Popular Science Monthly/Volume 40/December 1891/Type-Casting Machines
By P. D. ROSS.
IN the composing-room of the New York Tribune some forty type-casting machines have been used for several years. The foreman informed me in October last that all the ordinary reading-matter in the Tribune was being "set" by these inventions, and expressed himself perfectly satisfied with the working of the machines. As a rule, he said, not one of them was out of order, and on the average each did the work of three fair compositors. In a printed circular issued by the patentees of the machine the foreman, Mr. G. W. Shafer, declares that, compared with what the same amount of setting would cost if done by hand by compositors, "the machines save the Tribune office sixty per cent—probably more."
My object in visiting New York at that time was to look into the type-casting process. The result of the visit was a conviction that the problem of setting type by machinery has been solved. Small printing establishments may not benefit from it for a few years. Large establishments, particularly large newspapers, may profit at an early date. The New York papers are looking to this. The business manager of the World, Mr. G. W. Turner, informed me that he had ordered one hundred machines. In the composing-room of the Brooklyn Standard-Union I saw six machines working. I was informed that orders for machines had been placed by the New York Sun, Herald, Times, and Mail and Express. Outside of New York, the Louisville Courier-Journal uses thirty machines, and says it saves fifty per cent of what hand composition used to cost it. The Providence Journal uses twelve machines, and claims to save two hundred and fifty dollars per week. The Chicago News says it is saving fifty per cent in the cost of composition. These are only some of the newspapers which state that they have been using the machines regularly and successfully during the past year. Four machines ordered by the Canadian Government have been used in the Government Printing Bureau at Ottawa for some months, and, in reply to a question in the House of Commons recently, the Secretary of State, Hon. J. A. Chapleau, said that they were satisfactory and economical.
All this goes to show that the type-casting principle has obtained a practical footing in the market. In discussing the subject, I propose to confine myself as much as possible to my personal experience and investigations. If I state anything which I do not know personally or have not been told at first hand by disinterested persons, I will give the source of my information.
What Type-casting is.—Before describing the type-casting principle, allow me to review briefly the process of type-setting by hand.
In this process the operator, technically called a compositor, has before him an oblong frame (or "case") divided into a number of small open boxes. One box contains the a's, another the b's, another the numeral 1's, another the numeral 2's, and so on. In his left hand the compositor holds a little steel receiving box, called a "stick." With his right hand he picks out from the "case" the letters he requires to form a word, and puts them one by one in his "stick." The stick is the same width as the column of his newspaper. Toward the end of each line in his stick he has to pad out the line with lead slugs so as to exactly fill the width of the stick; this is called "justifying." When he has a certain quantity of reading matter in his stick, say one tenth of a column in length, he transfers the type to a "galley" or long stick. By and by, when the galley is filled up, the type in it is transferred to the large receiving form called a "chase," in which the columns of the newspaper are made up to be placed on the printing-press. Such, very roughly described, is the process of type-setting by hand. After the paper is printed the compositor must pick out all the separate letters and numerals from the columns of type, and put them back in the proper boxes in his "case." This is called "distributing." The "distribution" occupies about one fifth of a compositor's whole working-time.
In all this, civilization is to-day where it was five hundred years ago, and almost where the Chinese were two thousand years ago. Alone of all the great inventions of man, type-setting has stood still from its birth until now. In war and in commerce, on our farms and in our workshops, in travel and in our homes, almost every mechanical process, once slowly and laboriously effected by manual or animal labor, has been quickened generation after generation by new appliances or inventions, save and except the work of type-setting. That is as slow now as when Coster or Gutenberg did the first European type-setting early in the fifteenth century. Printing has otherwise moved with the rest of the world. Our printing-presses, our power, our folding and pasting machines, all are wonderfully improved. Nothing in all the world has developed more marvelously than the printing-press. But type-setting has stood still. The ordinary composing-room of to-day can work no faster and no better than the composing-room of the fifteenth century.
With the type-casting machine should come a new era. The operator needs only the intelligence which is required in a good compositor. He does not require more than one tenth the training. Thus equipped, he can, I believe, do steadily and regularly the work of three fair hand compositors. He does not handle type; has no "stick"; is not required to do any justification nor any distributing. He sits in front of a machine and works a key-board and a lever, and the machine does everything else.
Now to outline the working of the type-casting machine. A key-board similar to that of a type-writer fronts the machine There is a key for each letter of the alphabet. The operator sits in front of the key-board. Let us suppose that he wishes to set the word "new." He touches the key n. The touch on the key releases from a magazine in rear of the machine a mold technically called a matrix, for the letter n. The matrix, which is of brass, slides down into a receiver near the key-board Next the operator touches the key e. A matrix for the letter e is released and slides down alongside the letter n. The operator touches the keyw. A matrix for w comes down and ranges itself alongside e. Now, in the receiver we have, what?—the word new in type? No, nothing of the kind. We have three little brass molds standing side by side, from which, if we poured molten metal into them, we would get the word new in a solid cast. But there is no type. The machine knows nothing of type whatever, though, for convenience' sake, we are calling it at present a type-casting machine.
But the time is not come to put molten metal into the three little molds or "matrices." An entire line should be set not merely a word. Suppose the line is to be, "new things come to pass." The operator proceeds to touch key after key for the successive letters until the matrices for the whole line are ranged side by side. Now at this point comes in what was for years the great problem in type-casting by machinery. As the end of a line of matrices or type is approached, it may not be possible to fit in an even word or syllable. Padding, or "justifying," becomes necessary. In setting by hand, the compositor does this with little lead slugs, called "spaces," inserted between words. How is this to be done by a machine? Inventors long stuck at it. But they have found out how. The process is simple in action, though difficult to describe without a model. Roughly speaking, the "spaces" or slugs which are used between each word in the line of matrices are compensating wedges, the bottoms of which project below the matrices. When the line of matrices requires justification," a touch on a lever by the operator causes the bottoms of the compensating wedges to be struck by a cross-bar, which forces the wedges up between the words until the line is solidly filled out.
The line of matrices or letter molds is then ready to receive a cast. Where is the molten metal? It is in the machine. This wonderful apparatus has a furnace for a heart and a melting-pot for a stomach. The furnace, consisting of a series of gas-jets, and the melting-pot, are in the lower part of the body of the machine. In the pot, stereotype metal is melted. The pot is not very large, because fresh metal may be put into it at any time when needed. The same metal may be used over and over again as often as desired; it does not deteriorate.
A jet of molten metal is thrown into the matrices by a force-pump worked by the automatic action of the machine. The metallic fluid, hardening almost in an instant, a property of
stereotype metals, forms a solid cast or bar, on the face of which is the line "new things come to pass," and the machine automatically ejects this cast or bar of letters into a receiver, into which it is followed line after line by new casts with wonderful rapidity, until in a short space of time a column of reading-matter in bars is ready for the press. The speed of the machine is measured by the speed of the operator at the key-board. It can work as fast as he can.
When a line of matrices has been utilized, the matrices must be returned to their channels ready for use again. This is accomplished by ingenious contrivances as soon as the cast has been made. The matrices being thus promptly returned, there is only need for a few of each letter. Thus a few dozen of the little brass molds do the work which in type-setting by hand needs a stock of from forty to fifty pounds of type.
The Rival Patents.—There are two type-casting machines on the market. These are the Mergenthaler or Linotype, and the Rogers or Typograph. The Linotype weighs a ton, covers floor space about six feet by six, stands seven feet high, and is sold for $3,000, or rented for $500 a year. I have seen an expert operator set at the rate of nearly eight thousand ems per hour on it from a phonograph communicating with his ear. The proprietors claim a regular practical speed of over four thousand ems an hour, which is four times the speed a good compositor averages by hand, if we include the time he must take for distributing. On the Linotype, the first time I ever touched a key-board, I set one hundred and fourteen ems of strange copy in six minutes, or at the rate of eleven hundred and forty ems an hour.
The Typograph weighs four hundred and fifty pounds, covers floor space four feet by four, is four feet six inches high, sells for $2,500, and rents for $365 a year. The proprietors claim a regular practical speed of three thousand to thirty-five hundred ems per hour. I have set one hundred and fourteen ems by the Typograph in nine minutes. At the end of each line the operator at the Typograph must stop to throw back the cap of the machine, a movement which restores the matrices to their magazines. The proprietors of the Typograph claim that it can work as fast as will ever be practically possible on any machine. In other words, they think that human beings will not be physically capable throughout a whole working day of requiring as great a steady speed as the Typograph can give.
The Typograph was submitted to a severe practical test in September, 1890, by the New York World. An eight-page section of the Sunday World, September 28th, was set by one machine working continuously day and night for one hundred and nineteen hours and thirty-five minutes, or nearly a week. The object of the test was to ascertain how the machine would bear a continuous steady strain. Three operators took eight-hour shifts at the work. The machine—I was informed both by the business manager of The World, Mr. Turner, and by one of the operators, the foreman of The World composing-room—stood the test almost perfectly. I measured the amount of setting done. It came to one hundred and fifty-six thousand ems of minion. As the machine was worked one hundred and nineteen hours, this shows an average speed of only thirteen hundred ems per hour. At first sight this might seem disappointing. There were reasons why it was not so. The three operators were compositors, and had had only three or four weeks' practice on the Typograph. Owing
to faults of the motor used to run the machine, it had to be worked by hand-power one quarter of the time. The three operators not only ran the machine, but they read the proofs, made the corrections, set the headings, and made up the "forms" ready for the press. Finally, the machine lost several hours' work through a fault in a casting. Taken as a whole, it seems to me the test was a conclusive proof of the practical success of the Typograph.
Type-setting by Machinery.—Type-casting is quite different from machine type-setting. Before contrasting type-casting with ordinary hand type-setting, it may clear the way to outline the principle of machine type-setting.
The type-setting machine has a reservoir of type, instead of a magazine of matrices as in the casting-machine; but, unlike the matrices, which return to their magazine the moment a line is cast from them, the type must go the whole way to the printing-press. Otherwise, the action of the type-setting machine is somewhat similar to that of the casting-machine. The type-setting machine is also worked by an operator at a key-board. When the operator touches a key, a type is released, just as a matrix is in the casting-machine, and slides into a receiver, where it is joined by other successive letters until words and lines are formed. As type is directly used, there is no furnace or melting-pot about the machine. This is the only advantage it has over the casting-machine, while compared with the latter it has serious drawbacks.
The type-setting machine seems to be a practical success, and an improvement on type-setting by hand; but, if for two reasons only, it is doomed to be superseded by the casting-machine. 1. It requires a heavy stock of type instead of a few matrices. 2. At least two attendants are required to each machine, one to operate the key-board, the other to justify the lines, attend to corrections and superintend matters generally, and to distribute the type again. Still, the business manager of the office in which the New York Forum is printed, informed me that through their use he is saving $1,700 a year in the setting of that monthly magazine, which does not require in a year as much composition as a daily paper in a month.
Comparison with Type-setting by Hand.—In any considerable quantity of straight reading matter, type-casting machines as compared with hand composition should, if working successfully, effect a saving of from one fourth to one third the cost of setting. Moreover, the setting is better. Perhaps this contention is best illustrated by figures. Those which I propose to give are based on the conditions prevailing in Canadian newspaper offices. Let us suppose an office in which one hundred and twenty thousand ems of straight reading matter are set per day in minion type. To fix ideas, we may describe this roughly as equal to about twenty-five ordinary newspaper columns. Many of the larger city papers in Canada print just about this quantity of reading matter per day. The union rate paid compositors in Canada is thirty-three and a third cents per one thousand ems. One hundred and twenty thousand ems would cost, therefore, about $40 for composition, apart from the cost of the type, foremen, office, etc. Forty dollars per day would come to $12,000 per year of three hundred working-days. Now, let us see what it would cost to do the same amount of setting by the type-casting machines. These are claimed by their proprietors to work at the rate of three thousand to five thousand ems per hour in regular use. Making allowance for the probability that operators could not keep up such a speed all day, that mistakes have to be corrected, and accidental stoppages might occur, we may admit that the machines can set an average of twenty-five hundred ems per hour during an eight-hour day, or twenty thousand ems per day each, which is little more than half what the inventors claim as practical. Six machines could at this rate set one hundred and twenty thousand ems per day. As already said, to set this by hand would cost $12,000. The cost of the machine work would be—
|Six machines at $500 rent each||$3,000|
|Six operators at say $14 per week||4,308|
Or equivalent to a saving of $3,132 on the setting by hand, or over twenty-five per cent. The estimate of $14 per week as a fair rate for operators of the machines is not too low for a Canadian office. First-class compositors certainly do not average more.
As a further illustration, I may give the actual figures of cost of a composing-room with which I am familiar. The setting amounts to about sixty thousand ems in a nine-hour day, done by ten to twelve compositors. A number of the hands are paid by the week, and the straight setting costs only about twenty-five cents per thousand; or, for sixty thousand ems, $15 per day—equal for three hundred days to $4,500 per year. There is also a foreman at $14, an assistant foreman at $12, and a couple of lads at $3 each. These four, costing $32 per week, or $1,664 a year, do all the setting of space advertisements. There is $2,000 worth of type, costing for interest say $140 per year, and requiring renewal at the rate of say $400 per year, or complete renewal once in five years. The cost of the composing-room is therefore somewhat as follows:
|Composition by hand||$4,500|
|Cost of type||540|
|Rent, heat, light, etc., say||500|
To set sixty thousand ems in a nine-hour day, or six thousand seven hundred ems per hour, would require say three type-casting machines, which at $500 rent would cost $1,500 per year, and the composing-room figures would be:
|Rent of machines||$1,500|
|Three operators at $14||2,184|
|Gas for machines, say||500|
|Rent, heat, light, etc||500|
The saving would apparently be some $856, or over twelve per cent, while less room would be required, cleaner and better work would be done, the labor better paid, and a higher class of operators employed. Later I will touch on some reasons why it might not be safe to depend on type-casting machines in so small a business. In a larger business there is little doubt in my mind that the use of the machines is preferable to hand composition.
Finally, it is much easier to learn to operate the type-casting machine than to learn to set type. To set type at the rate of a thousand ems an hour requires two or three years of constant practice. To set a thousand ems on the type-casting machine in an hour requires no previous practice. It can be done the first time a person touches a key-board. This seems a strong statement to make, but I have the best of reasons for knowing it to be true. I did it, as already described. Previous to making the attempt I had never touched a key-board but once, and that was a dummy-board. I had never touched a type-writer or any other instrument the use of which might qualify one for operating the type-casting machine. Being in the rooms of the Linotype Company in New York recently, I asked and received permission to try the machine; and picking up a printed clipping from which the operator had been setting, I went to work and in six minutes set one hundred and fourteen ems, equal to one thousand one hundred and forty ems per hour, stopping because the clipping then ended. I repeated similar experiments on other machines subsequently, with much the same average result. In short, I was able to do with the machine at sight and without practice what it would take me years to learn to do by hand. As to becoming expert on the machines, a number of operators whom I have questioned agree that from three to six months' practice enables one to attain a speed of three thousand to four thousand ems from ordinary copy.
In fact, as I have stated, the only limit of speed on the Linotype is the rate at which the operator can move his fingers. He can not work quite so rapidly as a type-writer, because at the end of each line of matrices he must stop to touch the lever which sends the line off to receive a cast. Supposing we allow twenty-five per cent of his time for this, which is surely a large proportion, we can get an idea of the possible practical rate of the machine by comparing it with the possibilities of a type-writer. Upon a type-writer, a rate of sixty words per minute from dictation is not very high. The Senate Hansard reporters of Canada employ several type-writers who average from sixty to seventy words and over for considerable periods of time. Allowing the speed of the operator on the type-casting machine to be twenty-five per cent less, we have at least forty-five words per minute as the practical rate of the machine. This is equal to seven thousand one hundred and five ems per hour. As already said, I saw one man at the Linotype set for half an hour from a phonograph at a rate of nearly eight thousand ems per hour, and the setting was as "clean" as that of the average compositor.
Summing up the comparison between hand setting and machine casting, I find: 1. The machine is much more easily learned.
2. No type is required. 3. Less space and fewer hands are needed in the composing-room. 4. Setting is cleaner, and probably one third cheaper. 5. Justification is automatic and perfect. 6. By changing the matrices, which can be done in half an hour, a different style of type becomes available. 7. "Leading" can be done much more quickly. 8. There is no "pi-ing," or mixing up of type. 9. Fewer typographical errors are likely. You do not have inverted letters, nor mistakes due to the type having been wrongly "distributed" in the case, which are a source of frequent typographical blunders.
Drawbacks and Possible Complications.—It will be asked, How is it that these remarkable machines have not at once sprung into popularity?—so cheap, so rapid, so easily learned, so economical! How is it that so little has been heard about them? Well, the patents were only perfected last year, and the machines are not yet being made fast enough to supply the demand. Meanwhile, there are many possible complications, the fear of which must cause the average printing-office to hesitate to try the machines. 1. The machines require power to drive them effectively. The failure of power for any reason would seriously interfere with them, although they can be driven by foot-power in an emergency. 2. They require gas or gasoline for their furnaces; the failure of the gas from leakage, or cold, or accident, would stop the machines.
3. The molten metal in the melting-pot must always be at a tolerably even temperature; otherwise the casting is bad, perhaps impossible. It is claimed that this difficulty has been overcome in the Linotype, and that the temperature of the molten metal is automatically kept at a temperature varying not more than 10° Fahr. A column of mercury is connected with the melting-pot, and when the temperature causes the mercury to ascend beyond a certain point, it lowers the gas-jets which supply the heat. When the mercury descends below a certain point, it turns on the gas more strongly. 4. The machines are composed of many parts, and if they get out of order in a town in which expert mechanics are not at once available, their usefulness is gone for that day at least. 5. They can set only straight reading matter, so that advertisements, display headings, cross lines, italics, etc., must be set by compositors. 6. If a mistake of a letter is made in setting by the machine, the whole line must be recast, unless (which is not likely) the mistake is noticed the moment it is made and the operator stops to rectify it by changing the matrix. However, a whole line can be reset and recast almost as quickly as a compositor can correct by hand a mistake in a type letter. 7. It is a more serious drawback that if, in correcting proofs, it is desired to insert additional words, a number of lines may have to be recast. 8. The matrices in which the casts are made are possibly liable to wear a little, and so to soon make bad casts. Of course this can be remedied by getting new matrices, which are not expensive. 9. In a small office where two or three machines might be employed, there would probably be only two or three expert operators; if one took ill, the machine would become almost useless for the time being.
Present Practical Availability.—A small printing-office is hampered in many ways with regard to the use of machines, nor can it safely, at present, take the chances of break-downs. Where only three or four machines can be used, the stoppage of one means a loss of twenty thousand ems of setting per day. That is serious enough; but if the cause of stoppage should affect all the machines, there must be a business dead-lock, because small concerns, or rather concerns in the smaller centers of population, can not at slight notice secure a staff of compositors to replace the machines, or arrange for publication elsewhere. Even, therefore, were the machines being manufactured as fast as desired, it is questionable whether they would find a market at present outside the large cities where expert mechanics can be had to attend them at a moment's notice, and where arrangements for special help or special publication can be made in an hour, if necessary. But I think that in any office setting one hundred thousand ems a day, or over, it would pay the proprietors to at once procure machines sufficient to do at least half their setting, retaining a certain number of compositors with them. I can see no reason why this should not be a fairly safe experiment and a financial success.
The machines are available on a very liberal basis. Either company leases them at a moderate rental, agrees to take them back if not satisfactory, to keep them in repair while used, and to replace them with new machines in case of improvement of the patent.
The typographical unions admit that the machines must be accepted as a practical fact. The International Typographical Union, at its last annual meeting in the United States, recommended its subordinate unions, in cities in which the machines come into use, to prepare a scale of prices for the work done, and to urge that union compositors be employed as operators. This is a sensible acceptance of a new order of things.
In conclusion, this is to be observed: There are theoretical objections to the machines in many small details which have not been touched on in this article, partly because I wish to present a clear general idea of the subject unencumbered by trivialities; partly because to handle them would require complicated and technical descriptions likely to confuse those who have not seen the machines, or who are not familiar with type-setting or stereotyping methods and appliances. With regard to such possible objections, it should be remembered that the type-casting principle scarcely now requires to defend itself against fanciful opponents. It has been tried, and not found wanting. As was stated at the outset of this article, a large number of Linotypes have been successfully employed for years in the composing-room of a leading New York paper. I have tried to deal with the chief possibilities of failure in the machine and it has been noticed that these possibilities seem to be chiefly in connection with printing establishments of limited extent and means. Few of the drawbacks, it appears to me, would be serious in a large office employing machines, and located in centers where the prompt assistance of expert mechanics can be had, and my conviction of this is borne out by the New York Tribune's experience. Such a test as the Linotype has received in that office during five years is the most conclusive answer to technical or theoretical objections to the principle of type-casting. The real problem with a publisher should be, not whether the machines are a success when used on a large scale, but whether his own business is large enough to justify him in introducing them into his own office. To use an exaggerated illustration, there is no question but that a steam-locomotive is an infinitely more useful powerful, and, on a proper scale, more economical affair than a wheelbarrow; but a laborer building a bit of roadway may do better with the wheelbarrow.