Chapter II: Part 2

+--------+---------------+-------+--------+----------+--------+
| Metal. |Denomination of|Length.|Breadth.|Thickness.|Average |
| |intended coin. | | | |Weight. |
+--------+---------------+-------+--------+----------+--------+
| | |Inches.|Inches. | Inches. |Troy oz.|
|Gold | Sovereign | 24 | 1·375 | 1·000 | 320 |
| | Half-Sovereign| 24 | 1·125 | 1·000 | 250 |
+--------+-----------------------+--------+----------+--------+
| | Crown | 22 | 2·750 | 1·000 | 300 |
| | Half-Crown | 22 | 2·500 | 1·000 | 240 |
| | Florin | 21 | 2·125 | 1·000 | 220 |
|Silver | Shilling | 21 | 1·437 | 1·000 | 150 |
| | Sixpence | 21 | 1·125 | 1·000 | 120 |
| | Fourpence | 21 | 0·875 | 1·000 | 96 |
| | Threepence | 21 | 0·875 | 1·000 | 96 |
+--------+-----------------------+--------+----------+--------+
| | Penny | 24 | 2·500 | 0·375 | 100 |
|Bronze | Halfpenny | 24 | 3·000 | 0·375 | 117 |
| | Farthing | 24 | 3·000 | 0·375 | 117 |
+--------+---------------+-------+--------+----------+--------+

It is to be regretted that crowns, half-crowns, and fourpences are no longer coined. It has been represented to me by many manufacturers and others who employ vast numbers of men, that the disuse of the fourpenny piece has caused considerable inconvenience in the weekly payments to their workpeople. The objection to this coin would appear to be that it is so nearly the diameter of the threepenny piece; but this is obviated by the fact that the edge of the latter is plain, while that of the fourpenny is crenated or “milled.”

The assay pieces,[14] when cut from the bars, are placed in the divisions of a tray going from left to right, so that there can be no mistake as to the number of the pot from which the metal comes; each assay piece is then placed in a small envelope, marked with a distinctive mark, to characterise the pot from which it came. The assay pieces are sent in equal numbers to the two non-resident assayers, who determine, irrespectively of each other, the amount of gold found in each piece. Suppose we trace one assay piece, and imagine that 120 are undergoing the same process at the same time. The assay piece is flattened out into a kind of ribbon, and from it three pieces are cut, each weighing half a gramme. Each piece of this weight takes the name of a _thousand_, and is represented by the figures 1,000; but of course it may be called a pound, a ton, or by any other denomination. Each piece is wrapped in paper, with three times its weight of silver in two pieces, both metals being of the same thickness. The parcels, as they may now be called, are ready for the next process, and are arranged in their proper places from left to right in the divisions of a tray, and taken to the assaying furnace. In the assaying furnace is placed a Payen[15] muffle, or kind of oven, perforated at places so as to allow a limited amount of atmospheric air to pass from the interior through its sides into the furnace. The muffle is surrounded with fuel, so that it is kept at a good—almost white—heat. On the floor of the muffle is sprinkled some bone ash—obtained by burning bones to whiteness—and on this bone ash is placed a set of 40 cupels, or little cups, made of compressed bone ash, and about as big as a florin, and so deep as to hold about half a teaspoonful of water. When the cupels have been in the muffle long enough to become red hot, a piece of lead of about nine times the weight of the gold to be assayed is put into each cupel, taking care not to let any lead fall over, or it would destroy the muffle. So soon as the lead has melted, the paper parcels containing the gold and the silver are placed one in each cupel of melted lead. In a short time the gold and silver melt together, and, as it were, dissolve in the lead; then a kind of circulation of the fluid metallic mixture is observed to take place; and during this circulation, the lead, as it is presented to the surface, meets with the oxygen of the atmosphere, and combining with it, forms oxide of lead, which at that temperature is fluid, and drains into the substance of the cupel, carrying with it the copper and other impurities contained originally in the gold, leaving in the cupel a button which contains all the pure gold and the silver. The cupellation occupies about ten minutes, and at the end of that time the little mass of fused metal is observed to brighten up, the signal by which the assayer knows that the process is finished, and he withdraws the cupels one by one, tilting the fluid globule on one side, that it may incorporate with it any small globule which may chance to be upon the side of the cupel. Some assayers close the doorway of their muffles as the time of the brightening up of the assay approaches, to avoid the access of the atmosphere till the buttons have become solid, because the button of metal—by its silver—absorbs oxygen; and giving off this oxygen at the moment of solidification, spurts or opens, leaving the button hollow or ragged. This precaution is not thought necessary by all assayers, nor, indeed, is this system invariably followed, for some prefer to wrap the gold and silver in the lead foil which is to be used, and do not care to have the precious metals of identical thickness; these assayers use a smaller proportion of silver, viz., 15 parts to 6 of standard gold, or 17·50 to that which they judge to be _fine_, with a view to avoid the danger of “spurting;” they also save time by withdrawing the whole batch of cupels on a tray by means of a peel. The button is taken from the cupel as soon as it has been detached, either by dropping a little water on to it, or by allowing it to cool spontaneously, and it is then hammered out into a strip and annealed. It is next rolled into a ribbon or fillet, and again annealed, after which it is curled up into the form of a letter S. Some prefer to roll the piece into a kind of cushion, but in such a case it is more difficult to remove all the silver; the proper plan is to expose as large a surface as possible to the action of the acid in the next operation.

[Footnote 14: See pages 9, 53.]

[Footnote 15: That is, made by Payen, of Paris.]

The S-shaped fillet is put into a flask of nitric acid of specific gravity 1·23, and the flask is placed in a little cup of brass, which stands over a small gas-burner, while the neck of the flask enters a kind of flue, through which a current of air is continually passing into the chimney of the furnace. The flask being placed in its position, the gas is lighted, and a gentle heat applied, when the nitric acid dissolves out the silver from the S-shaped fillet, and is known to have done its work when red fumes cease to be evolved. The acid is then poured off, and the remaining sponge of metal is washed with distilled water, and boiled with concentrated nitric acid, which removes the remainder of the silver. The sponge of gold is now washed with water, to remove the nitrate of silver, and is then heated to redness in a capsule to render it tough, for in its spongy state it is so rotten that it will not bear to be touched; and although it remains spongy, it is toughened by being heated to redness. It is now called a _cornet_. This spongy state is a consequence of the addition of silver, the presence of which, and its subsequent removal, produce a separation of the particles of the gold. The gold is thus formed into a kind of network or sponge, so that the acid can get at every part of it, and remove any metal which is soluble in the acid. Diluted acid is used first, because if there be any lead left it is dissolved out, and because the action of the strong acid is so violent, that part of it would probably be carried out from the flask. Nitric acid, when boiling, is liable to form bubbles of gas, which expanding, give rise to what is called _bumping_; hence it is usual to put a charred pea into the flask, and this, floating on the surface, causes a more even flow of vapour and gases, and so prevents bumping. It is probable that the charred pea determines the boiling at one particular temperature, for it is found that water, if it be floated in a fluid of a higher boiling point than itself, may be raised considerably above its ordinary boiling point, although remaining quiescent till disturbed by a rod or point. The cornet is next weighed, and as it has been begun under the idea that it was a _thousand_—1,000—all that it weighs short of 1,000 is the alloy which has been removed, the object having been to determine how much pure gold was present in the alloy. There is, however, a source of error in the process which requires to be explained.

It is found to be practically impossible to remove every trace of silver from the cornet; it is therefore necessary to make an allowance, the amount of which is determined by a _proof_. The proof consists of a mixture of gold and silver of known proportions, so that if all the silver be removed from a thousand—1,000—the remaining cornet should weigh exactly 916·6. Four of these proofs are worked with each batch of assays—a batch being 120 assays—under precisely similar circumstances, every precaution being taken that the four shall be equally distributed over different parts of the furnace, &c. Now suppose that the proofs, instead of weighing 916·6, as they would weigh if the whole of the silver were removed, were found each to weigh 916·9; it would be known that 000·3 of silver had been retained, and must be deducted from each of the whole batch of assays. If the assays of coins exceed the limits of from 917·6 to 915·6, they are repeated, as it is assumed that an error has arisen. The convenience of considering the ½ gramme as 1,000 consists in the fact that 1,000 parts of standard gold contain 916·67 of gold and 83·33 of alloy; so that the result having been arrived at without any calculation, one source of error is avoided and time is saved.

Messrs. Johnson, Matthey, and Co. have invented a tray of platinum capsules or thimbles, into which the assay pieces are placed for treatment with acid, instead of into the ordinary glass vessels over gas-burners. The advantages of this invention are too manifest to require elaboration; but saving of acid, gas, labour, and risk of error are amongst the chief of them; and besides, the plan has been in successful operation in their own assay offices in Hatton Garden, as well as in other important assay offices, for several years past. The first cost would seem to be the principal objection to this plan of Messrs. Johnson, Matthey, and Co.; but this is really a small matter, for the apparatus becomes stock in trade, and its cost should be viewed simply as so much capital, whose interest is paid by the saving effected in glass, but more especially by the smaller amount of acid actually employed, while the platinum can at all times be sold for nearly its original cost as old metal.

The mode of operation is as follows:—A stand of slate is so arranged that a means of heating is made to rest on its base. For this purpose a jet of gas is preferred; where, however, gas cannot be obtained, an ordinary oil lamp or a charcoal fire may be used. On a shelf over the source of heat are placed two or three receptacles of platinum, each communicating with a vessel made of porcelain, provided with three necks and an overflow pipe. There is a kind of sieve or tray of platinum, so arranged as to carry from 16 to 100 thimbles of platinum, and provided with a handle, so that this tray, with its charge, can be manipulated at pleasure. The thimbles vary in size according to taste, but each one is cut or slit at the bottom, so that the solution of silver as it is formed may by its density fall out, and allow the clean acid to take its place.

When an operation is to be performed, the tray filled with the charged thimbles—that is, containing the assay pieces—is placed in one of the platinum receptacles or boilers, and heated to a fitting temperature; when the desired effect is produced, the tray is lifted into another receptacle, and again heated; this may be carried to three times if necessary. The products of decomposition of the acid go, with the acid which evaporates, into the porcelain vessel, where the free acid falls through the overflow pipe into a proper chamber, while the acid fumes pass into the flue through the third neck before spoken of. The parted assays having been washed by several immersions in boiling distilled water, without removal from their thimbles, have now to be dried and annealed in a platinum muffle, so formed as to fit into an ordinary muffle, and, after annealing, to be weighed in the usual manner, having saved at least 75 per cent. of the usual trouble.

It is almost needless to add that the system of _proofs_ above described must be also used with this process.

The assay for silver is not so tedious, as it is finished at the point where it leaves the muffle on the cupel; but up to this point it passes through precisely the same process as the gold.

It will have been observed that the event of the process of assaying is to exhibit the proportion of _bullion_ which may be present in a given weight of mixed metals, but that it does not demonstrate the character of the substance which forms the alloy, because this is assumed to be copper or silver. In practice it is found that the alloy may be a mixture of copper or silver with lead, mercury, antimony, arsenic, tin, or zinc, in varying proportions, but whose sum does not exceed the rate per centum in relation to the bullion which is allowed by law. The existence of minute proportions of lead, mercury, antimony, arsenic, tin, or zinc is a matter of great concern, for these, when present, add indefinitely to the difficulties of the coining departments.

Under the present system, complete analyses of bullion imported for coining are a practical impossibility. I would therefore propose to abolish the office of non-resident assayer, and then to appoint two resident assayers, who should conduct their operations in independent laboratories—already existent—and a condition of whose appointment should be an agreement to conduct, when so directed by the Master of the Mint, complete analyses of the bullion submitted for assay. The information thus obtained would enable the melter, by processes subsequently explained,[16] to remove all those substances which are _now_ found to be fatal to the coining of certain bullion. The appointment of officers possessing such ability would be an equal guarantee with the present for the faithful preservation of the integrity of the coinage. This suggestion has been adopted and recommended—so far as regards the resident assayers—by Messrs. C. W. Fremantle and C. Rivers Wilson, in their “Reports on the Mint,” 1870. Indeed two assistant assayers have been recently appointed.

When the assay reports arrive, the Master determines whether the metal has been found within the limits, and if he be satisfied he writes on the assay reports “Passed,” and signs the reports. The Deputy-Master retains the reports as his warrant, and then issues his order for the delivery of the bars by the melter to the rolling room, as recommended in my report dated 29th January, 1859; but, as the accounts are kept by weight, every set of bars is weighed by the officer who receives them into that room before he gives them to his men to work into fillets.

When, in 1856, I took charge of some departments in the Royal Mint, I found that the system of weighing was extremely loose. Officers were plainly told that if they ventured to satisfy themselves as to the weight of bullion they had received—in fact, to determine whether they received what was charged to them—“they would be paid out.” This state of things led men to accept _any_ weight. But I objected and resisted; for I found that the average on each day’s work as received amounted to an habitual minus of five ounces on the weight charged on silver, while on gold it was seldom so little as one ounce. In illustration I will state one case which occurred. I delivered 7920·00 ounces of gold to the Mint Office. When that gold was received, the official weigher gave me credit for 7918·15 ounces, which was a deduction of no less than 1·85 ounces. I appealed to the Master, who by written order directed the gold, the weight of which was thus disputed, to be weighed by Mr. Pilcher, the officer of the weighing room, in the presence of witnesses whom the Master nominated. Mr. Pilcher complied with this order, and gave a certificate, signed by the deputed witnesses, in which he stated that the gold under dispute weighed 7919·98 ounces—that is, that it differed in weight as charged by me to the Mint Office 9·60 grains instead of 888·00 grains, the difference which the official weigher had deducted from the bullion I had delivered to him. The custom had been to weigh silver to 0·50 ounce, and gold to a pennyweight; but I introduced the system of weighing silver to 0·10 ounce, and gold to 0·01 ounce, and at the same time induced the Master to order new balances of superior construction for the coining department, and one specially devised by Mr. James M. Napier for use in the Mint Office. Of this balance, received after having been ordered for some years, it will be sufficient to say that it appears to be extremely accurate when properly used, and is then capable of great results; but, unfortunately, the system is such that those who are admittedly unfit may be promoted to important posts, just because they are senior, and not because of superior fitness for the work. This fine balance, therefore, becomes equal to a good rifle in the hands of a bad marksman. Great accuracy having been enforced, the weighing is now improved, but is still far from perfect, because the officers are compelled to abide by the decision of the weigher at the Mint Office, who, as beforesaid, may be unfit for his office, while the officers themselves are debarred from all checks, by the removal from the coining department of the set of standard weights, which were bought on purpose that those gentlemen might check their own weights as to accuracy. The Master’s order for the removal of these weights was in the following terms:—

[Footnote 16: See pages 81-82.]

“The standard weights hitherto kept in the weighing room
are permanently transferred to the Mint Office. A new set
of standard weights to be made for the Mint Office. The
coining department weights to be examined by Mr. ——
——, and compared with those in the Mint Office (after
these have been corrected), and reported on by him. Such
examinations to be repeated every six weeks.”

It should be observed that this order was preceded by one written on the 2nd February, 1865, which directed that the Mint Office weigher should be the final judge, but that any officer might demand the re-weighing of any bullion, while the second weighing was to be final. All appeals to pass through Mr. John Graham, the order proceeding to direct that—

“The final decision is not to be called in question by any
other officer of the coining department.”

The latter determination, I was informed, was intended to apply to me personally. Its immediate effect was a deduction of 0·25 ounce from the first gold I delivered, and a consequent alteration of my book to that extent by Mr. John Graham, who made a note in writing as follows:—

“This difference (error) may be accounted for by the
acknowledged error of three grains heavy of the Mint Office
500 oz. weight. See Œrtling’s Report.
(Signed) “J. G.”

Such being the circumstances, I submitted; but the result could be in one direction only, for it had long been maintained that the reported losses were but “an hallucination,” and were, in fact, a “mere difference of weighing.”

It is true that the weights are periodically adjusted, but there is adjusting and adjusting. These facts have been dwelt upon because they have a most important bearing on the subject of loss and gain by coining, to be afterwards treated of.

Between 1856 and 1866 the old scales were removed, to make room for the superior balances of Messrs. De Grave, Short, and Fanner, the eminent scale-makers of St. Martin’s-le-Grand, of whose balances it is impossible to speak too highly, and of whom it is but fair to state that the Prize Medal was awarded to them at the International Exhibition, 1862, for their superior workmanship in balances.

Since it is necessary for the officer in charge of any department to ascertain the exact amount he gives to his workmen, and to satisfy the Master that they have returned—to within the limit of the weight of one blank of whatever denomination of coin he may be working—the bullion which they received, it became imperative to select the best balance; and in practice that made by Mr. S. R. Short, of the firm above named, was found to be the most serviceable. Mr. Short has introduced minor improvements from time to time, as experience has dictated, but the balance about to be described, after years of wear, determines to within one grain when charged with 1,200 ounces troy. The mode of proving the accuracy of a balance is to weigh as usual, and having arrived at a just determination of the weight of the matter to be weighed, to change the weights to the pan in which the matter has been weighed, and to place the matter in the pan previously occupied by the weights, so that both are now made dependent from the reverse ends of the beam; if the results be identical with the former results there can be no doubt as to the accuracy of the beam; should there be any variation, the balance must be adjusted by minute alterations of the knife-edges at its ends.

Mr. Napier has made his balance after the principle of Mr. Cotton’s balance, so that it requires no adjusting screws, but I have no personal experience of the benefits of this omission.

BY Mʳ. S. R. SHORT, JUNʳ.]

The balance used in the rolling room of the Royal Mint is specially adapted for the purpose. Such a balance, as already noticed, must be accurate, as must also all the balances used in the Mint; but this balance requires an arrangement by which the pans can be released from the beam and held firmly while being loaded. Mr. S. R. Short has been peculiarly happy in effecting this object, as will be seen by reference to the steel engraving, which exhibits a fine illustration of Mr. Short’s balance. The raising of the handle A causes the partial revolution of the cam B—represented by dotted lines—and this, as it diminishes, releases the lever C, which, by the rods D acting on the point E, permits the approach of the joints on the rods F towards the centre. The rods F work on a centre fixed beneath the table on which the balance stands, so that the closing of their one end causes the opening of the other. Thus the claws G recede from the wedges H, upon which they had been fastened, and so release the pan, for the claws act simply as a man’s hand would act in fixing the pan while it is being loaded. Immediately upon the opening of the claws another action takes place by the continued movement of the handle A, which causes the cam I to rotate upon the friction roller J, affixed to a lever proceeding from the shaft K, which at L is cut with eccentric bearings, so that by the partial revolution of K, caused by the pressure of I upon J, the series of supports M sink downwards from the pans N, thus obliging the pans, with the rods which support them, to be suspended from the supports S, which rest on the frame R. So soon as the pans are thus suspended, the continued motion of the handle A causes the cam O to permit the falling of the lever P, which at Q supports upon a friction roller the iron framework R. This framework terminates upwards at the points near S in sugar-loaf cones of gun-metal, which are intended to relieve the knife-edges of the beam from the pressure of their planes while the beam T is out of use, or while it is being loaded, for at these times the knife-edges would become seriously damaged. By the lowering of the frame R the centre knife-edge U is permitted to rest on its plane of steel _c_, which is beautifully bedded on an arch of gun-metal, as will be seen if the enlarged representation of this part, and which is shown to the right, be examined. In fact, the steel plane _c_ is secured by wedges exactly shaped to it, but for the sake of illustration these wedges have been replaced in the enlarged portion of the engraving by upright pillars shown against _c_, with screws travelling through them, and intended to fix _c_ rigidly in a given position. At the same instant that U comes into rest upon _c_, the knife-edges V at the ends of the beam T receive the planes W, so that the beam has now to support the weight of the matter placed in the scale-pans, and at this time is determined the actual weight of whatever that matter may be. The weigher, by an indicator suspended from the beam at X, reads the oscillations of the beam on an ivory plate at Y. At the time of weighing, the beam with its dependents is carried by the four-legged frame of iron _b_. The frame R passes through friction rollers of brass at _d_, which rollers maintain it in its relative position to the beam. The weighing having been determined, the handle A is reversed, and the beam is placed at rest; but the vibration of the machinery in the rolling room causes a continuous chatter of the knife-edges V against the planes W, therefore at each end of the beam a small screw Z is made just to touch the lower edge of the beam when it is at rest. Final adjustments which are very minute may be arranged by _f_; while the box _e_ on the pan against G is intended to receive small pieces of lead, which are used to compensate for the loss by the wearing of the pans. The shaft K is supported from the table by _a_. The beam of the balance represented in the steel engraving measures 48·00 inches from knife-edge to knife-edge; that is to say, it measures 24·00 inches from the centre knife-edge to the knife-edge at either end. Mr. S. R. Short has made for the rolling room another balance for weighing silver, and as it differs in some important details, it is deemed wise to demonstrate those alterations in Fig. 3, where the supporting frame R will be seen to carry three upright pillars S¹, S², S³. The pillars S¹ and S³ are represented as being placed on the farther side of the beam, which at those parts is broken away, that the pieces of steel against which the pillars are made to touch may be seen. S² touches against one of these pieces of steel near the centre knife-edge U, and in front of the beam. The pillar S² terminates in a point, which passes into an inverted cone, while the point of S³ is made to pass into an inverted =V=, so that when the beam is raised by these pillars from its centre knife-edge, S² obliges it to take a specific position with regard to its distance from the centre of the plane on which its centre knife-edge rests, and S³ causes it to take a perfectly parallel line. S¹ is simply a plane which rises against a plane and steadies the other end of the beam, which by this arrangement stands as on a tripod, and is, therefore, prevented from moving, and in addition has its knife-edge invariably placed upon the same part of the plane at each time of weighing, while the frame R, by its continued lowering, drops the planes with equal accuracy on the end knife-edges V.

The sovereign bars having been weighed by the officer, and given by him to his men, are wrought in sets of twenty; each set is called a _batch_, and each bar in the batch undergoes precisely the same process. The bar is passed into the opening A of the breaking-down mill Fig. 4, where it receives a considerable compression, for the rollers B, seizing its end, drag it forward, while they roll back and retard the progress of that part of the bar which is not between them. The result is that the bar is lengthened, but not widened materially, so that length is gained at the expense of the thickness, which is regulated by the distance between the rollers. The rollers are driven by shafts and adjusting couplings C, which are themselves driven by the wheels D. The distance between the rollers is determined by the action of the lever E′, which, by the endless wheels on its axle at E fitting into geared wheels, gives motion to powerful screws shown at F, which terminate in cups on the upper part of the upper brasses of the rollers B, as may be seen at G. The upper brasses are kept always against the ends of the screws by weights which are beneath the mill, but from which levers and rods terminate at the lower part of the upper brasses, at about the position indicated by I, so that the upper roller has motion either upwards or downwards at pleasure, but the motion upwards is arrested by the powerful screws F, and this point once determined by the reading of the scale H, is fixed by the clamp J. The thickness, therefore, of each bar in a batch is determined within certain limits; and when each bar in the batch has been rolled, the mill is altered, refixed, and again the rolling goes on till each bar has passed seven times at varying pinches through these rollers. Owing to the wear of the moulds in which the bars are cast—and which is largely due to the presence of minute portions of antimony in the gold—the bars are never of uniform thickness; hence bars of every denomination are passed through the rollers on their edge, so as to reduce them to one uniform thickness, otherwise the fillets resulting would be ragged, and of unequal widths, which defect would cause them to produce blanks out of _remedy_[17] as to weight. The rollers are set face to face; the graduated scale H is then fixed at zero; if now the rollers are separated until there be a space of 1·00 inch between them, the scale H will indicate 50; if, however, it be desired to read a higher figure, part of another revolution must be performed, causing the scale to read, say 31·50 of that revolution, when the rollers will be so far apart that sovereign bars which pass between them will, allowing for the expansion after compression, be found to be exactly 1·375 inches wide.

[Footnote 17: See p. 39.]

If now these bars be rolled on their sides, the mill must be turned down till the scale reads 45, when, if the bars be passed through, they will measure as under for width and thickness at each successive pinch. A _spring-pinch_ means the passing the bar once more through the mill without altering the distance between the rollers.

A sovereign bar passed Becomes a fillet or ribbon—
through the rolling mill— Inches wide. Inches thick.
At 45·00 1·450 0·775
” 35·00 1·540 0·610
” 26·00 1·610 0·460
” 19·00 1·665 0·335
” 14·00 1·696 0·250
” 10·50 1·712 0·194
” 8·50 1·765 0·148
” 7·00 1·778 0·129
” 1st spring-pinch. 1·778 0·127
” 2nd ” ” 1·778 0·120
” 3rd ” ” 1·779 0·118
” 4th ” ” 1·781 0·117

When the bars have passed through the rollers at 10·50 they have become of great hardness, and of considerable length, say 6 or 8 feet. They are then taken to the shears K, where their hollow ends are cut off, and the bars cut into lengths of 18 inches; or rather, such was the practice when it was considered wise to obtain as much coined money as was possible from the bars; but recently, under other considerations, the old plan of the moneyers has been re-introduced, that of shearing the hollow ends from the bars before rolling. In this process there is needless waste, and it is to be hoped that the proper practice will be reverted to, for under that system the ends averaged 4·26 per cent., while under the re-introduced one it amounts to 7·03[18] per cent. on the bars, and for the following reason. When the metal is poured into the moulds, it almost immediately solidifies, and while solidification is going on contracts in volume, leaving a kind of cup of metal or hollow part on the top of the bar. In the act of rolling, the bar maintains an equable width until this hollow part is reached, when it suddenly expands, and at this point the workman shears off the defective part of the bar, whereas in the other case he shears off the hollow end until his eye fails to discover the effects of the shrinkage; hence the loss of 2·77 per cent. on the produce of fillets, as well as the loss in wages for melting the bars. The shears may be regarded as large scissors driven by a drum on the same shaft that carries the driving wheel for D. The drum is eccentric, so that at each revolution the shears are caused to open and shut. The bar being placed between the jaws K, the long end L of the shears is raised by the drum, and the piece is cut off. The shears may be closed and the end of L suspended, when out of use, by a hook on the end of the screw worked by the lever M. The length at which a bar is to be cut off is regulated by the gauge shown at N.

[Footnote 18: See page 112.]

The sheared bars are placed in copper tubes, the tops of which are luted on with clay. It is imperative that the copper tubes should be made without solder, because this fuses at a temperature below that which is required to anneal the gold; if present it would run down upon the hot gold, and cause it to fuse and alloy with the solder, thus spoiling the work and entailing expense upon the coiner. The tubes which are used in the Royal Mint are made by Messrs. Benhams and Froud, of Chandos Street, Strand, who, after considerable pains, arrived at a method of making the tubes in such a manner as to entirely satisfy the requirements of the Mint. The tubes A are placed on an iron carriage B, which is then run into the furnace, as shown in Fig. 5. The door of the furnace C is closed by raising the counterpoise D; the heat of the furnace is regulated by the damper E. The apron F is sometimes of use in annealing very long silver bars. After remaining in this furnace for twenty minutes, the carriage is withdrawn, and the tubes, taken with tongs, are plunged into cold water, to cool the gold as rapidly as possible. The rapid cooling of gold and silver gives to each metal a peculiar character, which is of value in the after processes, and prevents the access of the atmosphere, which, in prolonged cooling, would cause the oxidation and consequent removal of so much copper that the alloyed metal would become too rich in gold for circulation as coin. The annealing of some metals is effected not so much by the continued heat as by the slow cooling; it is therefore wise to raise the metal to its full heat as rapidly as possible, and then so to arrange matters that it may cool very slowly. This method does not hold good in the case of the precious metals and of copper, for they become, under such treatment, so soft, malleable, and pasty as to stick to the machinery, and thus to cause considerable trouble and loss. After annealing, the bars, which are now called _fillets_, go again to the breaking-down mill, through which they are passed with the scale indicating 8·50, then at 7·00, and after this are submitted to another pinch without altering the scale at all; so that what is called a _spring-pinch_ is given, with the intention of effecting the reduction of the fillet to one uniform thickness, for the breaking down and subsequent rolling cause the fillets to become much thicker in their middle than at their sides. The spring-pinches reduce this, while at the same time they diminish the elasticity of the metal, and fit it for the other mills. It will be seen, by reference to the above table, that the widening of the fillet is very trifling; but width may be gained to any desired extent, at the will of the workman, if the bars be submitted to a heavy pinch instead of a series of light ones. The fillet having been submitted to the fourth spring-pinch, is gauged on its side by a steel instrument, of which Fig. 6 is a representation. It is a hollow wedge, which is graduated to the thousandth of an inch. Supposing that the opening from A to B were extended until it were one inch wide at A, the space would be divided between it and B into 1,000 parts, and then every fillet passed into this opening would stop at a given point, say, for instance, at 140; such being the case, every part of the same fillet should be arrested precisely at the same point. The fillets are reduced till they measure 117 on this gauge, and are consequently 0·117 inches thick. They are then passed to the next mill, where they receive four light pinches, and then to a third mill, where they receive two more very light pinches, and by means of these six pinches are reduced in thickness to 0·075 inches. They then pass to another mill, still finer than any of the preceding, and here are submitted to four very light pinches, by which they are reduced to 0·058 inches, and are finally finished at the sixth or gauging mill, where they receive three pinches, and are then 0·053 inches thick by 1·829 wide.

The gauging mill is of different construction from the other mills, as may be seen by reference to Fig. 7, where the rollers A are seen in the act of reducing a fillet. The upper roller is fixed in brasses loosely clamped together; the upper brass, B, being firmly bolted to the main frame of the mill by the screws _c_, while the lower one C, which carries the weight of the roller when it is running empty, is supported by spiral springs shown at D. The lower roller works on a brass, E, which rests on a wedge shown at F; the brass being cut to fit the wedge, so that it may become similar to a solid mass, irrespective of any motion given to the wedge. By this mode of adjusting, a difference of the 0·001 of an inch may be made with ease between the distance of the rollers, and, consequently, in the thickness of fillets which may pass between them. The wedge F is moved forward and backward by the screw G, which itself has motion from the gear work H, by the handle I. Directly a fillet is passed between the rollers, the topmost one is forced against its upper brasses, and further upward motion becomes impossible. The weight of metal in this roller gives rise to irregularities in the thickness of the fillets which pass from this mill, causing much trouble and some expense; therefore it is proposed to support the upper roller by a similar arrangement to that which affects the lower roller, and to keep it rigidly against the upper brass, but with just so much pressure as its own weight would induce were circumstances reversed, and by these means to relieve the fillet from the weight of the roller, because that has an undue influence on each end of it. The gauger of fillets requires other tests besides that of the thickness of the edge of the fillet, so he punches out a blank from an occasional fillet by a hand-press, the cutter of which is shown at J, Fig. 7, worked by the handle K, through the screw L. The blank falling through the bolster of the cutter is caught at M, and is then weighed in the hand scales N, against a standard weight, from which it must not vary more than 0·50 grain. O, Fig. 7, is the gauge actually used by the workman; Fig. 6 represents the standard gauge used only by the officer in charge to check the work at its various stages. He has in addition a gauge of great accuracy, by which to measure the fillets at any point, as to width and thickness. This gauge will be more intelligible by reference to Fig. 8. A is the handle, which is hollow; B is a lever attached to the flat rod of copper C, which at D is cut with a rack, into which a pinion E is made to work. The pinion E works on a shaft, the upper end of which carries a hand F, provided with a vernier G. If now the handle A be firmly held by the hand, while the thumb be made to press the lever B towards the end of the handle, the rod C is set in motion, and causes the hand G to travel in the direction of H. The rod C rests on another rod I, made of steel, and so long as to pass into the handle of the instrument. The ends of the rods C and I are fitted at _b_ with steel shoulders, and are then continued, as represented, to _a_. If it be desired to measure the thickness of a fillet, the points _a a_ and _b b_ are caused to open by pressure applied to B, and the fillet is placed between the points _a_, when a spring fixed in the box K brings back the rod C as soon as B is gently released, and encloses the fillet. The separation of the points _a_ by the fillet causes the hand or indicator G to stand at a point from zero, which is then read. The scale is divided into 500 parts; and if the points be opened 0·50 inch, the hand makes one revolution; so that the ·001 of an inch is gained by one reading. But each 0·001 is subdivided by the vernier into ten, so that a ten-thousandth part of an inch is read without trouble. To measure the diameter of a blank coin, or the width of a fillet, it must be placed between the points _b_; but since the extreme graduation of this gauge is 0·50 inch, it is necessary, if it be desired to measure a larger diameter, to press back the lever B till the zero of the vernier G reaches 0·500 on the scale H, and hold it there while a clamp is made fast at the spot indicated by the star (*), to prevent the motion of C without I. When the clamp is fixed the rod I must be drawn out till the zero of the vernier reaches that of the scale H, when the screws J must be tightened to retain I in its new position, with half an inch permanent opening between the points _b_ and between the points _a_. In a new measurement, that permanent 0·500 must be added to the reading. This arrangement admits of measuring up to 3·5000 inches, to which limit the gauge is extremely accurate. The instrument was invented by myself, because I found it difficult to convince the men that the fillet was thickest in its middle, and consequently heavier there than it should be; and, unfortunately, that the workmen habituated themselves to attributing this fault to each other, when its existence was proved. The fact was, that under the system which had prevailed, the men—with a view to make bad work for a specially-designed reason—would set the upper rollers at varying angles, so that a fillet at one mill would have one thin and one thick edge; and when that fillet passed through the next mill, the angle of the roller being altered, would make both edges alike, but the effect of the manœuvre was to push the metal into the middle of the fillet, and thus to unfit it for the draw-bench. Hence the necessity to fix where the blame should rest, and the production of this instrument, under my direction, by Mr. C. Becker, of 30, Strand, at once overcame all those difficulties.

The fillets are weighed from the rolling room to the drag room, where they are finally adjusted; for with every energy, discretion, and skill, FILLETS CANNOT BE OBTAINED OF UNIFORM THICKNESS BY SIMPLE ROLLING. In the drag room the fillets are taken to the small shears, Fig. 9, by which one end of each fillet is trimmed so as to render it square. The plates A are fixed to the head of a T-shaped lever, which is caused to oscillate by a cam beneath the floor. The plates A shut against a face of steel fixed to a block, and held by the screws shown at B; if therefore the end of the fillet be passed between the plate A and the face of B, each oscillation causes the cutting off of so much as protrudes, the pieces cut off falling into the box C, which has now been enlarged so far as to form a pan all round the top with a view to catch all of them. D forms part of a chain, by which the shears are thrown out of motion. The fillets, having been trimmed so as to render their ends square, are next passed to the extent of about two inches between the rollers of a flatting mill, shown in Fig. 10, which reduce that part of the fillet to about two-thirds its thickness. A A represent a pair of small rollers, the upper one of which is cut with three flat faces, so that it has three rounding and three flat surfaces; hence, when both rollers are revolving, there are spaces with openings between them; but when the rounding faces come down, those openings are much narrowed, so that any fillet placed between them becomes thinned to just such an extent as may be deemed necessary. The rollers travel in opposite directions, so as to cause the expulsion of a fillet placed between them. The reverse motion is gained as follows:—The upper roller is driven by B, which receives its motion from the little pinion C, carried on the shaft which also supports D. D reverses the motion of E, which is driven from the drum F. E also drives G, which gives motion to the lower roller. The fly-wheel H is borne at the extreme end of the shaft which carries F and E. The fillets are rested on J while being _flatted_, and are, after flatting, placed in the trough K, from which they are taken to a rolling mill in the drag room, of precisely the same construction as that exhibited at Fig. 7, to be passed twice through at equal pinches, with a view to render them still more accurate than they were when leaving the rolling room, as well as to reduce them to the exact thickness at which the trier has found they will produce the best work at the draw-bench. The rolling mill in the drag room was provided with steel rollers. Steel rollers are of somewhat recent invention, and seem to have received a high character from those whose opinion may be modified by further experiments more accurately made. My opinion, founded on experience, is to the effect that they are not worth their extra cost; but that their usefulness may be more fully developed when they shall have been fitted with the arrangements proposed for the gauging mills before explained.[19] After these alterations have been made, it may fairly be questioned whether steel rollers will, under circumstances every way similar, produce better work than is produced by the ordinary chilled-iron rollers; in other words, I believe that steel is not a better substance for rollers than chilled cast-iron. Those who have to sell, and those who have to use, have, of course, different motives; he who has to use a machine should judge calmly, and not be led away because the invention is new. It is amongst these considerations that I am convinced that steel rollers do not save money; for if they wear longer without getting out of order, they also require a longer time to put them again in order. Rollers, made from whatsoever substance, cannot reasonably be expected to produce a fillet from every part of which blanks of equal weight can be struck, because it is not yet possible to produce a compound of equal hardness throughout; but if the construction of the mill be altered, steel rollers may approach nearer to that perfection which is gained by the draw-bench, but they can never replace it. The fillets are taken from the mill to the draw-bench.

[Footnote 19: See page 23.]