Popular Science Monthly/Volume 29/June 1886/Primitive Clocks
By FREDERIC G. MATHER.
THE story is that King Alfred had no better way to tell the time than by burning twelve candles, each of which lasted two hours; and, when all the twelve were gone, another day had passed. Long before the time of Alfred, and long before the time of Christ, the shadow of the sun told the hour of the day, by means of a sundial. The old Chaldeans so placed a hollow hemisphere, with a bead in the center, that the shadow of the bead on the inner surface told the hour of the day. Other kinds of dials were afterward made with a tablet of wood or straight piece of metal. On the tablets were marked the different hours. When the shadow came to the mark IX., it was nine o'clock in the morning. The dial was sometimes placed near the ground, or in towers or buildings. You see, in the picture, two sundials
that are on the Gray and Black Nunnery in Ottawa, the capital of Canada. The old clock on the eastern end of Faneuil Hall in Boston was formerly a dial of this kind; and on some of the old church-towers in England you may see them to-day. Aside from the kinds mentioned, the dials now in existence are intended more for ornament than for use. In the days when dials were used, each one contained a motto of some kind, like these: "Time flies like the shadow"; or, "I tell no hours but those that are happy."
But the dial could be used only in the daytime; and, even then, it was worthless when the sun was covered with clouds. In order to measure the hours of the night as well as the hours of the day, the Greeks and Romans used the clepsydra, which means, "The water steals away." A large jar was filled with water, and a hole was made in the bottom through which the water could run. The glass, in those days, was not transparent. No one could see from the outside how much water had escaped. So there were made, on the inside, certain marks that told the hours as the water ran out; or else a stick with notches in the edge was dipped into the water, and the depth of what was left showed the hour. Sometimes the water dropped into another jar in which a block of wood was floating, the block rising as the hours went on. Once in a while, some very rich man had a clepsydra that sounded a musical note at every hour.
Another way of measuring time among the ancients was by the sand, or hour-glass. This was made of pear-shaped bits of hollow glass with a very small opening between them. It held just sand enough to run from the upper into the lower pear in the space of one hour. The glass was then turned the other side up and the sand ran back, also taking an hour. You have seen glasses of this kind where the sand runs out in three minutes. They are used for boiling eggs. King Charlemagne, a thousand years ago, had a glass of this kind that ran for twelve hours without turning. It was marked on the outside with red lines to show the escape of the sand. Hour-glasses were so common after this that they were carried in the pocket like watches. Every minister had one to mark the length of his sermon, which was a very serious matter in England during the protectorate of Cromwell, very few sermons being as short as one hour. It is said of one minister that when the sand ran out of his glass he turned it over, saying, "I know that you are all good fellows, so let's have another glass." Once, when the preacher had turned his glass a second time, showing that he had already preached two hours, the sexton asked him to lock the door and put the key on the nail when he was through, because the few people that were left wanted to go home to dinner. We also read that, in the early history of New York, the soldiers who defended the city used hour-glasses to tell when they should go on guard.
We have seen that the dial could be used neither at night nor in cloudy weather. We have also noticed that the hour-glass had to be watched so that it might be turned at the very moment the sand ran out. And we have also seen how inconvenient it was to measure time by the running of water. None of these ways was accurate enough, for minutes and even hours would be lost. A better means of measuring time was sought for; and this was found by means of a clepsydra, in which the water drove a wheel that marked the hours by a hand. The old Romans used this water-clock; but, when their empire was destroyed, all Western Europe forgot the existence of such a thing. In the year 807 a. d., the Caliph of Bagdad, Haroun-al-Raschid, sent to Charlemagne a water-clock of this kind. Soon after we learn that, instead of the running water, a weight was used for turning the wheel. But whether the clock was run by water or by a weight it was always a hard matter to have the hours of the same length. The escapement, which we shall speak of presently, made one hour more nearly the length of every other hour. The machine for telling the hours was, for many years, called the horologe, or "hour-teller." The word "clock" was applied only to the bell that struck the hours. It sounds very much like the Saxon, French, and German words that mean "bell." About nine hundred Years ago horologes were brought into England by the Catholic clergy. Very large horologes were built into the towers at Canterbury Cathedral, in 1292; at Westminster, in 1290; at Exeter Cathedral, in 1317—the striking part of which is still in use; at the cathedrals of Wells and Peterborough; and at St. Albans Abbey in 1326. A smaller horologe was made for Charles V of France in 1370, by a German named Vick.
Horologes, or clocks, would have remained in this imperfect state until to-day if it had not been for the invention of the pendulum, which means "something that swings." You all remember the story of Galileo, who, when a boy, watched the chandelier as it swung to and fro in the cathedral at Florence. The young boy noticed that it moved with great regularity. If it had moved all the way around the point where it was held, or suspended, it would have made a circle; but as it moved only a small part of the way, it moved in what is called the "arc" of a circle. Galileo saw that it took just as long a time to go from one end of the arc to the other as it did to return. This is called isochronism, or "equal times." In 1620, several years after Galileo's discovery, Huygens first used the pendulum to regulate the movement of a clock. You may see how this is done by looking at Fig. 1. We have here the simplest form of clock-work, or "movement," as it is called. A wheel, with teeth on the edge, turns on a pin, i, by the force of the weight h, the string being wound about what is called a "barrel" at i. If there is no way of stopping the wheel, it will run down very fast and very unevenly. Here is just where the pendulum becomes useful. The pendulum is a long wire, a c, the part c being enlarged into what is called a "bob." The pendulum swings on the point a. It has an arm, d g, fastened to it and swinging with it. The points of this arm are called the "pallets." When the pendulum is in the position marked by the black line you will see that the wheel is stopped by the pallet d. But, when the pendulum swings to the place marked by the dotted line, the pallet d moves out to e. This lets the wheel move a little; but, before it moves a notch, the pallet g has moved to f and catches the wheel below. When the pendulum swings from b back to c, f is moved to g, and the pallet d stops the wheel from going any farther. So that, while the pendulum has gone from c to b and back again, only one tooth Fig. 1. of the wheel has escaped instead of two. The arm of the pendulum which acts upon the teeth of the wheel and the wheel itself are called the "escapement," because they let only a little of the power in the weight escape at a time—just as the hour-glass allowed but a little of the sand to escape at once, and as the clepsydra allowed only a little of the water to run out at a time.
The earliest form of an escapement was that of Vick. It was a small wheel that was turned back and forth by a twisted string. Afterward it was turned by a spiral spring, the wheel being always horizontal, or running at right angles to the other wheels, that were vertical. A new "scape-wheel," as it is called, was invented by Dr. Hooke, which moved vertically, or in the same plane with the other wheels. This is the wheel that is shown in Fig. 1. You will see by the figure that, when the bob is at b, and the tooth of the wheel comes on the pallet f, it will throw f over to g and help the bob to move from b to c. This is called the "recoil" escapement, because the force of the wheel gives such a sudden jerk to the pendulum. The cheaper clocks frequently have the recoil escapement. Very much of this jerking motion is saved by the "dead-beat" escapement, invented by Graham, an Englishman. It is so called because the tooth of the wheel falls dead upon the pallet and stays there until the pendulum starts back and releases it. The teeth of the dead-beat scape-wheel are of a different shape from those shown in Figs. 1 and 2. The "gravity" escapement is so called because another weight beside the principal weight gives an impulse, or motion, to the pallet. There are many other kinds of escapements, that are too difficult to be explained here. I have described only the simpler kinds.
In Fig. 1 the pendulum is made very much shorter than it should be, so that it will not take up the whole of the page. At the earth's equator it should be about thirty-nine inches long, to "vibrate," or go from c to b in one second. At the latitude of Washington, where the force of gravity is greater, the length is thirty-nine and one tenth inches. At London, which is still farther north, the length is thirty-nine and one seventh inches. A pendulum of the right length in London would lose two and one quarter minutes a day at the equator. The pendulum that vibrates from c to b in two seconds must be four times the length of a one-second pendulum. The pendulum of the great clock at Westminster moves once in two seconds. It is nearly fifteen feet long, and it weighs seven hundred pounds—the heaviest in the world. The heavier and longer the pendulum, the more regularly will the clock move. But pendulums may be too long and too heavy. Almost all of the clocks that were made before the year 1800 had pendulums about Fig. 2. thirty-nine inches long, and they stood with their cases over five feet high—usually in the corner of the room. They were so clumsy that only the machinery was peddled about from place to place—the nearest cabinet-maker being called upon to make the case. By-and-by it was found that, if, in Fig. 1, the pendulum would go from c to b in one second, it would go from c to b, back again to c—or twice as fast—if it were one quarter as long. After that, clocks were made short enough to stand on a shelf.
It had also been found that the bob of the pendulum, when moving in the arc of a circle, was not reliable; but that all the trouble was avoided if it moved in the arc of a cycloid (or "like a circle"). This arrangement is shown in Fig. 2. The pendulum hangs from a fixed point, a, where it is fastened securely. The upper end of the wire is beaten into a very thin spring. When the bob b moves back and forth, it does not move in the arc of the circle c d, but on the dotted line ef. Great care is taken in preparing the spring at a, so that the bob will have no other motion than that from e to f. Should it move sidewise, or twist about, the clock will be spoiled. The bob was formerly flat, like a small plate, or round, like a ball. It was then a difficult matter to run the pendulum-wire through the exact center, and therefore the best bobs are now made in the form of a cylinder. A nut at the end of the wire keeps the bob from slipping off. If the nut is turned to the right, the pendulum is shortened, and the clock goes faster. If it is turned to the left, the clock goes slower. Sometimes it is necessary to regulate the pendulum without stopping it. This is done by placing small weights on the parts of it that project. In order to keep them of the same length, both in summer and in winter, pendulums were often made of wood; but it has been found that if the bob is made of bars of iron and zinc, or brass and steel, in the form of a gridiron, the different expansions of the two metals keep the pendulum at the right length. The pendulum-rod sometimes ends in a cup of mercury at the bob. When the heat expands the rod, the mercury is forced upward in the cup and nearer the fixed end of the pendulum. The object of both the gridiron pendulum and the mercurial is to bring the "center of oscillation" as near as possible to the "center of gravity." Another kind of a pendulum is called the "rotary," because the bob moves in a circle instead of going from side to side, but this is not thought to be at all reliable.
From what has been said already, you will see that the weight h (Fig. 1) would soon run away with the scape-wheel unless the pallets d e f g dodged in and out among the teeth and stopped it from going so fast. The pendulum, too, instead of moving back and forth between b and c, would stop half-way between them in a vertical or up and-down line, like the plummets that the bricklayers use. A clock with simply the scape-wheel and the pendulum will soon run down; you must therefore have more wheels and a heavier weight to move them, or else your wheels will not move evenly enough to carry the minute-and hour-hands over the "face" that is outside. In Fig. 3 you will see that we have added other wheels; but you will recognize the scape-wheel in c, and the weight hanging to the wheel a. As it descends, the weight pulls the wheel a in the direction of the arrow. The wheel A turns with the wheel a, and it has seventy-eight "teeth," as the cogs are called. At b is a small wheel called a "pinion," with six "leaves," as the cogs are called. The large wheel, B, has also seventy-eight teeth; and the pinion c has also six leaves. While A is turning round once, B and b turn thirteen times, because b has one thirteenth as many teeth as A. In the same way C and c turn thirteen times as fast as B and b. I have a clock before me in which the wheel A turns once in one hundred and thirty minutes, or two hours and ten minutes. The wheel B turns in ten minutes, and the wheel C in ten thirteenths of a minute. You will see that the scape-wheel C does not always take exactly a minute to go round. This scape-wheel has forty-two teeth, which is more than the usual number. If there were sixty teeth, and the pendulum marked one second at each swinging, the scape-wheel would turn once every minute. But this is not necessary; besides, the scape-wheel must be small enough for the pallets to take in about nine teeth between them, and yet be able to swing clear of them altogether.
The series of wheels in Fig. 3 is called the "train." You can not see the train in the clock so plainly as it is drawn in the picture, because
one wheel is placed behind the other in order to take as little room as possible. Sometimes, instead of only one wheel, B, between A and C, there will be two or three wheels—all of them smaller. The train of wheels is then harder to move, and the weight must be heavier. If the weight drops two inches in twenty-four hours, it will need a space of sixteen inches if it is to run eight days. The length of time that the clock will run depends upon three things: 1. The length of the pendulum; 2. The space through which the weight falls; 3. The number of wheels in the train, and the number of teeth in each wheel. We have already seen how the length of the pendulum can be regulated. If the weight has a small space allowed for its fall, the clock may be made to run longer by increasing both the weight and the number of teeth. The number of teeth may be increased by increasing the number of wheels, or by putting in new wheels.
The wheel D, Fig. 3, is called the "center wheel," because it turns once in an hour. It has thirty-six teeth. In former times the wheel A turned once in twelve hours; and the axle, or "arbor," a, went through a hole in the face of the clock. A hand on the end of the arbor passed over certain figures on the face which marked the hours from one to twelve. This hand was called the hour-hand; but, as it could not mark the minutes, the center wheel, D, was so made that it would turn once in an hour, and thus, by carrying a hand over the face outside, marked the minutes. After this change was made no one cared whether the wheel A turned in one hour or in three hours, or whether the wheel C turned in one half minute or in two minutes, if only the wheel D turned in exactly one hour. At d is a "cannon" pinion that sticks to the arbor by friction. The minute-hand, which is placed upon the pinion, may thus be moved without turning the wheel D or any of the other wheels.
We must now provide an hour-hand. The cannon-pinion a (Fig. 4), with twelve leaves, runs on the arbor of the center wheel;Fig. 4. but it could not be drawn in Fig. 3, because it is behind the center wheel, D. These twelve leaves, A (Fig. 4), run into thirty-six teeth in the wheel B. You will notice that the teeth and the leaves are not drawn in the picture. On the farther side of B is the pinion b, with twelve leaves which run into the forty-eight teeth of the wheel C. The wheel C and the pinion b are marked with dotted lines, because they are behind the pinion a and the wheel B. If a turns once in an hour, B will turn once in three hours, and C once in twelve hours. If what is called a "barrel" is placed over the cannon-pinion of the center wheel, and one end of it is fastened to the wheel C, the other end that comes through the face of the clock will carry the hour-hand. These wheels, in Fig. 4, are independent of the wheels in Fig. 3, except that a, in Fig. 4, fits upon the arbor d, of D, in Fig. 3 so loosely that you may turn the hour-and the minute-hand whenever you choose, and yet tightly enough to turn about with the wheel D if they are not disturbed. You can, therefore, move the two hands of the clock without disturbing any of the wheels in Fig. 3.
We have seen that the weight must keep pulling, or the clock will stop. Sometimes, instead of the weight, a spring is used, especially if the clock is small. The spring simply pushes the wheel A in the direction of the arrow (Fig. 3). When the spring is used the clock may have a pendulum escapement, or it may have a wheel escapement like that of a watch. But if the pressure of the spring is removed, or if the weight (should there be one) is lifted, the clock will stop. When you wind up the clock it is the same thing as taking away the weight, or the spring, while you are winding. How, then, can you wind it and still keep it going? This is done by what is called a "going-barrel," or "maintaining-works." In Fig. 3 you will notice that the wheel A turns in the direction of the arrow when the weight pulls down. When you wind up the clock the force of the weight is taken off. A strong spring is placed on the side of the wheel A that pushes it along in the direction of the arrow for the few seconds that you take in winding. Another wheel, or barrel, a, is placed on the large wheel A, and on this the string that holds the weight is wound. This wheel you turn in the opposite direction to that of the arrow. At the same time the spring pushes A in the direction of the arrow. You will sometimes see an old clock with an endless chain so arranged that, by pulling on a small weight, you may lift a large weight, and thus wind the clock. Others of the old time-pieces have weights that are hung by chains with rings at the upper end. When the weight has run down you can pull on the ring and the weight is lifted. You will find that all the best clocks, and all the watches, have the "maintaining-works."
The striking part of a clock is a very interesting study. It has a train of wheels and a weight entirely separate from the train that tells the hours and minutes by the hands. The large wheel, B, in Fig. 5, really consists of two wheels fastened together. The larger or outer wheel has seventy-eight teeth that run into a pinion, with thirteen leaves. The cord that holds the weight is wound on the axle of, on which A is also fastened. There are thirteen pins on the surface of A. They can not be seen, because they are on the other side of the wheel; but they have been drawn in the picture so that the explanation may be more easily understood. As the wheel A turns, each pin strikes the end of the lever c, which, when it is released, springs back and strikes the bell d. The smaller wheel, B, has notches all about it—first, one notch; then two notches close together; then three notches close together; and so on until you find twelve notches all in one place. This makes seventy-eight notches in all. Behind the wheel B is a pinion that you can not see. It is turned by the wheel A, but it is entirely independent of B, although it turns on the same axis. This independent pinion turns a wheel almost as large as B, which itself turns a small pinion that carries the "fly-fan." The use of the fan is to keep an even motion. The large wheel that we have spoken of turns once at every stroke of the bell. In Fig. 5, a wire, c, runs over to the center wheel, D. In Fig. 3, a pin on the center wheel pushes up this wire when the clock is ready to strike. If the end of the wire (in Fig. 5) rests at the four notches, it shows that four o'clock
has been struck. If the center wheel pushes the wire up again, or pulls it out from the notch where it is resting, the large wheels at B are released; the weight commences to turn A and B, and the pins in A set the hammer c to striking the bell d. It keeps on striking until five has been struck. The wire then drops into a notch and holds the striking-wheel fast until the center wheel moves the wire again—thus saying that it is time to strike six. The wheels then turn again until the wire comes down and stops them. Alarm-clocks have an arrangement by which the spring that sounds the alarm is let loose at the hour when the owner wishes to be awakened.
The boys who went to school in New England sixty years ago had no such device to waken them in cold winter mornings as the modern alarm-clock; they had to waken each other, in order to have a good start in kindling their fires, so that they could enjoy an hour's hard study, and sometimes a recitation, before breakfast. But it was not always convenient for one to keep awake in order to waken his companions. The one who was on guard was as sleepy as any of the rest: so the inventive brains of the Yankee boys were set to work to find some way of giving an alarm at the right time. Let it be remembered that, while primitive alarm-clocks were to be had in Europe, and while "repeating"-watches were a luxury in America, neither of them were to be found in New England as it was then. Even if the repeating-watch had been in general use, it was valueless, except to tell the time in the dark when one was awake. The invention of the alarm-clock was, therefore, a greater advance in the history of clock-making than was the invention of time-locks in the history of lock-making. The essential feature of the time-lock is a chronometer that turns a wheel containing a pin so adjusted that it will reach a certain point in a fixed time. Then a "dog" drops down, removes the obstruction, and allows the bolt to be shoved back. Two chronometers are used, so that, in case one runs down, the other will do the work. They are hung on springs, for fear that they will run down if the burglars should use dynamite, or some other explosive, to give them a sudden jar.
The Yankee boys, at the time that I have spoken of, were equal to the difficulty of awakening at the exact time. They invented a contrivance which was an indication of what was coming in both the alarm-clock and the time-lock. Indeed, it was so nearly a combination of the two that we must take away from the more modern inventors some of the credit and bestow it upon the boys.
In order to explain the plan more clearly, I ought first to say that the watches worn by both the men and the boys were of the large and coarse pattern known as "bull's-eyes"—a name given because the crystals were very thick, and bulged out something like the lens of a dark-lantern. The watches of this kind were not only very thick, but they were very large in diameter. The springs were very strong, and the hands were very stout. Therefore, the power that moved the hands was much greater than the power that moves the hands in the watches that are made to-day.
The boys prepared a board, a b c d, Fig. 6, about a foot square. Toward the upper edge, at e, they scooped out a place large enough for the watch to drop into, and have the face even, or flush, with the surface of the board. The face of the watch was then fastened to the board. The crystal was opened, or taken away entirely, and thus the hands traveled around just as if they had been on the board itself. A small wooden lever, j g, was fastened to the board by a nail, f, that acted as a fulcrum. Another lever, g i, had a fulcrum at h, and touched the first lever at g. The board was kept at a slant on the table by the prop n, or else by a pile of books behind it. The lever g j was so adjusted that the minute-hand of the watch would pass over the end, j; but when the given hour-hand, v, for instance, came round, it would strike j to the left. The effect would be this: g would move to the right and i to the left, thus pushing the weight at i from the little shelf on which it was balanced, and causing it to tumble toward the floor.
You can imagine that the force set in motion by the hour-hand of the watch, even of a "bull's-eye," was not enough to start a very heavy weight. Therefore, the dropping of the weight at i was not enough of a noise to awaken the boys, but the force that was exerted was enough, applied at the end of a long lever, to transfer itself to a point where it would do more good. The weight i, in dropping, pulled a string that was fastened on the long arm of the lever k m. This lever was fastened to the edge of the same table that held the square board by a gimlet, or nail, as a fulcrum, at l. When i dropped, it pulled k down and pushed m up. The sudden jerk at m pushed over a nicely balanced table, upon which had been placed nearly all the chairs and other furniture in the room. This certainly made enough noise to awaken the occupants of the room, and it is not likely there was much sleep after that. It was a great deal of trouble to adjust so nicely all the different parts of this primitive alarm-clock; but it never failed to work when care was taken with all the details. Let us praise the boys for studying out a scheme which others have adopted and called their own. They preferred to lie in bed as long as possible, and did not propose to keep awake all night, if any machinery could be devised to do the awaking for them.
A few words in regard to the dial on the face of the clock. The dial of a clock, if it is a cheap one, is made of wood and painted white. If the dial is small and expensive, it is made of copper on which is baked a white enamel surface. The figures are marked in black paint, which is sometimes burned or "baked in." The usual size of the figures from I to XII is one third of the distance from the outer circle toward the center. If the face of the clock is white, the figures and the hands should be black. If the face is black, or any dark color, the figures and hands should be either white or gilt. The dials of tower-clocks are frequently illuminated by gas or electricity, so that the time may be easily determined at night.