Chapter II: Part 2

The adjusting of the matrix to the mould is technically called "fitting," and requires great skill. If type are cast from unfitted matrices, be the letters ever so cleverly designed and perfectly cut, when assembled in the printed page they will present a very ragged appearance. Some letters will appear slanting backward, others forward, some be above the line, others below; some will perforate the paper, while others will not print at all; the distances between the letters will everywhere be unequal, and some will print on but one edge. Indeed, a single letter may have half of these faults, but when the matrices are properly fitted, the printed page presents a smooth and even appearance.

The mould for this purpose is made of hardened steel, and in it is formed the body of the type. The printing end is formed in the matrix. The mould is provided at one end with guides and devices for holding the matrix snugly against it while the type is being cast, and for withdrawing the matrix and opening the mould when the type is discharged. At the opposite end from the matrix is an opening through which the melted metal enters. The moulds are made adjustable so that each character is cast the proper width, the opening of course being wider for a "W" than for an "i." Only one mould is necessary for one size of type, and with it all the matrices for that size may be used. Commercially, however, it is often necessary to make several moulds of the same size in order to produce the requisite amount of type.

After the adjustments are made, the casting of the type follows. Type are now cast in a machine which is automatic, after it is once adjusted to cast a given letter. The melted type metal is forced by a pump into the mould and the matrix, and when solidified, the type is ejected from the mould and moved between knives which trim all four sides. The type are delivered side by side on a specially grooved piece of wood, three feet long, called a "stick," on which they are removed from the machine for inspection. Type are cast at the rate of from ten to two hundred per minute, according to the size, the speed being limited only by the time it takes the metal to solidify. To accelerate this, a stream of cold water is forced through passages surrounding the mould, and a jet of cold air is blown against the outside.

The automatic casting machine performs six different operations. Formerly, all of them, except the casting itself, were done by hand, and each type was handled separately, except in the operation of dressing, or the final finishing, where they were handled in lines of about three feet in length.

After the type have been delivered to the inspector, they are examined under a magnifying glass and all imperfect type are thrown out. The perfect type are then delivered to "fonting" room, where they are weighed, counted, and put up in suitable packages in proper proportion of one letter with another, ready for the printer.

Formerly the various sizes of type were indicated by names which had developed with the history of type making. It was a source of considerable annoyance to printers that these old standards were not accurate, and that two types of supposedly the same size, and sold under the same name, by different makers, varied so much that they could not be used side by side. Of recent years the "point" system, by which each size bears a proportionate relation to every other size, has done much to remedy this trouble, and now nearly all type is made on that basis. An American point is practically one seventy-second of an inch. Actually it is .013837 inch. It was based on the pica size most extensively in use in this country. This pica was divided into twelve equal parts and each part called a point. All the other sizes were made to conform to multiples of this point. The point is so near a seventy-second of an inch that printers frequently calculate the length of the pages by counting the lines, the basis being twelve lines of 6 point, nine lines of 8 point, eight lines of 9 point, and six lines of 12 point to the inch. This calculation is really quite accurate.

The following table will show the old and new names for the various sizes:--

3-1/2 Point, Brilliant.
4-1/2 Point, Diamond.
5 Point, Pearl.
5-1/2 Point, Agate.
6 Point, Nonpareil.
7 Point, Minion.
8 Point, Brevier.
9 Point, Bourgeois.
10 Point, Long Primer.
11 Point, Small Pica.
12 Point, Pica.
14 Point, 2-line Minion or English.
16 Point, 2-line Brevier.
18 Point, Great Primer.
20 Point, 2-line Long Primer or Paragon.
22 Point, 2-line Small Pica.
24 Point, 2-line Pica.
28 Point, 2-line English.
30 Point, 5-line Nonpareil.
32 Point, 4-line Brevier.
36 Point, 2-line Great Primer.
40 Point, Double Paragon.
42 Point, 7-line Nonpareil.
44 Point, 4-line Small Pica or Canon.
48 Point, 4-line Pica.
54 Point, 9-line Nonpareil.
60 Point, 5-line Pica.
72 Point, 6-line Pica.

HAND COMPOSITION AND ELECTROTYPING

By J. Stearns Cushing.

The form of the book, the size of the type page, and the size and style of the type having been determined, the manuscript is handed to the foreman of the composing room, with all the collected directions in regard to it. He fills out a scheme of the work which tells the whole story,--somewhat as shown in illustration opposite page 42.

Under the heading "Remarks," in the scheme shown, are noted general directions as to capitalization, punctuation, and spelling (whether Webster, Worcester, or English spelling--which means generally not much more than the insertion of the "u" in words like "favor," "honor," etc., and the use of "s" instead of "z" in words like "recognize," "authorize," etc.). Sometimes these directions are given by the publisher, sometimes by the author, but more often by the superintendent or foreman of the printing-office. The office generally has a fairly well established system, which is followed in the absence of other orders. It is rarely the case that it is not the wisest course, if one is dealing with a reputable firm of printers, to leave all such details, except deciding the dictionary to be followed, to them. It is their business, and they will, if allowed, pursue a consistent and uniform plan, whereas few authors and fewer publishers are able, or take the pains, to do this. Too often the author has a few peculiar ideas as to punctuation or capitalization, which he introduces just frequently enough to upset the consistent plan of the printer. He will neither leave the responsibility to the latter nor will he assume it himself, and the natural result is a lack of uniformity which might have been avoided if the printer had been allowed to guide this part of the work without interference.

The compositors who are to set the type are selected according to the difficulty of the matter in hand, and each one is given a few pages of the "copy," or manuscript. The portion thus given each compositor is called a "take," and its length is determined by circumstances. For instance, if time is an object, small takes are given, in order that the next step in the forwarding of the work may be started promptly and without the delay which would be occasioned by waiting for the compositor to set up a longer take.

When the compositor has finished his take, the copy and type are passed to a boy, who "locks up" the type on the galley--a flat brass tray with upright sides on which the compositor has placed his type--and takes a proof of it upon a galley-or "roller"-press. This is the proof known as a "galley-proof," and is, in book work, printed on a strip of paper about 7 x 25 inches in size, leaving room for a generous margin to accommodate proof-readers' and authors' corrections, alterations, or additions.

Date: ____
Name and Address of Author: ____
Name and Address of Publisher: ____
Uniform with ____
Size of Page: ____
Type,--Old Style or Modern face: ____
Text in ____ leaded with ____
Foot-notes ____ in leaded with ____
Extract in ____ leaded with ____
Other Types: ____
Running Titles in ____
Left-hand Running Title: ____
Right-hand Running Title: ____

PROOFS to be sent as follows:

1st Rev. and Copy to ____
2d Rev. and Old Rev. to ____

(Put Changes of Orders as to Proofs in this column.)

F. Proofs: ____
When begun: ____ When to be completed: ____
REMARKS. ____]

The galley-proof, with the corresponding copy, is then handed to the proof-reader, who is assisted by a "copy-holder" (an assistant who reads the copy aloud) in comparing it with the manuscript and marking typographical errors and departures from copy on its margin. Thence the proof passes back again to the compositor, who corrects the type in accordance with the proof-reader's markings. Opposite page 44 is a specimen of a page proof before correction and after the changes indicated have been made.

New proofs are taken of the corrected galley, and these are revised by a proof-reader in order to be sure that the compositor has made all the corrections marked and to mark anew any he may have overlooked or wrongly altered. If many such occur, the proof is again passed to the compositor for further correction and the taking of fresh proofs. The reviser having found the proof reasonably correct, and having marked on its margin any noticed errors remaining, and also having "Queried" to the author any doubtful points to which it is desirable that the latter's attention should be drawn, the proof--known as the "first revise"--and the manuscript are sent to the author for his reading and correction or alteration.[2]

[Footnote 2: If the book is to be illustrated, the
author or publisher should be particular to
indicate the position of all cuts by pasting proofs
of them on the margin of the galley-proofs nearest
the place desired. The time occupied by the
"make-up" in "overrunning" matter for the insertion
of cuts is charged as "author's time," and they can
be inserted at less expense in the galley-proofs
while making-up the type into pages than at any
other time. All alterations, so far as practicable,
for the same reason, should also be made in the
galley-proofs, especially those which involve an
increase or decrease in the amount of matter, since
changes of this nature made in the page-proof
necessitate the added expense of a rearrangement of
the made-up pages of type.]

On the return of the galley-proofs to the printer, the changes indicated on the margins are made by compositors selected for the purpose, and the galleys of type and the proofs are then turned over by them to the "make-up." The "make-up" inserts the cuts, divides the matter into page lengths, and adds the running titles and folios at the heads of the pages.

At this stage the separate types composing the page are held in place and together by strong twine called "page cord," which is wound around the whole page several times, the end being so tucked in at the corner as to prevent its becoming unfastened prematurely. The page thus held together is quite secure against being "pied" if proper care is exercised in handling it, and it can be put on a hand-press and excellent proofs readily taken from it. A loosely tied page, however, may allow the letters to spread apart at the ends of the lines, or the type to get "off its feet," or may show lines slightly curved or letters out of alignment. The proof of a page displaying such conditions often causes the author, unlearned in printers' methods, much perturbation of mind and unnecessary fear that his book is going to be printed with these defects. These should in reality be no cause for worry, since by a later operation, that of "locking-up" the "form" in which the pages will be placed before they are sent to the electrotyping department, the types readily and correctly adjust themselves.

Proofs of these twine-bound pages are taken on a hand-press, passed to the reviser for comparison with the galley-proofs returned by the author, and if the latter has expressed a wish to see a second revise of the proofs, they are again sent to him. For such a "second revise" and any further revises an extra charge is made. The proofs to which an author is regularly entitled are a duplicate set of the first revise, a duplicate set of "F"-proofs,--to be mentioned later,--and one set of proofs of the electrotype plates; though it may be added that the last is not at all essential and is seldom called for.

Usually the author does not require to see another proof after the second revise, which he returns to the printer with his final changes and the direction that the pages may be "corrected and cast," that is, put into the permanent form of electrotype plates. Some authors, however, will ask to see and will make alterations in revise after revise, even to the sixth or seventh, and could probably find something to change in several more if the patience or pocketbook of the publisher would permit it. All the expense of overhauling, correcting, and taking additional proofs of the pages is charged by the printer as "author's time." It is possible for an author to make comparatively few and simple changes each time he receives a new revise, but yet have a much larger bill for author's changes than another who makes twice or thrice as many alterations at one time on the galley-proof, and only requires another proof in order that he may verify the correctness of the printer's work. The moral is obvious.

After the pages have been cast, further alterations, while entirely possible, are quite expensive and necessarily more or less injurious to the plates.

The author having given the word to "cast," the pages of type are laid on a smooth, level table of iron or marble called an "imposing stone." They are then enclosed--either two or three or four pages together, according to their size--in iron frames called "chases," in which they are squarely and securely "locked up," the type having first been levelled down by light blows of a mallet on a block of smooth, hard wood called a "planer." This locking-up of the pages in iron frames naturally corrects the defects noted in the twine-bound pages, and not only brings the type into proper alignment and adjustment, but prevents the probability of types becoming displaced or new errors occurring through types dropping out of the page and being wrongly replaced.

When the locking-up process is completed, the iron chase and type embraced by it is called a "form." A proof of this form is read and examined by a proof-reader with the utmost care, with a view to eliminating any remaining errors or defective types or badly adjusted lines, and to making the pages as nearly typographically perfect as possible. It is surprising how many glaring errors, which have eluded all readers up to this time, are discovered by the practised eye of the final proof-reader.

The form having received this most careful final reading, the proof is passed back to the "stone-hands"--those who lock up and correct the forms--for final correction and adjustment, after which several more sets of proofs are taken, called "F"-proofs (variously and correctly understood as standing for "final," "file," or "foundry" proofs). A set of F-proofs is sent to the author to keep on file, occasionally one is sent to the publisher, and one set is always retained in the proof-room of the printing-office. These proofs are characterized by heavy black borders which enclose each page, and which frequently render nervous authors apprehensive lest their books are to appear in this funereal livery. These black borders are the prints of the "guard-lines," which, rising to the level of the type, form a protection to the pages and the plates in their progress through the electrotyping department; but before the plates are finished up and made ready for the pressroom, the guard-lines, which have been moulded with the type, are removed.

After several sets of F-proofs have been taken, the form is carried to the moulding or "battery" room of the electrotyping department, where it leaves its perfect impress in the receptive wax. Thence it will later be returned to the composing room and taken apart and the type distributed, soon to be again set up in new combinations of letters and words. The little types making a page of verse to-day may do duty to-morrow in a page of a text-book in the higher mathematics.

After the type form has been warmed by placing it upon a steam table, an impression of it is taken in a composition resembling wax which is spread upon a metal slab to the thickness of about one-twelfth of an inch. Both the surface of the type and of the wax are thoroughly coated with plumbago or black lead, which serves as a lubricant to prevent the wax from adhering to the type.

As the blank places in the form would not provide sufficient depth in the plate, it is necessary to build them up in the wax mould by dropping more melted wax in such places to a height corresponding to the depth required in the plate, which is, of course, the reverse of the mould, and will show corresponding depressions wherever the mould has raised parts. If great care is not taken in this operation of "building-up," wax is apt to flow over into depressions in the mould, thereby effacing from it a part of the impression, and the plate appears later without the letters or words thus unintentionally blotted out. The reviser of the plate-proofs must watch carefully for such cases.

The mould is now thoroughly brushed over again with a better quality of black lead than before, and this furnishes the necessary metallic surface without which the copper would not deposit. Then it is "stopped out" by going over its edges with a hot iron, which melts the wax, destroys the black-lead coating, and confines the deposit of copper to its face.

After carefully clearing the face of the mould of all extraneous matter by a stream of water from a force-pump, it is washed with a solution of iron filings and blue vitriol which forms a primary copper facing. It is then suspended by a copper-connecting strip in a bath containing a solution of sulphate of copper, water, and sulphuric acid. Through the instrumentality of this solution, and the action of a current of electricity from a dynamo, copper particles separate from sheets of copper (called "anodes," which are also suspended in the bath) and deposit into the face of the mould, thus exactly reproducing the elevations and depressions of the form of type or illustrations of which the mould is an impression. After remaining in the bath about two hours, when the deposit of copper should be about as thick as a visiting card, the mould is taken from the bath and the copper shell removed from the wax by pouring boiling hot water upon it. A further washing in hot lye, and a bath in an acid pickle, completely removes every vestige of wax from the shell. The back of the shell is now moistened with soldering fluid and covered with a layer of tin-foil, which acts as a solder between the copper and the later backing of lead.

The shells are now placed face downward in a shallow pan, and melted lead is poured upon them until of a sufficient depth; then the whole mass is cooled off, and the solid lead plate with copper face is removed from the pan and carried to the finishing room, where it is planed down to a standard thickness of about one-seventh of an inch. The various pages in the cast are sawed apart, the guard-lines removed, side and foot edges bevelled, head edge trimmed square, and the open or blank parts of the plate lowered by a routing machine to a sufficient depth to prevent their showing later on the printed sheet.

Then a proof taken from the plates is carefully examined for imperfections, and the plates are corrected or repaired accordingly, and are now ready for the press.

Although, owing to the expense and to the fact that the plate is more or less weakened thereby, it is desirable to avoid as much as possible making alterations in the plates, they can be made, and the following is the course generally pursued. If the change involves but a letter or two, the letters in the plate are cut out and new type letters are inserted; but if the alteration involves a whole word or more, it is inadvisable to insert the lead type, owing to its being softer and less durable than the copper-faced plate, and it will therefore soon show more wear than the rest of the page; and so it is customary to reset and electrotype so much of the page as is necessary to incorporate the proposed alteration, and then to substitute this part of the page for the part to be altered, by cutting out the old and soldering in the new piece, which must of course exactly correspond in size.

As a patched plate is apt at any time to go to pieces on the press, and may destroy other plates around it, or may even damage the press itself, it is generally considered best to cast a new plate from the patched one. This does not, however, apply to plates in which only single letters or words have been inserted, but to those which have been cut apart their whole width for the insertion of one or more lines.

The plates having been finally approved, they are made up in groups (or "signatures") of sixteen, and packed in strong boxes for future storage. Each box generally contains three of these groups, or forty-eight plates, and is plainly marked with the title of the book and the numbers of the signatures contained therein.

The longevity of good electrotype plates is dependent upon the care with which they are handled and the quality of paper printed from them; but with smooth book paper and good treatment it is entirely possible to print from them a half million impressions without their showing any great or material wear.

COMPOSITION BY THE LINOTYPE MACHINE

By Frederick J. Warburton.

The Linotype, pronounced by _London Engineering_ "the most wonderful machine of the century," was not the product of a day. Its creator, whose early training had never touched the printer's art, was fortunately led to the study of that art, through the efforts of others, whose education had prepared them to look for a better method of producing print than that which had been in use since the days of Gutenberg; but his invention abolished at one stroke composition and distribution; introduced for the first time the line, instead of the letter, as the unit of composition; brought into the art the idea of automatically and instantly producing by a keyboard solid lines of composed and justified type, to be once used and then melted down; rendered it possible to secure for each issue new and sharp faces; abolished the usual investment for type; cheapened the cost of standing matter; removed all danger of "pieing," and at the same time reduced greatly the cost of composition. The story is an interesting one.

In the autumn of 1876, Charles T. Moore, a native of Virginia, exhibited to a company of Washington reporters a printing machine upon which he had been working for many years, and which he believed to be then substantially complete. It was a machine of very moderate dimensions, requiring a small motive power, and which bore upon a cylinder in successive rows the characters required for printed matter. By the manipulation of finger keys, while the cylinder was kept in continuous forward motion, the characters were printed in lithographic ink upon a paper ribbon, in proper relation to each other; this ribbon was afterwards cut into lengths, arranged in the form of a page, "justified," to a certain extent, by cutting between and separating the words, and then transferred to a lithographic stone, from which the print was made. Such print was not, of course, of the highest character, but it was a beginning; and the machines were used in Washington and New York, mainly in the transcription of stenographic notes taken in law cases and in the proceedings of legislative committees. A number of these machines was built, but mechanical difficulties became so frequent that the parties interested resolved, very wisely, before proceeding to build upon a large scale, to put the machine into the hands of a thorough mechanical expert, so that it might be tried out and a determination reached as to whether or not it was a commercially practical one. At the head of the little company of men who nurtured this enterprise and contributed most largely by their labors and means to its development, were James O. Clephane, a well-known law and convention reporter, and Andrew Devine, then the Senate reporter of the Associated Press. In their search for an expert, a Baltimore manufacturer named Hahl, who had constructed some of these machines, was consulted, and upon his recommendation his cousin, Ottmar Mergenthaler, was selected to undertake the work, and thus the future inventor of the Linotype was discovered.

Mergenthaler was born in 1854, in Wuertemberg, Germany, had been a watchmaker, and at this time was employed upon the finer parts of the mechanical work done in Hahl's shop. The contract was that Mergenthaler was to give his services at a rate of wages considerably beyond what he was then receiving, and Hahl was to charge a reasonable price for the use of his shop and the cost of material. The task undertaken, however, proved to be a far larger one than had been anticipated, and the means of the promoters were exhausted long before the modifications and improvements continually presented had been worked out. The circle of contributors was therefore necessarily widened, and indeed that process went on for years, enough, could they have been foreseen, to have dismayed and disheartened those who were there "in the beginning." Mergenthaler and Moore, assisted by the practical suggestions of Clephane and Devine, continued to work upon the problem for about two years, by which time the lithographic printing machine had become one which indented the characters in a papier-mache strip, and this being cut up and adjusted upon a flat surface in lines, the way was prepared for casting in type metal. The next step of importance was the production of the "bar indenting machine," a machine which carried a series of metal bars, bearing upon their edges male printing characters, the bars being provided with springs for "justifying" purposes. The papier-mache matrix lines resulting from pressure against the characters were secured upon a backing sheet, over this sheet was laid a gridiron frame containing a series of slots, and into these slots type metal was poured by hand to form slugs bearing the characters from which to print. This system was immediately followed by a machine which cast the slugs automatically, one line at a time, from the matrix sheets.

It was in this work that Mergenthaler received the education which resulted in his great invention and in due time he presented his plans for a machine which was known as the "Band" machine. In this machine the characters required for printing were indented in the edges of a series of narrow brass bands, each band containing a full alphabet, and hanging, with spacers, side by side in the machine. The bands tapered in thickness from top to bottom, the characters being arranged upon them in the order of the width-space which they occupied. By touching the keys of a keyboard similar to a typewriter, the bands dropped successively, bringing the characters required into line at a given point; a casting mechanism was then brought in contact with this line of characters, molten metal forced against it through a mould of the proper dimensions, and a slug with a printing surface upon its face was thus formed. This was recognized as a great advance and was hailed with delight by the now largely increased company. The necessary funds were provided and the building of the new machine undertaken. But Mergenthaler continued active, and before a second of the "Band" machines could be built, he had devised a plan for dealing with the letters by means of independent matrices. These matrices were pieces of brass measuring 1-1/4 inches by 3/4 of an inch and of the necessary thickness to accommodate the character, which it bore upon its edge in intaglio; they were stored in the newly devised machine in vertical copper tubes, from the bases of which they were drawn, as required, by a mechanism actuated by finger keys, caught by the "ears" as they dropped upon a miniature railway, and by a blast of air carried one by one to the assembling point. Wedge spacers being dropped in between the words, the line was carried to the front of the mould, where "justification" and casting took place.

Success seemed at last to have been reached, and now the problem was, first, how to obtain means to build machines, and second, how to persuade printers to use them. The first of these was the easier, although no slight task; the second was one of great difficulty. The field for the machine then in sight was the newspaper, and the newspaper must appear daily. The old method of printing from founder's type, set for the most part by hand, was doing the work; a revolutionary method by which the type was to be made and set by machine, although promising great economies, was a dangerous innovation and one from which publishers naturally shrank. They could see the fate which awaited them if they adopted the new system and it proved unsuccessful. However, a number of newspaper men, after a careful investigation of the whole subject, determined to make the trial; and the leaders of these were Whitelaw Reid of the _New York Tribune_, Melvin Stone of the _Chicago News_ (to whom succeeded Victor F. Lawson), and Walter N. Haldeman of the _Louisville Courier-Journal_. Into these offices, then, the Linotype went. To Mr. Reid belongs the honor of giving the machine a name--line of type--Linotype, and of first using it to print a daily newspaper. Of the machine last described, two hundred were built, but before they were half marketed, the ingenious Mergenthaler presented a new form, which showed so great an advance that it was perforce adopted, and the machines then in use, although they gave excellent results, were in course of time displaced. The new machine did away with the air blast, the matrices being carried to the assembling point by gravity from magazines to be hereafter described, and the distributing elevator was displaced by an "arm" which lifted the lines of matrices, after the casting process, to the top of the machine to be returned to their places.

The improvements made in the Linotype since Mergenthaler's time (who died in 1899 at the early age of forty-five) have been very great; indeed, almost a new machine has been created in doing what was necessary to adapt it to the more and more exacting work which it was called upon to perform in the offices of the great American book publishers. These improvements have been largely the work of, or the following out of suggestions made by, Philip T. Dodge, the patent attorney of the parties interested in the enterprise from the beginning, and later the president of the Mergenthaler Linotype Company. They went on year after year under the supervision of a corps of gifted mechanical experts, the chief of whom was John R. Rogers, the inventor of the Typograph, until from the machine of Mergenthaler, supplying through its ninety keys as many characters, a machine appeared yielding three hundred and sixty different characters from the like keyboard. The magazines, too, were capable of being charged with matrices representing any face from Agate (5-point) to English (14-point), and even larger faces for display advertising and for initial letters, by special contrivances which cannot be described without carrying this article beyond reasonable limits. Among the ingenious devices added are: the Rogers systems of setting rule and figure tables, box heads, etc.; the reversal of the line so as to set Hebrew characters in their proper relation; the production of printers' rules of any pattern; the making of ornamental borders; a device for the casting of the same line an indefinite number of times from one setting. The machine was also greatly simplified in its construction.

The amount of money expended in the enterprise before the point of profit was reached was very great; it aggregated many millions of dollars; but the promoters had faith in the success of the machine and taxed themselves ungrudgingly. Among those who contributed largely to the ultimate result by substantial aid and wise counsel in the conduct of the business the name of D. O. Mills should be particularly mentioned.

It was Mergenthaler's great good fortune to have had as his supporters many men of the character of those mentioned above, and in thus being relieved of all financial anxiety and permitted to work out thoroughly and without delay every idea that suggested itself either to him or to the ingenious men who had been drawn into the enterprise. His profits, too, were proportionate to the company's success, and although he did not live to enjoy them for his natural term of years, he had the satisfaction of knowing that a handsome income would continue to flow into the hands of his wife and children.

The company's principal works are situated in the Borough of Brooklyn, New York City, and have a space devoted to manufacturing purposes of about one hundred and sixty thousand square feet. Approximately one hundred Linotypes, besides a large number of smaller machines and a vast quantity of supplies, are turned out from there every month; but the growing demand from abroad for American-built machines has led to the consideration of plans for an entirely new establishment, to be built in accordance with the latest modes of factory construction. About ten thousand Linotypes are now in daily use.

The machine as at present built is shown in part by the accompanying cut, and its operation may be briefly described as follows:--

The Linotype machine contains, as its fundamental elements, several hundred single matrices, which consist of flat plates of brass having on one edge a female letter or matrix proper, and in the upper end a series of teeth, used for selecting and distributing them to their proper places in the magazine. These matrices are held in the magazine of the machine, a channel of it being devoted to each separate character, and there are also channels which carry quads of definite thickness for use in tabular work, etc. The machine is so organized that on manipulating the finger keys, matrices are selected in the order in which their characters are to appear in print, and they are assembled in line side by side at the point marked _G_ in the illustration, with wedge-shaped spaces between the words. This series of assembled matrices forms a line matrix, or, in other words, a line of female type adapted to form a line of raised printed characters on a slug which is cast against them. After the matrix line has been composed, it is automatically transferred to the face of a slotted mould, as shown at _K_, and while in this position the wedge spaces are pushed up through the line, and in this manner exact and instantaneous justification is secured. Behind the mould there is a melting pot, _M_, heated by a flame from a gas or oil burner, and containing a constant supply of molten metal. The pot has a perforated mouth which fits against and closes the rear side of the mould, and it contains a pump plunger mechanically actuated. After the matrix line is in place against the front of the mould, the plunger falls and forces the molten metal through the mouth pot into the mould, against and into the characters in the matrix line. The metal instantly solidifies, forming a slug having on its edge raised characters formed by the matrices. The mould wheel next makes a partial revolution, turning the mould from its original horizontal position to a vertical one in front of an ejector blade, which, advancing from the rear through the mould, pushes the slug from the latter into the receiving galley at the front. A vibrating arm advances the slugs laterally in the galley, assembling them in column or page form ready for use. To insure absolute accuracy in the height and thickness of the slugs, knives are arranged to act upon the base and side faces as they are being carried toward the galley. After the matrices have served their purpose in front of the mould, they are shifted laterally until the teeth in their upper ends engage the horizontal ribs on the bar _R_; this bar then rises, as shown by the dotted lines, lifting the matrices to the distributor at the top of the machine, but leaving the wedge spacers, _I_, behind, to be shifted to their box, _H_. The teeth in the top of each matrix are arranged in a special order, according to the character it contains, the number or relation of its teeth differing from that of a matrix containing any other character, and this difference insures proper distribution. A distributor bar, _T_, is fixed horizontally over the upper end of the magazine and bears on its lower edge longitudinal ribs or teeth, adapted to engage the teeth of the matrices and hold the latter in suspension as they are carried along the bar over the mouths of the magazine channels by means of screws which engage their edges. Each matrix remains in engagement with the bar until it arrives at the required point, directly over its own channel, and at this point for the first time its teeth bear such relation to those on the bar that it is permitted to disengage and fall into the channel. It is to be particularly noted that the matrices pursue a circulatory course through the machine, starting singly from the bottom of the magazine and passing thence to the line being composed, thence in the line to the mould, and finally back singly to the top of the magazine. This circulation permits the operations of composing one line, casting from a second, and distributing a third, to be carried on concurrently, and enables the machine to run at a speed exceeding that at which an operator can finger the keys. A change from one face of type to any other is effected by simply drawing off one magazine and substituting another containing the face required, so that the variety of faces needs to be limited only by the number of them which the printer chooses to carry in his stock.

Matrices are also made bearing two characters, as the ordinary body character and the corresponding italics, or a body character and a small capital or a black face, and either of these is brought into use as desired by the touching of a key, so that if, for instance, it is required to print a word in italics or black face at any part of the line being composed, it is effected in this way, and composition in the body letter is resumed by releasing the key.

The latest pattern of machine is supplied with two magazines, superimposed one above the other, each with its own distributing apparatus. The operator can elect, by moving a lever, from which magazine the letter wanted will fall--the same keyboard serving for both. It is thus possible to set two sizes of type from one machine, each matrix showing two characters as described above.

COMPOSITION BY THE MONOTYPE MACHINE

By Paul Nathan.

Though for more than half a century machines adapted for the setting of type have been in use, it is only within a few years that the average printer of books has been enabled to avail himself of the services of a mechanical substitute for the hand compositor. The fact seems to be that despite the ingenuity that was brought to bear upon the problem, the pioneer inventors were satisfied to obtain speed, with its resultant economy, at the expense of the quality of the finished product. Thus, until comparatively recently, machine composition was debarred from the establishments of the makers of fine books, and found its chief field of activity in the office of newspaper publishers and others to whom a technically perfect output was not essential so long as a distinct saving of time and labor could be assured. Thanks, however, to persistent effort on the part of those inventors who would not be satisfied until a machine was evolved which should equal in its output the work of the hand compositor, the problem has been triumphantly solved, and to-day the very finest examples of the printed book owe their being to the mechanical type-setter.

The claim is made for one of these machines, the monotype, that, so far from lowering the standard of composition, its introduction into the offices of the leading book printers of the world has had the contrary effect, and that it is only the work of the most skilful hand compositor which can at every point be compared with that turned out by the machine. The fact that the type for some recent books of the very highest class, so-called "editions de luxe," has been cast and set by the monotype machine would seem to afford justification for this claim, extravagant as at first glance it may appear.

The monotype machine is, to use a Hibernicism, two machines, which, though quite separate and unrelated, are yet mutually interdependent and necessary the one to the other. One of these is the composing machine, or keyboard, the other the caster, or type-founder. To begin with the former: this is in appearance not unlike a large typewriter standing upon an iron pedestal, the keyboard which forms its principal feature having two hundred and twenty-five keys corresponding to as many different characters. This keyboard is generally placed in some such position in the printing office as conduces to the health and comfort of the operator, for there is no more noise or disagreeable consequence attendant on its operation than in the case of the familiar typewriter, which it so markedly resembles.

It has been said that the machines are interdependent; yet they are entirely independent as to time and place. The keyboard, as a matter of fact, acts as a sort of go-between betwixt the operator and the casting-machine, setting the latter the task it has to perform and indicating to it the precise manner of its performance. A roll of paper, which as the keyboard is operated continuously unwinds and is rewound, forms the actual means of communication between the two machines. The operator, as he (or she, for in increasing numbers women are being trained as monotype operators) sits facing the keyboard, has before him, conveniently hanging from an adjustable arm, the "copy" that has to be set in type. As he reads it he manipulates the keys precisely as does an operator on a typewriter, but each key as it is depressed, in place of writing a letter, punches certain round holes in the roll of paper. Enough keys are depressed to form a word, then one is touched to form a space, and so on until just before the end of the line is reached (the length of this line, or the "measure," as it is termed, has at the outset been determined upon by the setting of an indicator) a bell rings, and the operator knows that he must prepare to finish the line with a completed word or syllable and then proceed to justify it. "Justification," as it is termed, is perhaps the most difficult function of either the hand or the machine compositor. On the deftness with which this function is discharged depends almost entirely the typographic excellence of the printed page. To justify is to so increase the distance between the words by the introduction of type-metal "spaces" as to enable the characters to exactly fill the line. To make these spaces as nearly equal as possible is the aim of every good printer, and in proportion as he succeeds in his endeavor the printed page will please the eye and be free from those irregularities of "white space," which detract from its legibility as well as from its artistic appearance.

That the monotype should not only "justify" each line automatically, but justify with a mathematical exactness impossible of attainment by the more or less rough-and-ready methods of the most careful human type-setter is at first thought a little bewildering. The fact remains, however, that it does so, and another triumph is to be recorded for man's "instruments of precision."

Monotype justification is effected as follows: an ingenious registering device waits, as it were, on all the movements of the operator, with the result that when he has approached as close to the end of the line as he dare go, he has merely to glance at a cylindrical dial in front of him. The pointer on this dial signifies to him which of the "justifying keys" he must depress. He touches them in accordance therewith, and the line is justified, or rather it _will_ be justified when, as will be seen later on, the casting machine takes up its part of the work. That is the outward manifestation; it remains to be seen in what manner the machine accomplishes its task. Firstly, the machine automatically notes the exact width of the space left over at the line's end; then, also automatically, it records the number of spaces between the words already set which form the incompleted line; finally, it divides the residuary space into as many parts as there are word-spaces, and allots to each of these one of the parts. Thus if there is one-tenth of an inch to spare at the end of the line and ten word-spaces, then one-hundredth of an inch added to each of these spaces will justify the line with mathematical accuracy. But the machine will do something more wonderful than this. It will separately justify separate parts of the same line. The utility of this is comprehended when it is pointed out that when the "copy" to be set consists of what is technically termed "tabular" matter, the various columns of figures or so forth composing it are not composed vertically but horizontally and so each section must of necessity be justified separately.

Should the compositor be required to "over-run illustrations," as the term goes, in other words to leave a space in which the "block" for a cut may be inserted, so that it may have type all around it or on one side of it only, the machine offers no difficulty at all. All that the operator has to do in this case is to carry the composition of each line as far as necessary and then complete it with a row of "quads," or spaces. Thus, when the composition is cast by the casting-machine the space into which the block is to fit is occupied by a square of "quads." These have only to be lifted out, the block inserted, and the trick is done.

We will then imagine that the operator has finished his task. Of the bank of two hundred and twenty-five keys in front of him (the equivalent of a full "font" of type, with figures, italics, and symbols complete), he has depressed in turn those necessary to spell out the words of his copy, he has put a space between the words he has justified in accordance with the dictates of the justifying dial, has arranged the spaces for the insertion of blocks or illustrations, and as the result of his labors he has merely a roll of perforated paper not unlike that which operates the now familiar pianola or piano-player. Yet this roll of paper is the informing spirit, as it were, of the machine. Its production is the only portion of the work of the monotype for which a human directing agency is necessary, every other function being purely automatic.

The roll of perforated ribbon is lifted off the keyboard and put in place on the casting-and setting-machine. As it is swiftly unwound it delivers to the casting-machine the message with which the operator has charged it. Through the perforations he has made compressed air is forced. Now, as has been explained, the holes correspond to the characters or typographic symbols of the "copy," and the jet of air forced through them sets in motion the machinery, which controls what is known as the "matrix-case," a rectangular metal frame about five inches square, which contains two hundred and twenty-five matrices, or little blocks of hardened copper, each one of which is a mould corresponding to a character on the keyboard. This frame is mounted horizontally on a slide, which by an ingenious mechanical movement brings any one of the two hundred and twenty-five matrices over what is termed the mould. The particular matrix thus placed in position is determined by those particular holes punched in the paper ribbon at the keyboard, through which the compressed air is at that precise moment being forced.

The mould referred to is closed by the matrix, a jet of molten metal is forced in, and in an instant the type is cast, its face being formed by the matrix, its body or shank by the mould. The cast type is ejected and takes its place in the galley, to be followed by another and that by yet others in their regular rotation. It must, however, be pointed out that the composition emerges from the machine hind part foremost and upside down as it were. This enables the justification holes, which were originally punched at the _end_ and not at the beginning of each line, to direct the proper casting of the spaces in the lines to which they correspond.

It will be seen, therefore, that the casting portion of the monotype machine is actually automatic. It performs all its operations without human assistance or direction. Occasionally it will stop of its own accord and refuse to work, but this merely means that it has found something amiss with the perforated instructions, a mistake as to the length of a line or so forth, and it refuses to continue until the workman in charge of it puts the error right, then it starts on again and continues on its even course, casting letters and spaces and punctuation marks, and arranging them first in words, then in lines, next in paragraphs, and finally in a column on the galley.

The casting-machine works at so high a rate of speed (casting from one hundred and forty to one hundred and fifty characters per minute) that it can in its output keep well ahead of the operator on the keyboard. This, however, so far from being an inconvenience or leading to any loss of time, is an advantage, for four casting-machines, which can easily be looked after by one man and a boy, can cope with the work of five keyboard operators, or if all are engaged on the same character of composition two casters can attend to the output of three keyboards. This suggests a reference to the facilities offered by the machine for the production of matter composed in various faces of type. The machine casts practically all sizes in general use from five-point, or "pearl," to fourteen point, or "English." Owing to the number of characters included in the matrix-case, it can at the same time set upper and lower case, small capitals, and upper and lower case italics, or any similar combination of two or even three different faced alphabets. To change from one complete set of matrices to another is a simple operation, performed in about a minute of time, while the changing of mould, which insures a corresponding change in the size of the "body" of the type, takes about ten minutes.