in the process. The deposit is generally scratch-brushed or burnished down to a regular reflecting surface.

Sec. 133. Cold Silvering.

A thin but brilliant coat of silver may be readily applied to small articles of brass or copper in the following way. A saturated solution of sodium sulphite (neutral) is prepared, and into this a 10 per cent solution of nitrate of silver is poured so long as the precipitate formed is redissolved. A good deal of silver may be got into solution in this way. Articles to be silvered need only to be cleaned, brushed, and dipped in this solution till a coat of the required thickness is obtained.

I must admit, however, that the coating thus laid on does not appear to be so permanent as one deposited by simple immersion from the cyanide solution, even though it is thicker. The cyanide plating solution will itself give a good coat of silver if it is used boiling, and if a little potassium cyanide be added.

For purposes of instrument construction, however, a thin coat of silver is seldom to be recommended, on account of its liability to tarnish and its rapid destruction when any attempt is made to repolish it. For these reasons, nickel or gold plating is much to be preferred.

Sec. 134. Gilding.

This art deserves to be much more widely practised than is usual in laboratories. Regarded as a means of preserving brass, copper, or steel, it is not appreciably more "time robbing" than lacquering, and gives infinitely better results. Moreover, it is not much more expensive. Strange as it may seem, the costliness of gilding seldom lies in the value of the gold deposited; the chief cost is in the chemicals employed to clean the work, and in interest on the not inconsiderable outlay on the solution and anode.

The easiest metal to gild is silver, and it is not unusual to give base metals a thin coating of silver or copper, or both, one after the other, before gilding, in order to secure uniformity. To illustrate the virtue of a thin layer of gold, I will mention the following experiment. About three years ago I learned for the first time that to "clean" the silver used in a small household required at least an hour's labour per diem. I further ascertained that most of this time is spent on the polishing part of the process.

As this seemed a waste of labour, I decided to try the effect of gilding. In order to give the proposal a fair trial I gilt the following articles: half a dozen table spoons and forks, a dozen dessert forks and spoons, and a dozen tea spoons. These were all common electroplated ware. They were weighed before and after gilding, and it was with difficulty that the increase of weight was detected, even though a fine bullion balance was employed. On calculating back to money, it appeared that the value of the gold deposited was about threepence. Assuming that an equal weight of silver had been accidentally dissolved by the free cyanide during the plating—which is unlikely—the total amount of gold deposited would be worth, say, sixpence.

After three years' continuous use the gilding is still perfect, except at the points on which the spoons and forks rest, where it is certainly rather shabby. Meanwhile the "gold" plate only requires to be washed with hot water and soap to keep it in perfect order, a much more cleanly and expeditious process than that of silver cleaning.

Sec. 135. Preparing Surfaces for Gilding.

Ordinary brass work—rough or smooth—may for purposes of preservation be dipped, scratch-brushed, and gilt at once. Seven years ago the writer gilt the inside of the head of a copper water still, and simply scratch-brushed it; it is to-day in as good order as when it was first done. If it is intended to gild work from the first, with the view of making an exceptionally fine job of it, "gilding metal," i.e. brass containing one to one and a quarter ounces of zinc to the pound of copper may be specified. From its costliness, however, this is only desirable for small work.

Iron and steel are generally given a preliminary coating of copper, but this may be dispensed with though with no advantage—by using a particular process of gilding.

Base metals, zinc, pewter, lead, etc, are first coppered in a cyanide of copper solution, as will be described under the head of Copper-plating. If it is intended to gild soldered articles, the preliminary coating of copper is essential.

The most convenient vessel for holding a gilding solution is undoubtedly one formed of enamelled iron. Particularly useful are the buckets and "billies" (i.e. cylindrical cans) made of this material. These vessels may be heated without any fear of a smash, and do not appear to be appreciably affected by gilding solutions—at all events during several days or weeks. The avoidance of all risk of breakage when twenty or thirty pounds' worth of solution is in question is a matter of importance.

Under no circumstances is it desirable to use anything but the purest gold and best fused cyanide (called "gold" cyanide) in the preparation of the solutions. The appearance of a pure gold deposit is far richer than of one containing silver, and its resistance to the atmosphere is perfect; moreover, in chemico-physical processes one has the satisfaction of knowing what one is dealing with.

Sec. 136. Gilding Solutions.

The strength of solution necessary for gilding brass, copper, and silver is not very material. About one to two pounds of "gold" potassium cyanide (? 96 per cent KCN) per gallon does very well. The gold is best introduced by electrolysing from a large to a small gold electrode. One purchases a plate of pure gold either from the mint or from reliable metallurgists (say Messrs. Johnson and Matthey of London), and from this electrodes are cut.

The relative areas of the electrodes do not really much matter. I have used an anode of four times the area of the cathode. The solution is preferably heated to a temperature of about 50 deg. C, and a strong current is sent through it, say twenty amperes to the square foot of anode. The electrodes must be suspended below the surface of the solution by means of platinum wires. If the gold plates are only partly immersed, they dissolve much more rapidly where they cut the surface, possibly on account of the effect of convection currents, though so far as the writer is aware no proper explanation has yet been given.

After a time gold begins to be deposited on the cathode in a powdery form, for which reason it is a good plan to begin by wrapping the latter in filter paper. The process has gone on for a sufficient time when a clean bit of platinum foil immersed in the place of the cathode becomes properly gilt at a current density of about ten amperes per square foot.

The powdery gold deposited on the cathode while preparing the solution can be scraped off and melted for further use, or the whole cathode may now be used as an anode. The platinum foil testing cathode may also be "stripped" by making it an anode, and is for this reason preferable to German silver or copper, which would contaminate the solution while the "stripping" process was in progress.

For general purposes a current density of say ten to fifteen amperes per square foot may be used, but this may be considerably varied, so long as the upper limit is not greatly overpassed. During gold-plating there is a considerable advantage in keeping the electrodes moving or the solution stirred.

After immersing the cleaned and scratch-brushed articles, depositing may go on for about three minutes, after which they are removed from the bath and examined, in order to detect any want of uniformity in the deposit.

The articles should be entirely immersed; if this is not done, irregularity is apt to appear at the surface. Platinum wires employed as suspenders, and coated along with the articles to be gilt, may also be cleaned without loss by making them anodes. If, on examination, all is found to be going on well, reimmerse the cathodes, and continue plating till they appear of a dull yellowish brown (this will occur in about four minutes), then remove them, rinse and scratch-brush them, and replace them in the bath.

When a second coat appears to be getting rather brown than yellowish brown, i.e. of the colour of wet wash-leather, the removal, followed by scratch-brushing, may be repeated, and for nearly all laboratory purposes, the articles are now fully gilt.

The coating of gold deposited from a hot cyanide solution is spongy in the extreme, and if the maximum wear-resisting effect is to be obtained, it is advisable to burnish the gold rather than to rely upon the scratch brush alone.

If the area of the cathode exceeds that of the anode the solution is said to grow weaker, and vice versa. This may be remedied in the former case by an obvious readjustment; the latter introduces no difficulty so far as I know except when plating iron or steel.

The student need not be troubled at the poor appearance of the deposit before it is scratch-brushed. Heavy gold deposits are almost always dull, not to say dirty, in appearance till the burnisher or scratch brush is applied. On the other hand, the deposit ought not to get anything like black in colour.

The following indications of defects may be noted—they are taken from Gore. I have never been really troubled with them.

The deposit is blackish. This is caused by too strong a current in too weak a bath. This may be remedied to some extent by stirring or keeping the cathode in motion. The obvious remedy is to add a little cyanide of gold.

The gold anode gets incrusted. This is a sign that the bath is deficient in potassium cyanide. The gold anode gets black and gives off gas. The solution is deficient in cyanide, and too large a current is being passed.

If a bright surface is desired direct from the bath, some caustic potash (say 2 per cent) may, according to Gore, be added, or the articles may be plated only slightly by using a weak current and taking them out directly they show signs of getting dull. By a weak current I mean one of about five amperes per square foot.

The deposit is said to be denser if the solution be heated as directed; but the bath will gild, though not quite so freely when cold.

To gild iron or steel directly, dilute the bath as above recommended some five or six times, add about 1 per cent of potassium cyanide, and gild with a very weak current (say two or three amperes per square foot) in the cold. Frequent scratch-brushing will be found requisite to secure proper adherence.

It is generally recommended to gild brass or German silver in solutions which are rather weak, but in the small practice which occurs in the laboratory a solution prepared as suggested does perfectly for everything except iron or steel. The scratch-brushing should be done over a large photographic developing dish to avoid loss of gold. It is a good plan to rinse the articles after leaving the bath in a limited quantity of distilled water, which is afterwards placed in a "residue" bottle, and then to scratch-brush them by hand over the dish to catch fine gold. When any loose dust is removed the articles may be scratched in the lathe without appreciable further loss.

Silver-gilt articles tend to get discoloured by use, but this discoloration can be removed by soap and water. After long use a gold cyanide bath tends to alter greatly in composition, In general, the bath tends to grow weaker, from the fact that there is a strong temptation to gild as many articles at once as possible.

It is therefore a good plan to keep a rough profit and loss account of the gold in order to find the quantity in solution. Fifty dwts. per gallon (or 78 grms. per 4.5 litres) is recommended. A gallon of solution of this strength is worth about eleven pounds sterling in gold and cyanide, and a serviceable anode will be worth about 10 pounds. (Fine gold is worth nominally four pounds four shillings and eleven pence ha'penny per oz.) Gold may be easily obtained containing less impurity than one part in ten thousand.

Sec. 137. Plating with Copper.

Copper may be deposited from almost any of its salts in reguline form, the sulphate and nitrate being most usually employed. In the laboratory a nearly saturated solution of sulphate of copper with 1 or 2 per cent of sulphuric acid will answer most purposes. A current density of, at most, fifteen amperes per square foot may be used, either for obtaining solid deposits for constructional purposes or for calibrating current measuring instruments by electrolysis. A copper anode is of course employed.

When coppering with a view to obtaining thick deposits it is a good plan to place the electrodes several inches apart, and, if possible, to keep the liquid stirred, as there is a considerable tendency on the part of copper deposits to grow out into mossy masses wherever the current density exceeds the limit mentioned. As the masses grow towards the anode the defect naturally tends to increase of itself, hence the necessity for care. The phenomenon is particularly marked at the edges and corners of the cathode.

If the deposit becomes markedly irregular, the best plan is to stop the process and file the face of the deposit down to approximate smoothness. In coppering it is of the utmost importance that the cathode be clean and free from grease; it must never be touched (by the finger, for instance) from the time it is scratch-brushed till it is immersed in the plating bath. Any grease or oxidation tends to prevent the copper deposit adhering properly.

A copper deposit oxidises very easily when exposed to the air. Consequently if the surface be required free from oxide, as, for instance, when it is to be silvered or gilt, it must be quickly washed when withdrawn from the coppering bath, scratch-brushed, and transferred immediately to the silvering or gilding bath.

If the surface is to be dried, as in electrolysis calibrations, it must be rinsed quickly with boiling water and pressed between sheets of filter paper. Another method which has been recommended is to rinse the copper in—water slightly acidulated with sulphuric acid (which prevents oxidation), then in distilled water, and to dry by blotting paper and in front of a fire, taking care not to make the plate too hot. The wash water is sufficiently acidulated by the addition of two or three drops of acid per litre. So far as I know, the method of washing in acidulated water was first proposed by Mr. T. Gray.

Sec. 138. Coppering Aluminium.

A good adherent deposit of copper on aluminium used to be considered a desideratum in the days when it afforded the only means of soldering the latter. Many receipts have been published from time to time, and I have tried, I think, most of them. On no occasion, however, till this year (1896), have I succeeded in obtaining a deposit which would not strip after it was tinned and soldered, though it is not difficult to get apparently adherent deposits so long as they are not operated upon by the soldering iron. The best of the many solutions which have been proposed in years gone by is very dilute cupric nitrate with about 5 per cent of free nitric acid.

The problem of electroplating aluminium which I have indicated as awaiting a solution has at last found one. In the Archives des Sciences physiques et naturelles de Geneve for December 1895 (vol. xxxiv. p. 563) there is a paper by M. Margot on the subject, which discloses a perfectly successful method of plating aluminium with copper. The paper itself deals in an interesting way with the theory of the matter—however, the result is as follows.

(1) The aluminium articles are boiled for a few minutes in a strong solution of ordinary washing soda. The aluminium surface is thus corroded somewhat, and rendered favourable to the deposit of an adherent film of copper. After removal from the soda solution the aluminium is well washed and brushed in running water.

(2) The articles are dipped for thirty seconds or so in a hot 5 per cent solution of pure hydrochloric acid.

(3) After dipping in the hydrochloric acid, the work is instantly plunged into clean water for about one second, so as to remove nearly, but not quite, all of the aluminium chloride.

(4) The work is transferred to a cold dilute (say 5 per cent) solution of cupric sulphate slightly acidulated with sulphuric acid. The degree of acidulation does not appear to be very important, but about one-tenth per cent of strong acid does well.

If the preliminary processes have been properly carried out the aluminium will become coated with copper, and the process is accompanied by the disengagement of gas. It appears to be a rule that if gas is not given off, the film of copper deposited is non-adherent. The work must be left in the copper sulphate solution till it has received a uniform coating of copper.

(5) When this is the case the work is removed—well washed so as to get rid of the rest of the aluminium chloride, and then electroplated by the battery in the ordinary copper sulphate bath.

If the operation (4) does not appear to give a uniform coat, or if gas is not evolved from every part of the aluminium surface, I find that operations (2) and (3) may be repeated without danger, provided that the dip in the hydrochloric acid is shortened to two or three seconds.

The copper layer obtained by Margot's method is perfectly adherent—even when used as a base for ordinary solder—though in this case it can be stripped if sufficient force is applied.

Since the solder recommended by M. Margot for aluminium contains zinc, it does not run well when used to unite aluminium to copper, brass, iron, etc. In this case, therefore, I have found the most advantageous method of soldering to be by way of a preliminary copper-plating.

The success of M. Margot's method depends in my experience on obtaining just the proper amount of aluminium chloride in contact with the aluminium when the latter is immersed in the copper sulphate solution.

Sec. 139. The process of copper-plating from sulphate or nitrate may, according to Mr. Swan (Journal of the Royal Institution, 1892, p. 630), be considerably accelerated by the addition of a trace of gelatine to the solution. As success appears to depend upon hitting the exact percentage amount of the gelatine, which must in any case be but a fraction of one per cent, and as Mr. Swan refrains from stating what the amount is, I am unable to give more precise instructions. A few experiments made on the subject failed, doubtless through the gelatine content not having been rightly adjusted. Mr. Swan claims to be able to get a hard deposit of copper with a current density of 1000 amperes per square foot, but seems to recommend about one-tenth of that amount for general use.

The solution employed is a mixture of nitrate of copper and ammonium chloride—proportions not stated. Electrolytic copper, as generally prepared, is very pure, but this is a mere accident depending on the impurities which, as a rule, have to be got rid of. Electrolysis seems to have no effect in purifying from arsenic, for instance.

Roughly speaking, about 11 grms. of copper are deposited per ampere hour from cupric salt solutions. When the current density is too high the anode suffers by oxidation, and this introduces a large and very variable resistance into the circuit.

Sec. 140. Alkaline Coppering Solution

Coppering Base Metals. It is often desirable to coat lead, zinc, pewter, iron, etc, with a firm and uniform layer of copper preparatory to gilding or silvering. If copper or brass articles are soldered with soft solder it is found that the solder does not become silvered or gilt along with the rest of the material, but remains uncoated and of an ugly dark colour. This defect is got over by giving a preliminary coating of copper.

This is done in an alkaline solution, generally containing cyanogen and ammonia. The following method has succeeded remarkably well with me. The receipt was taken originally from Gore's Electro-metallurgy, p. 208. A solution is made of 50 grms. of potassium cyanide (ordinary commercial, say, 75 per cent) and 30 grms. of sodium bisulphite in I.5 litres of water. Thirty-five grammes of cupric acetate are dissolved in a litre of water, and 20 cubic centimetres of the strongest liquid ammonia are added. The precipitate formed must be more or less dissolved to a strong blue solution. The cyanide and bisulphite solution is then added with warming till the blue colour is destroyed. This usually requires the exact amount of cyanide and bisulphite mentioned, but I have not found it essential to entirely destroy the colour.

The solution contains cuprocyanide of sodium and ammonium (?), which is not very soluble, and this salt tends to be deposited in granular crystalline masses on standing. However, at a temperature of 50 deg. C. the above receipt gives an excellent coppering liquid, which will coat zinc with a fine reguline deposit. Brass or copper partly smeared with solder will receive a deposit of copper on the latter as well as on the former, and, moreover, a deposit which appears to be perfectly uniform.

In using the bath the anode tends, as a rule, to become incrusted, and this rapidly increases the resistance of the cell, so that the current falls off quickly. The articles should be scratch-brushed and plated for about two minutes with a current density of about ten amperes per square foot.

As soon as the deposit begins to look red the articles are to be removed and rebrushed, after which the process may be continued. About five minutes' plating will give a copper deposit quite thick enough after scratch-brushing to allow of a very even gilding or silvering.

Aluminium appears to be fairly coated, but, as usual, the copper strips after soldering. Iron receives an excellent and adherent coat.

I do not think that the formation of a crust upon the anode can be entirely prevented. According to Gore, its formation is due to the solution being too poor in copper, but I have added a solution of the acetate of copper and ammonium till the colour was bright blue without in any way reducing the incrustation. If the solutions become violently blue it is perhaps as well to add a little more cyanide and bisulphite, but I have not found such an addition necessary. The process is one of the easiest and most satisfactory in electro-metallurgy.

Sec. 141. Nickel-plating.

An examination of several American samples of nickel-plated goods has disclosed that the coating of nickel is, as a rule, exceedingly thin. This is what one would expect from laboratory repetition of the processes employed.

Commercial practice in the matter of the composition of nickelling solutions appears to vary a good deal. Thin coatings of nickel may be readily given in a solution of the double sulphate of nickel and ammonia, which does rather better if slightly alkaline. Deposits from this solution, however, become gray if of any thickness, and, moreover, are-apt to flake off the work. The following solution has given very good results with me. It is mentioned, together with others, in the Electrical Review, 7th June 1895.

The ingredients are:-

Nickel sulphate 5 parts

Ammonia sufficient to neutralise the nickel salt.

Ammonium tartrate 3.75 parts

Tannin 0.025 parts

Water 100 parts

The nickel sulphate and ammonia are dissolved in half the water, the ammonium tartrate in the other half with the tannin. The solutions are mixed and filtered at about 40 deg. C. This solution works well at ordinary temperatures, or slightly warm, with a current density of ten amperes per square foot. In an experiment made for the purpose I found that plating may go on for an hour in this solution before the deposit begins to show signs of flaking off. The deposit is of a fine white colour.

The resistance of the bath is rather high and rather variable, consequently it is as well to have a current indicator in circuit, and it may well happen that five or six volts will be found requisite to get the current up to the value stated. For nickelling small objects of brass, such as binding screws, etc, it is very necessary to be careful as to the state of polish and uniformity of their surfaces before placing them in the plating bath. A polished surface will appear when coated as a polished surface, and a mat surface as a mat surface; moreover, any local irregularity, such as a speck of a foreign metal, will give rise to an ugly spot in the nickelling bath. For this reason it is often advisable to commence with a coat of copper laid on in an alkaline solution and scratch-brushed to absolute uniformity.

An examination of the work will, however, disclose whether such a course is desirable or not; it is not done in American practice, at all events for small brass objects. These are cleaned in alkali and in boiling cyanide, which does not render a polished surface mat, as weak acid is apt to do, and are then coated with a current density of about ten amperes per square foot.

In spite of what is to be found in books as to the ease with which nickel deposits may be polished, I find that the mat surface obtained by plating on an imperfectly polished cathode of iron is by no means easily polished either by fine emery, tripoli, or rouge. Consequently, as in the case of brass, if a polished surface is desired, it must be first prepared on the unplated cathode. In this case, even if the deposit appears dull, but not gray, it may be easily polished by tripoli and water, using a cork as the polisher. Scratch-brushing with brass wire, however, though possibly not with German silver wire, brightens the deposit, but discolours it. When the deposit becomes gray I have not succeeded in polishing it satisfactorily.

Soldered brass or iron may be satisfactorily coated with nickel by giving it a preliminary coating of copper in the cyanide bath. On the whole, I recommend in general that iron be first coated with copper in the alkaline bath, scratch-brushed, and then nickel-plated, and this whether the iron appears to be uniform or not. Much smoother, thicker, and stronger coats of nickel are obtained upon the copper-plated surface than on the iron one, and the coating does not become discoloured (? by iron rust) in the same way that a coating on bare iron does. The copper surface may be plated for at least an hour at a density of ten amperes per square foot without scaling.

Scales or circles divided on brass may be greatly improved in durability by nickel—plating. For this purpose the brass must be highly polished and divided before it is nickelled.

The plating should be continued for a few minutes only, when a very bright but thin coat of nickel will be deposited; it then only remains to wash and dry the work, and this must be done at once. If the nickel is deposited before the scale or circle is engraved, very fine and legible divisions are obtained, but there is a risk that flakes of nickel may become detached here and there in the process of engraving.

142. Miscellaneous Notes on Electroplating.

Occasionally it is desirable to make a metallic mould or other object of complex shape. The quickest way to do this is to carve the object out of hard paraffin, and then copy it by electrotyping. Electrotype moulds can be made in many ways. The easiest way perhaps is to take a casting in plaster of Paris, or by means of pressure in warm gutta-percha.

In cases where the mould will not draw, recourse must be had to the devices of iron-founders, i.e. the plaster cast must be made in suitable pieces, and these afterwards fitted together. This process can occasionally be replaced by another in which the moulding material is a mixture of treacle and glue. The glue is soaked in cold water till it is completely soft. The superfluous water thrown away, one-fourth part by volume of thick treacle is added, and the mixture is melted on the water bath; during which process stirring has to be resorted to, to produce a uniform mixture.

This liquid forms the moulding mixture, and it is allowed to flow round the object to be copied, contained in a suitable box, whose sides have been slightly oiled. The object to be copied should also be oiled. After some hours, when the glue mixture has set, it will be found to be highly elastic, so that it may be pulled away from the mould, and afterwards resume very nearly its original form.

One drawback to the use of these moulds lies in the fact that the gelatine will rarely stand the plating solution without undergoing change, but this may be partially obviated by dipping it for a few seconds in a 10 per cent solution of bichromate of potash, exposing it to the sunlight for a few minutes, and then rinsing it.

In order to render the surface conducting, it is washed over with a solution of a gold or silver salt, and the latter reduced in situ to metal by a suitable reagent. A solution of phosphorus is the most usual one (see Gore, Electro-metallurgy, p. 216). Such a mould may be copper-plated in the sulphate bath, connection being made by wires suitably thrust into the material.

Plaster of Paris moulds require to be dried and waxed by standing on a hot plate in melted wax before they are immersed in the plating bath. In this case the surface is best made conducting either by silvering it by the silvering process used for mirrors, or by brushing it over with good black lead rendered more conducting by moistening with an ethereal solution of chloride of gold and then drying in the sun.

The brushing requires a stiff camel's-hair pencil of large size cut so that the hairs project to a distance of about a quarter of an inch from the holder. The brushing must continue till the surface is bright, and is often a lengthy process.

The same process of blackleading may be employed to get a coat of deposited metal which will strip easily from the cathode.

In all cases where extensive deposits of copper are required, the growth takes place too rapidly at the corners. Consequently it is often desirable to localise the action of the deposit. A "stopping" of ordinary copal varnish seems to be the usual thing, but a thin coat of wax or paraffin or photographic (black) varnish does practically as well.

I do not propose to deal with the subject of electrotyping to any extent, for if practised as an art, a good many little precautions are required, as the student may read in Gore's Electro-metallurgy. The above instructions will be found sufficient for the occasional use of the process in the construction of apparatus, etc. There is no advantage in attempting to hurry the process, a current density of about ten amperes per square foot being quite suitable and sufficiently low to ensure a solid deposit.

Sec. 143. Blacking Brass Surfaces.

A really uniform dead-black surface is difficult to produce on brass by chemical means. A paste of nitrate of copper and nitrate of silver heated on the brass is said to give a dead-black surface, but I have not succeeded in making it act uniformly. For optical purposes the best plan is to use a paint made up of "drop" black, ground very fine with a little shellac varnish, and diluted for use with alcohol. No more varnish than is necessary to cause the black to hold together should be employed.

In general, if the paint be ground to the consistency of very thick cream with ordinary shellac varnish it will be found to work well when reduced by alcohol to a free painting consistency.

A very fine gray and black finish, with a rather metallic lustre, may be easily given to brass work. For this purpose a dilute solution of platinum tetrachloride (not stronger than 1 per cent) is prepared by dissolving the salt in distilled water. The polished brass work is cleaned by rubbing with a cork and strong potash till all grease has disappeared, as shown by water standing uniformly on the metal and draining away without gathering into drops.

After copious washing the work is wholly immersed in a considerable volume of the platinum tetrachloride solution at the ordinary temperature. After about a quarter of an hour the brass may be taken out and washed. The surface will be found to be nicely and uniformly coated if the above instructions have been carried out, but any finger-marks or otherwise dirty places will cause irregularity of deposit. If the process has been successful it will be found that the deposit adheres perfectly, hardly any of it being removed by vigorous rubbing with a cloth. If the deposit is allowed to thicken—either by leaving the articles in the solution too long or heating the solution, or having it too strong—it will merely rub off and leave an irregular surface.

This process succeeds well with yellow brass and Muntz metal, either cast or rolled, but it does not give quite such uniform (though still good) results with gun-metal, on which, however, the deposit is darker and deader in appearance.

A book might be written (several have been written) on the art of metal colouring, but though doubtless a beautiful and delicate art, it is of little service in the laboratory. For further information the reader may consult a work by Hiorns.

Sec. 144. Sieves.

Properly graded sieves with meshes of a reliable size are often of great use. They should be made out of proper "bolting" cloth, a beautiful material made for flour-millers. Messrs. Henry Simon and Company of Manchester have kindly furnished me with the following table of materials used in flour-milling.

Sieves made of these materials will be found to work much more quickly and satisfactorily than those made from ordinary muslin or wire gauze.

Relative Bolting Value of Silk, Wire, and Grit Gauze

Threads per inch Trade No. Trade No. Trade No. of Approximate. of Silk. of Wire. Grit Gauze.

18 0000 18 16

22 000 20 20

28 00 26 26

38 0 32 34

48 1 40 44

52 2 45 50

56 3 50 54

60 4 56 58

64 5 60 60

72 6 64 66

80 7 70 70

84 8 80 80

94 9

106 10

114 11

124 12

130 13

139 14

148 15

156 16

163 17

167 18

170 19

173 20

Sec. 145. Pottery making in the Laboratory.

When large pieces of earthenware of any special design are required, recourse must be had to a pottery. Small vessels, plates, parts of machines, etc, can often be made in the laboratory in less time than it would take to explain to the potter what is required. For this purpose any good pipeclay may be employed. I have used a white pipe-clay dug up in the laboratory garden with complete success.

The clay should be kneaded with water and squeezed through a cloth to separate grit. It is then mixed with its own volume or thereabouts of powdered porcelain evaporating basins, broken basins being kept for this purpose. The smoothness of the resulting earthenware will depend on the fineness to which the porcelain fragments have been reduced. I have found that fragments passing a sieve of sixty threads to the inch run, do very well, though the resulting earthenware is decidedly rough.

The porcelain and clay being thoroughly incorporated by kneading, the articles are moulded, it being borne in mind that they will contract somewhat on firing. [Footnote: The contraction depends on the temperature attained as well as on the time. An allowance of one part in twelve will be suitable in the case considered.] The clay should be as stiff as is convenient to work, and after moulding must be allowed to get thoroughly dry by standing in an airy place; the drying must not be forced, especially at first, or the clay will crack.

Small articles are readily fired in a Fletcher's crucible furnace supplied with a gas blow-pipe; the furnace is heated gradually to begin with. When a dull red heat is attained, the full power of the blast may be turned on, and the furnace kept at its maximum temperature for three or four hours at least, though on an emergency shorter periods may be made to do.

The articles are supported on a bed of white sand; after firing, the crucible furnace must be allowed to cool slowly. It must be remembered that the furnace walls will get hot externally after the first few hours, consequently the furnace must be supported on bricks, to protect the bench.

The pottery when cold may be dressed on a grindstone if necessary. This amateur pottery will be found of service in making small fittings for switch-boards, commutators, and in electrical work generally.

Pottery made as described is very hard and strong, the hardness and strength depending in a great degree on the proportion of powdered porcelain added to the clay, as well, of course, as on the quality of both of these materials.

It is a good plan to knead a considerable quantity of the mixture, which may then be placed in a well-covered jar, and kept damp by the addition of a little water.

Pottery thus made does not require to be glazed, but, of course, a glaze can be obtained by any of the methods described in works on pottery manufacture. The following glaze has been recommended to me by a very competent potter:-

Litharge

7 parts by weight

Ground flint

2 parts by weight

Cornish stone or felspar

1 parts by weight

These ingredients are to be ground up till they will pass the finest sieve—say 180 threads to the inch. They are then mixed with water till they form a paste of the consistency of cream. They must, of course, be mixed together perfectly. The ware to be glazed is dipped into the cream after the first firing; it is then dried as before and refired. The glaze will melt at a bright red heat, but it will crack if not fired harder; the harder it is fired the less likely is it to crack.

If colouring matters are added they must be ground in a mill free from iron till they are so fine that a thick blanket filter will not filter them when suspended in water. This remark applies particularly to oxide of cobalt.

APPENDIX

PLATINISING GLASS

IN the Philosophical Magazine for July 1888 (vol. xxvi. p. 1) there is a paper by Professor Kundt translated from the Sitzungsberichte of the Prussian Academy. This paper deals with the indices of refraction of metals. Thin prisms were obtained by depositing metals electrolytically on glass surfaces coated with platinum. The preparation of these surfaces is troublesome. Kundt recounts that no less than two thousand trials were made before success was attained. A detailed account of the preparation of these surfaces is not given by Kundt, but one is promised—a promise unfortunately unfulfilled so far as I am able to discover. A hunt through the literature led to the discovery of the following references: Central Zeitung fuer Optik und Mechanik, p. 142 (1888); Dingler's Polytechnik Journal, Vol. cxcv. p. 464; Comptes Rendus, vol. lxx. (1870).

The original communication is a paper by Jouglet in the Comptes Rendus, of which the other references are abstracts. The account in Dingier is a literal translation of the original paper, and the note in the Central Zeitung is abbreviated sufficiently to be of no value. The details are briefly as follows:-

One hundred grams of platinum are dissolved in aqua regia and the solution is dried on the sand bath, without, however, producing decomposition. Though the instructions are not definite, I presume that the formation of PtCl4 is contemplated.

The dried salt is added little by little to rectified oil of lavender, placed on a glass paint-grinding plate, and the salt and oil are ground together with a muller. Care is required to prevent any appreciable rise of temperature which would decompose the compound aimed at, and it is for this reason that the salt is to be added gradually. Of course the absorption of water from the air must be prevented from taking place as far as possible. Finally, the compound is diluted by adding oil of lavender up to a total weight of 1400 grams (of oil).

The liquid is poured into a porcelain dish and left absolutely at rest for eight days. It is then decanted and filtered, left six days at rest, and again decanted (if necessary). The liquid should have a specific gravity of 5 deg. on the acid hydrometer. (If by this the Baume scale is intended, the corresponding specific gravity would be 1.037.) A second liquid is prepared by grinding up 25 grams of litharge with 25 grams of borate of lead and 8 to 10 grams of oil of lavender. The grinding must be thoroughly carried out.

This liquid is to be added to the one first described, and the whole well mixed. The resulting fluid constitutes the platinising liquid, and is applied as follows:-

A sheet of clean glass is held vertically, and the liquid is painted over it, carrying the brush from the lower to the upper edge. The layer of oil dries slowly, and when it is dry the painting is again proceeded with, moving the brush this time from right to left; and similarly the process is repeated twice, the brush being carried from top to bottom and left to right. This is with the object of securing great uniformity in the coating. Nothing is said as to the manner in which the glass is to be dried.

The dried glass is finally heated to a temperature of dull redness in a muffle furnace. The resinous layer burns away without running or bubbling, and leaves a dull metallic surface. As the temperature rises this suddenly brightens, and we obtain the desired surface (which probably consists of an alloy of lead and platinum). It is bright only on the surface away from the glass.

I have not had an opportunity of trying this process since I discovered the detailed account given by Jouglet; but many modifications have been tried in the laboratory of the Sydney University by Mr. Pollock, starting from the imperfect note in the Central Zeitung, which led to no real success.

It was found that it is perfectly easy to obtain brilliant films of platinum by the following process, provided that the presence of a few pin-holes does not matter.

The platinum salt employed is what is bought under the name of platinic chloride; it is, however, probably a mixture of this salt and hydro-chloro-platinic acid, and has all the appearance of having been obtained by evaporating a solution of platinum in aqua regia to dryness on the water bath. A solution of this salt in distilled water is prepared; the strength does not seem to matter very much, but perhaps one of salt to ninety-nine water may be regarded as a standard proportion. To this solution is added a few drops of ordinary gum water (i.e. a solution of dextrin). The exact quantity does not matter, but perhaps about 2 per cent may be mentioned as giving good results.

The glass is painted over with this solution and dried slowly on the water bath. When the glass is dry, and covered with a uniform hard film of gum and platinum salt free from bubble holes, it is heated to redness in a muffle furnace. The necessary and sufficient temperature is reached as soon as the glass is just sensibly red-hot.

The mirrors obtained in this way are very brilliant on the free platinum surface. If the gum be omitted, the platinum will have a mat surface; and if too much gum be used, the platinum will get spotty by bubbles bursting. There does not appear to be any advantage in using lead.

It is quite essential that the film be dry and hard before the glass is fired.

THE END