Precision Clock Cases.—The casing of a precision clock is only secondary in importance to the compensation of its pendulum. The best construction of an efficient case can be ascertained only by most careful study of the conditions under which the clock is expected to be a standard timekeeper, and often the entire high accuracy sought by refined construction is sacrificed by an inefficient case and mounting.

The objects of casing a precision clock are as follows:

Now if we hold the above considerations in view we can readily see that cast iron, wood and glass, with joints of wash leather (which is kept soft by a wax cement which does not become rancid with age), are the preferable materials.

The advantages of using cast iron for the pillar or body of the case are that it can be cast in such a shape as to require very little finishing afterwards, and that only such as planing parallel surfaces in iron planing machines. It makes a stiff column for mounting the pendulum when it rests upon a masonry foundation from below. Plates of glass can be clamped against the planed surfaces of iron piers (by putting waxed wash leather between the glass and the iron) so as to make air-tight joints without difficulty.

The mass of iron symmetrically surrounding the steel pendulum is the safest protection the clock can have against casual magnetic disturbances. In the language of electricians it “shields” the pendulum.

Suppose, then, we adopt as the first type of precision clock case which our present knowledge suggests, that of an iron cylinder or rectangular box resting on a masonry pier, and which has a table top to which the massive pendulum bracket is firmly bolted. This type admits of the weights being dropped in small cylinders outside of the cast iron cylinder or box. These weight cylinders, of course, end in the table top of the clock case above and in the projecting base of the flange of the clock case below.

With this construction it is a simple matter to cover the movement with a glass case, preferably made rectangular, with glass sides, ends and top, with metal cemented joints. The metal bottom, edges of this rectangular box can be ground to fit the plane surface of the top of the clock case. Then, by covering the bottom edges with such a wax as was used in making the glass plates fit the iron case in front or back, we can secure an air-tight joint at the junction of the rectangular top glass case with iron case. In practice the wax to be used may be made by melting together and stirring equal parts of vaseline and beeswax. The proportions may be varied to give a different consistency of wax, and it may be painted on with a brush after warming over a small flame.

If the clock case will be exposed to a comparatively high temperature, say 95° F., then the beeswax can be 3 parts to 1 of vaseline. The good quality of this cement wax is that it does not change with age, or at least for several years, is very clean, and can be wiped off completely with kerosene, or turpentine, or benzine. In all joints meant to be air-tight, the use of rubber packing is to be avoided. It answers well enough at the start, but after several months it is sure to crack and leak air.

By an air-tight joint I do not mean a joint which will not leak air under any pressure which may be applied. It is not necessary that our pendulum should vibrate in a vacuum; all we want is that the pressure inside the clock case should be uniform; that it should not vary with the barometer outside. In actual practice we find it best to have the pressure inside the case as nearly as possible equal to the average atmospheric pressure outside. Now, if the barometer in a given locality never sinks below 27.5 inches, it is not necessary that the vacuum in the clock case be less than that represented by 29.5 inches of mercury pressure. So, too, if it were desirable to have the pressure inside the case greater than that outside, owing to some special form of joint which made the clock case less liable to leak out than to leak in, it might be that an inside pressure would be efficient at 31 inches of mercury. By not having the inside pressure vary but slightly from the outside, the actual pressure of air will not exceed one inch of mercury, or, say, pound pressure to the square inch. This is a pressure which causes quite an insignificant strain upon any joint.

There are objections, however, to the use of air in an enclosed space for precision clocks and so the attempt has been made to use hydrogen. Air is, comparatively speaking, heavy. It is 14½ times as heavy as hydrogen gas, for instance. The pendulum, therefore, in moving through its arc has to push aside 14 times as much weight as it would have to in case it were surrounded by hydrogen. Then what might be called the “case friction” is greater than if we used hydrogen. By “case friction” I mean resistance and a disturbance to the pendulum depending on the effect of the currents of air produced by driving the air before the pendulum against the sides and front of the case. It is a well-established observation that small, cramped cases disturb the clock’s rate more than large, roomy ones. This is because the air, having no room to go before the pendulum, is cushioned up against the side of the case at each pendulum swing, and acts as a resisting spring against the swing of the pendulum. By the time the pendulum has reached the end of its vibration the air has escaped upwards and downwards perhaps so that it no longer has its spring power to restore the loss of energy to the pendulum. This “case friction” is most pernicious in its action when associated with free falling weights in the clock case. Clock weights should always fall in separate compartments, and never in such a manner that they can affect the space in which the pendulum swings.

But this is a digression to explain the term “case friction” in its use in horology.

Precision clocks, almost without exception, have electric break-circuit attachments within the case. Most of these break-circuits are constructed so that there is a small spark every time the circuit is broken. The effect of such a spark in air is to convert a small portion of the air in the immediate neighborhood of the spark into nitrous acid gas. After several months there might be a considerable quantity of this gas in the case, with the certain result of rusting the nicer parts of the escapement.

Many attempts have been made to run a clock in an almost complete vacuum of air; but the volume to be exhausted is so large, and the leakage is so sure to occur after a time, that the attempt is now pretty generally abandoned. It will be inferred from what has preceded that a full atmosphere of hydrogen would only offer one-fourteenth the resistance to the pendulum that air would, and all the disturbances arising from the surrounding mediums would be only one-fourteenth for hydrogen of that which we would expect for air. Every consideration, therefore points to the use of hydrogen as the medium with which to fill our clock cases. It is inert, it forms no compounds under the influence of the electric spark, the case friction is no greater than would exist if we made an air vacuum of only about 1 inch of mercury, and hydrogen gas may be readily prepared. The method from dilute sulphuric acid and scrap zinc is the handiest, and it will be found described in almost any chemistry textbook or encyclopedia. Should the horologist wish to know something of the chemistry of the process, without previous study, he will find it described in very simple language in any primary chemistry. The practical details of filling a clock case with hydrogen gas I have not yet worked out. It is evident that since hydrogen is 14½ times lighter than air, that by attaching a small tube to the source of hydrogen and to the top of the clock case, and another small outlet tube at the bottom of the clock case, that by gravity alone the hydrogen would fill the upper part of the case and drive the air before it out at the bottom. The hydrogen should be dry. To insure this it should pass through a tube containing quicklime, which, if it is a foot long and two inches in diameter, will be sufficient. No burning light or electric spark must be put into the case while filling, because the mixture of hydrogen with the air is very explosive when ignited. Great care must be used in making all joints when attempting to maintain an atmosphere of hydrogen as it leaks readily through the pores of wood iron and all joints. It is, therefore, better to treat the case friction as a constant element and simply keep it constant.

The above discussion has not considered the temperature question. It is important that the changes of temperature in a clock case should be as slow as possible and as small as possible. Professor Rogers, of the Harvard College Observatory, has shown that such bars as are used in pendulum rods of clocks are often several hours in taking up air temperatures many degrees different from that in which they were swinging. We have at the top of the pendulum a thin spring for suspension whose temperature decides its molecular friction; then we have the pendulum rod, and lastly the large bob, all of which take up any new temperature with different degrees of slowness. Now obviously no compensation can be made to act unless the temperatures are the same for all parts of the pendulum, and vary at the same rate. A number of years ago, there was a long discussion as to the temperature at the top and bottom of clock cases. It was shown that this regularly amounted to several degrees in the best clocks. It was to lessen this difference that at the Harvard College Observatory the Bonds built a deep well in the cellar, purposing to put the clock at its bottom. The idea was a good one, and were it not for the difficulty in getting at clocks in wells, and keeping water out, it would doubtless find favor where the utmost accuracy is desired.

A better plan is to run the clock at a high temperature, say 95° to 100° F. The oil is more liquid, the temperature can be more easily maintained, it can all take place in lighted, dry rooms, and the means for doing this we shall now consider.

Our iron case must now be housed in another outside case, which had better be of wood, with glass windows for seeing the clock face. A single thickness of wood would conduct heat too rapidly. It must therefore be made of two thicknesses, with an air space between. If the air space is left unfilled, the circulation of the air soon causes the inner wooden layer to be of the same temperature as the outer. It is necessary to prevent this circulation of air therefore by means of some substance which is a non-conductor of heat and which will prevent the air from circulating. The very best thing to be used in this connection is cotton batting, which has been picked out until it is as light and fibrous as possible. Then if the doors and windows of the wooden case are made of two thicknesses of extra thick glass, and are firmly clamped, by screws through their sashes or some other means, to the frame of the case, we have the best form possible for our completed case of the type I have described. It now remains to provide a layer of hot water pipes inside the clock room, heated by circulating hot water from the outside. The flame under the water tank outside, whether of gas or kerosene, to be automatically raised or lowered by any such thermostat arrangements as are in common use with chicken incubators, when the temperature varies from the point desired. Experience teaches that the volume of water had better be considerable, if there is considerable difference in the annual variations of temperature according to the seasons. Thus in Massachusetts or Illinois the temperature is likely to vary from -30° F. to +110° F., and the heating arrangements must be suitable to take care of this variation.

The Waltham Watch Company’s clock room is an excellent example of the means taken to secure uniformity of temperature and absence of vibration.

The clock room, which is located in the basement of one of the buildings, is built with a double shell of hollow tile brick. The outer shell rests upon the floor of the basement, and its ceiling is within two or three inches of the basement ceiling. The inner shell is 10 feet square and 8 feet in height, measured from the level of the cellar floor. There is an 18-inch space between the walls of the inner and outer shell and a 9-inch space between the two ceilings. On the front of the building the walls are three feet apart to accommodate the various scientific instruments, such as the chronograph, barometer, thermostat, level-tester, etc. The inner house is carried down four feet below the floor of the basement, and rests upon a foundation of gravel. The walls of the inner house below the floor level consist of two thicknesses of brick with an air space between, and the whole of the excavated portion is lined, sides and bottom, with sheet lead, carefully soldered to render it watertight. At the bottom of the excavation is a layer of 12 inches of sand, and upon this are built up three solid brick piers, measuring 3 feet 6 inches square in plan by 3 feet in height, which form the foundation for the three pyramidal piers that carry the three clocks. The interior walls and ceilings and the piers for the clocks are finished in white glazed tiling. The object of the lead lining, of course, is to thoroughly exclude moisture, while the bed of sand serves to absorb all waves of vibration that are communicated through the ground from the various moving machinery throughout the works. At the level of the basement floor a light grating provides a platform for the use of the clock attendants.

Although the placing of the clock room in the cellar and the provision of a complete air space around the inner room would, in itself, afford excellent insulation against external changes of temperature, the inner room is further safeguarded by placing in the outer 18-inch space between the two walls a lamp which is electrically connected to, and controlled by, a thermostat. The thermostat consists of a composite strip of rubber and metal, which is held by a clamp at its upper end and curves to right or left under temperature changes, opening or closing, by contact points at the lower end of the thermostat, the electrical circuit which regulates the flame of the lamp. The thermostat is set so as to maintain the space between the two shells at a temperature which shall insure a constant temperature of 71 degrees in the inner clock house. This it does with such success that there is less than half a degree of daily variation.

The two clocks that stand side by side in the clock room serve to keep civil time, that is to say, the local time at the works. The clock to the right carries a twelve-hour dial and is known as the mean-time clock. By means of electrical connections it sends time signals throughout the whole works, so that each operative at his bench may time his watch to seconds. The other clock, known as the astronomical clock, carries a twenty-four-hour dial, and may be connected to the works, if desired. These two clocks serve as a check one upon the other. They were made at the works and they have run in periods of over two months with a variation of less than 0.3 of a second, or ¹/259000 part of a day. The third clock, which stands to the rear of the other two, is the sidereal clock. It is used in connection with the observatory work, and serves to keep sidereal or star time.

The rate, as observed at the Waltham works, rarely exceeds one-tenth of a second per day. That is to say, the sidereal clock will vary only one second in ten days, or three seconds in a month. The variation, as found, is corrected by adding or subtracting weights to or from the pendulum, the weights used being small disks, generally of aluminum.

Summing up, then, we find that the great accuracy obtained in this clock room is due to the careful elimination of the various elements that would exercise a disturbing influence. Changes of temperature are reduced to a minimum by insulation of the clock house within an air space, in which the temperature is automatically maintained at an even rate. Changes of humidity are controlled by the specially designed walls, by the lead sheathing of the foundation pit, by the preservation of an even temperature, and by placing boxes of hygroscopic material within the inner chamber. Errors due to vibration are eliminated by placing the clocks on massive masonry piers which stand upon a bed of sand as a shock-absorbing medium.

The astronomical clock is inclosed in a barometric case, fitted with an air pump, by which the air may be exhausted and the pendulum and other moving parts relieved from barometric disturbances. For it must be understood that variation in barometric pressure means a variation in the density of the air, and that the speed of the pendulum must necessarily be affected by such changes of density.

Restoring Old Cases.—Very often the watchmaker gets a clock which he knows will be vastly improved by varnish, but not knowing how to take off the old varnish he simply gives it a little sand paper or rubs it off with oil and lets it go at that. Varnishing such a clock thinly with equal parts of boiled oil and turpentine and allowing it to dry will often restore the transparency of the varnish; if uneven results are obtained a second coat may be necessary. Many of these old clocks have not been varnished for so many years that the covering of the wood looks like a cheap brown paint. To remove this in the ordinary way means endless labor, and if the case is inlaid with colored patterns of veneers, which are partly loosened by the glue drying out, the repairer is afraid to touch it for fear he will only make matters worse in the attempt to better them.

In the case of an old clock of inlaid marquetry, if the pieces of veneer have become partly loosened, the first thing to do is to make a thin, fresh glue. Work the glue under the veneer and then clamp it down tightly with a piece of oiled paper, or waxed paper, laid between the glue and the board used to clamp with and the whole firmly set down tight with screws or screw clamps. To make waxed paper dissolve parafine wax in benzine and flow or brush on the paper and let dry. After the glue has hardened comes the work of removing the varnish. To do this you will need some varnish remover, which can either be bought at the paint store, or made as follows:

Varnish Remover.—In doing such work the trick is to make sure that nothing put on the case will injure it, as a clock one hundred years old cannot be replaced. Therefore, if you are suspicious as to the varnish removers you can purchase, and do not want to take chances, you may make one of wood alcohol and benzole, or coal tar naphtha. Be sure you do not get petroleum naphtha, which is common gasoline. The coal tar naphtha is a wood product. The wood alcohol is also a wood product and the varnishes used upon furniture are vegetable gums, so that it will readily be seen that you are putting nothing on the antique with which it was not associated in its natural state. Equal parts of benzole and wood alcohol will dissolve gums instantaneously, so that if the oil has dried out of the varnish so much that the varnish has become opaque and only the rosins are left, the application of this fluid with a brush will cause instant solution, making the gums boil up and form a loose crust upon the surface of the wood, as the liquid evaporates, which it does very rapidly.

Varnishes containing shellac and some other gums are rather hard to dissolve and where an obstinate varnish is encountered it may be well to use wax in the varnish remover. This is done by shaving or chopping some parafine wax, dissolving it in the benzole, and when it is clear and transparent, add the wood alcohol. Upon the addition of the alcohol the wax immediately curdles so that the fluid becomes milky. In this condition it is readily brushed upon any surface and when the wax strikes the air it congeals and forms a crust which holds the liquid underneath and enables it to do its work instead of evaporating.

The wax also serves the purpose of allowing the workman to see just where he is putting his fluid and of holding it in position upon vertical surfaces or ceilings, round moldings, carved work and other places from which it will quickly run off. Only enough wax should be added to make it spread readily with the brush and after soaking it will be an easy matter to take a painter’s putty knife, a case knife, or a scraper and laying it nearly flat on the wood remove all the varnish at one operation, wiping off the knife as fast as it becomes too full. After the bulk of the varnish is off some of the fluid, without the wax, may be used upon a cloth to go over and smooth up by removing the spots and stripes of varnish left by the knife, or in moldings, etc., where the knife cannot be applied, and we have our bare wood, which, after drying and sand papering, is ready for a fresh coat of XXX coach varnish, which should dry in 24 hours and harden in a week.

A very little work and practice in this will enable the workman to rapidly and cheaply clean up and repair antiques in such a way that it will add greatly to his reputation.

To restore the gloss of polished wood it is not always the best plan to employ true furniture polish. The majority of the so-called polishes for wood are based on a mixture of boiled linseed oil and shellac varnish, made by dissolving shellac in alcohol in the proportion of four ounces of shellac to a pint of alcohol. A little of the dissolved shellac is poured on to a canton-flannel rag, a few drops of the boiled linseed oil are placed on the cloth, and the wood to be polished is rubbed vigorously. About half an ounce of camphor gum dissolved with the shellac in the alcohol will greatly facilitate the operation of polishing.

A soft woolen rag, moistened with olive oil and vigorously rubbed on dull varnished surfaces, like old clock cases, will brighten the surface wonderfully. Some workmen add a few drops of a strong solution of camphor gum in alcohol to the olive oil.

The polishing of cases is accomplished by applying several coats of the best coach painters’ rubbing varnish, when, after perfect drying, the surface is rubbed with a felt or a canton-flannel rag, folded flat, using water and the finest pulverized pumice-stone. This operation smooths the surfaces. The final polishing of such work is done by rubbing with rotten stone and olive oil with the smooth side of canton flannel. To remove the last traces of smear caused by the oil, an old, soft linen cloth and rye flour is used. Of course, fine work like we see on new cases of fine quality is not likely to be produced by one who is unaccustomed to it; a man must serve a good, long apprenticeship in the varnish finishing business before he is competent for it; and even then some polishers fail to obtain the fine results achieved by others. The great danger is that the rubber will cut through the varnish and expose the bare wood on edges, corners and even in spots on plane surfaces, before he has removed the lumps and streaks of varnish on adjacent portions of the work. Whenever the varnish is flat and smooth in any spot, you must stop rubbing there.

Black wood clocks which have become smoked and dull should have the cases rubbed with boiled oil and turpentine on a piece of soft woolen rag; afterwards polish off with a dry rag. If the gloss has been destroyed it will have to be varnished. Flow the varnish well on and use 1½-inch brush and be careful to get the varnish on even and so as not to trickle. This is easy if you are careful to keep the varnish thin and do not go over the varnish a second time after spreading it on. Thin with turpentine and put very little on the case; it is already smooth and a mere film will give the gloss. For white filling on the engraving on black cases use Chinese white or get a good white enamel at a paint store.

Gilding on wood cases is done by mixing a little yellow dry color with thin glue and painting the cases with the mixture; the color lets you see what you are doing. When the glue has dried until it is “tacky,” lay gold leaf on the painted portions and smooth down with cotton. If you have any holes do not attempt to patch them. It is easier and quicker to put on another sheet of gold leaf over the first one. After the gold is dry, it may be burnished with a bloodstone or smooth steel burnisher, or it may be left dead. Finish with colorless lacquer, very thin and smooth.

Imitation gold leaf, known to the trade as Dutch Metal, may be substituted for the gold leaf, if the latter is thought to be too expensive, but in such cases be sure to have the metal well covered with the lacquer, as unless this is done it will blacken in two or three years—sometimes in two or three months.

Bronze powder may be applied to the glue size with a tuft of cotton and well rubbed in until flat and smooth; then lacquer and dry. Never put on bronze paint, for the following reason: If we examine the bronze under a microscope we shall find that it is composed of flat scales like fish scales; if mixed as a paint they will be found lying at all angles in the painted work—many standing on edge. Such scales reflect the light away from the eye and make the work look dull and rough. If we rub these dry scales in gently on the sticky size, we will lay them all down flat and smooth, so that the work will glisten all over with an even color. Always lacquer bronzed work—yellow lacquer being the best—and put on plenty of lacquer.

Metal ornaments, when discolored, should be removed from the case, dipped in boiling lye to remove the lacquer, scratch brushed, dipped in ammonia to brighten, rinsed in hot water and dried in sawdust. They may then be lacquered with a gold lacquer, or plated in one of the gold plating solutions sold by dealers for plating without a battery and then lacquered, if bright. If they are of oxidized finish cleaning and lacquering is generally all that is necessary.

Oxidized metal cases, if badly discolored, should be sent to an electroplater to be refinished, as the production of smooth and even finishes on such cases, requires more skill than the clock repairer possesses, and he therefore could not do a good job, even if he had the necessary materials and formulÆ.

Marble cases are made of slabs, cemented together. Many workmen use plaster of paris by merely mixing it with water, though we rather think it better to use glue in the mixing, as plaster so mixed will not set as quickly as that mixed with water. After the case is cemented with the plaster, the workman can go over the joint with a brush and water colors, and with a little care should be able to turn out a job in which the joint will not be noticeable. Another cement much used for marble is composed of the white of an egg mixed with freshly slaked lime, but it has the disadvantage of setting very quickly.

Marble case makers use a cement composed of tallow, brick dust, and resin melted together, and it sets as hard as stone at ordinary temperatures.

It often happens that the marble case of a mantel clock is injured by some accident and its corners are generally the first to suffer. If the break is not so great as to warrant a new case or a new part the repairer may make the case a little smaller or file until the edges are reproduced, after which the polish is restored. Proceed as follows:

Take off from the damaged part as much as is necessary by means of a file, taking care however, not to alter the original shape of the case. Now grind off the piece worked with the file with a suitable piece of pumice-stone and water and continue the grinding next with a water stone until all the scratches have disappeared, paying special attention to the corners and contours. After this has been done take a hard ball of linen, moisten it, and strew over it either tripoli or fine emery and proceed to polish the case with this. Finish the polishing with another linen ball, using on it still finer emery and rouge. Now dry the case and finish the polishing with a mixture of beeswax and oil of turpentine. This method may be employed for all kinds of marble, or onyx and alabaster cases.

In cases where the fractures are very deep, so that the object cannot be made much smaller without ruining the shape, the damaged parts may be filled with a cement, prepared from finely powdered marble dust and a little isinglass and water, or fish glue will answer very well. Stir this into a thick paste, which fill into the deep places and permit to dry; after drying, correct the shape and polish as described.

If the pieces which have been broken off are at hand they may be cemented in place again. Wet the pieces with a solution of water and silicate of potash, insert them in place and let them dry for forty-eight hours. If the case is made of white marble use the white of an egg and a little Vienna lime, or common lime will answer.

To Polish Marble Clock Cases.—It frequently becomes the duty of the repairer to restore and polish marble clock cases, and we would recommend him to make a thin paste of the best beeswax and spirits of turpentine, clean the case well from dust, etc., then slightly cover it with the paste, and with a handful of clean cotton, rub it well, using abundant friction, finish off with a clean old linen rag, which will produce a brilliant black polish. For light colored marble cases, mix quicklime with strong soda water, and cover the marble with a thick coating. Clean off after twenty-four hours, and polish well with fine putty powder.

To Remove Oil Spots From Marble.—Oil spots, if not too old, are easily removed from marble by repeatedly covering them with a paste of calcined magnesia and benzine, and brushing off the magnesia after the dissipation of the oil; this may have to be repeated several times. Another recipe reads as follows: Slaked lime is mixed with a strong soap solution, to the consistency of cream; this is placed upon the oil spot, and repeated until it has disappeared. In place of this mixture, another one may be used, consisting of an ox gall, 125 grains of soapmaker’s waste lye and 62½ grams of turpentine, with pipe clay, to the consistency of dough.

Cutting Clock Glasses.—You will sometimes want a new glass for a clock. I get a lot of old 5 × 7 negatives and scald the film off in plain hot water, rinse well and dry. Now I lay my clock bezel on a piece of paper and trace around with a pencil, inside measure. Now remove the bezel and trace another circle around the outside of this circle about one-eighth inch. Now, lay the paper on a good, solid, smooth surface, glass on top, and with a common wheel glass-cutter follow around the outside line, free handed, understand. The paper with marked circle on is under the glass, and you can see right through the glass where to follow with the cutter. Now cut the margins of glass so as to roughly break out to one-half inch of your circle cut, running the cuts out on the side, then carefully break out.