What does the balance wheel do for accuracy? How do I set the hands on my watch and how do I wind it? What does "railroad-grade" mean?
Learn the basics of jeweling, configuration styles, movement sizes, winding and setting, dial construction, types of adjustments, watch mechanics, railroad standards, and more. There isn't nearly enough room on this page to cover everything, but it'll get you started.
Watch this video to see how the mainspring torque pushes through the gear train and the escapement to impel the balance wheel to oscillate. The design of the watch in this video is typical of nearly all American models during the railroad era, containing a mainspring, three-wheel gear train, escape wheel, lever pallet, and balance wheel assembly.
To see it full-size click the Enlarge icon in the lower right.
These are the primary components in most pocket watches, although the visual layout may change between models.
All watches from the railroad era used a balance wheel as a means of achieving accuracy. The earliest wheels were solid metal, while later models were bi-metallic with expansion cuts to adapt for changes in temperature and carried many weights that added mass to the wheel.
Its steady rate was controlled by an oscillating hairspring, and nearly all American models had some kind of regulator for fine adjustments.
Most American brands used at least one pair of these special screws to easily make bulk timing changes. They are usually found near the ends of the balance wheels arms and were made with a special interference thread to prevent them from turning without force. Any company that failed to install them on their watches (Elgin was one) invited a century of jewelers to molest the other weights to make timing changes.
When a watch inevitably slowed from thickening oils and carried no mean-time screws to make adjustments, the lazy choice was to simply grind off part of the balance weights to gain time. At some later point the watch was cleaned properly, which then ran way too fast because of the missing weight, so washers had to be installed to add mass - all of which thoroughly ruined the factory poise, or positional accuracy.
The hairspring is the rebound mechanism for the balance wheel, "breathing" in and out five times every second. The coils of a flat spring all lay on the same plane, anchored at either end. At the center is the collet, a brass split-ring that grips the balance staff, and at the tail is the stud, produced in many different styles, depending on the brand.
Prussian watchmaker Abraham Breguet is credited for the invention of the overcoil hairspring, which raises the final turn of the hairspring above the other coils. Whereas a flat hairspring applies lateral stress (and friction) to the balance staff when expanding or contracting, the overcoil minimizes this effect by moving a much smaller distance.
Any watch using an anchor or lever escapement requires a roller jewel, which is mounted in the roller table using shellac. On every cycle of the balance wheel this jewel gives an impulse to the pallet fork via the cup located at the end of the pallet arm. A vertical guide pin behind the cup prevented the pallet fork from drifting too early by meshing with a crescent-shaped cutout milled in the edge of the roller table.
Overbanking occurs when the roller jewel becomes misaligned with the pallet cup, which was possible with the earlier single roller design.
The double roller was developed to address this, using a smaller lower table with the cutout to mesh with the now-horizontal guard pin. This allowed the pallet to swing only when the roller jewel was in the cup, eliminating overbanking. The jewel is still mounted in the upper table.
The most common, found on most brands.
Used in hanging barrels with internal hooks.
Used on most Elgins with slotted barrels.
A watch barrel contains the mainspring and the cover for a going barrel snaps on tightly, so the spring must uncoil in that rigid space. Newer alloy mainsprings have sharp edges that will snag on anything inside this type of barrel, including any stamped numbers, making for an erratic letdown.
This is the most common design, using a slotted or T-end spring.
A hanging barrel consists of a barrel and cover that never touch each other, rotating on a common arbor and giving the spring more room. Waltham used this design on several of their models. The mainspring is allowed to unspool at a much smoother rate, producing more consistent torque.
Hanging barrels have an internal hook to anchor the mainspring.
A motor barrel has a pair of internal jewels between the winding arbor and the barrel and its cover. This increases accuracy by allowing the mainspring to unspool without friction, and when used with a hanging barrel makes for the smoothest possible letdown.
External jewels on the plate are for show and do not reduce friction or improve accuracy, though they make winding the watch easier.
The stopworks, or Maltese Cross, originated in Geneva, Switzerland in the 1700s as a means of limiting mainspring travel in pocket watches.
The device was located on the mainspring barrel, and consisted of a single-tooth drive wheel mounted on the mainspring arbor that mated with a driven wheel with a single closed spoke and several open ones. After a specific number of rotations the closed spoke would lock out the drive wheel and prevent further turning of the mainspring.
The purpose of stopworks was isochronism, the ability of a timepiece to run at the same rate regardless of changes in its power source. The red line in the graph represents the drive torque of a steel mainspring, and the stopworks allowed the watch to run only in the middle range. This was typically two-thirds of the entire linear force, shown in green.
Sadly, most stopworks have been tossed by lazy hacks as unnecessary.
A fully-jeweled railroad pocket watch contains seventeen jewels - a total of seven for the balance assembly and a pair for each piece of the gear train for ten more. In 1891 the Hampden Watch Co introduced the first 23-jewel watch in America, and the jewel race was on. Factories added jewels anywhere they could put them, like the mainspring arbor and unnecessary caps on the pivots.
Learn what jewels were made from, what they were for, where they were used, how they were mounted, and most importantly - how to determine jewel count just by looking at the watch.
Open face was by far the most common, with the winding stem at the 12:00 position, and was available in every size that American watch companies had to offer. This configuration eventually became a requirement as one of the primary railroad criteria.
There are four primary styles for open-face cases: threaded covers, snap-on covers, internal swing-ring (dustproof), and hinged clamshell.
Hunter (hunting) placed the stem at the 3:00 position, and was not railroad approved. The cases had a front lid to protect the dial and crystal, which opened by depressing the crown, along with several hinges, a dust cover, and springs for the front cover and the latch.
There is no such thing as a double hunter or a half hunter.
A case is either hunter or it's not, though demi-hunters have numerals on the front lid and a small crystal to view the position of the hands.
Simply put, sidewinders are hunter movements in open-face cases. Seldom purchased this way historically, this arrangement means the original gold-filled hunting case has likely been melted down.
Most collectors believe this configuration to be incorrect.
Conversion dials allowed hunter movements to be used in an open-face case by moving the seconds bit to the 3:00 position, returning the winding stem to 12:00. These began showing up after WWI, were usually made of metal or melamine, and could be railroad acceptable.
Reversible cases had a center ring that rotated within the case frame, allowing the movement to be displayed as if it were either a hunter or an open-face style. However, this does not convert the movement, since the winding pinion doesn't change positions. For example, an open-face movement with the winding stem at 12:00 used as a hunter profile will simply be laying on its side when displayed that way.
These are commonly called a "Muckle" case, after inventor E A Muckle.
No single part on an American pocket watch was as visible as the dial, and they came in an incredible array of colors, designs and fonts. See all the familiar variants in one place, learn what dials were made from, how they were mounted, discover who applied for patents, and don't forget the section on hand styles and colors at the bottom of the page.
Pendant (or stem-set) was the most common, even though the design required twice as many parts as a lever-set variant. The hands are set by popping up the crown, which did not meet the railroad standards of the day because the hands could accidentally be reset. The watch is wound by twisting the crown in a ratcheting motion.
Lever-set watches did meet railroad standards, as long as it was an open-face model. Setting the time meant removing the bezel and pulling out a small lever, which could be in several different locations around the dial rim. The watch still gets wound by twisting the crown.
Key-wind and key-set watches also didn't meet railroad requirements. The watch was wound with a small square key through the back and the hands were set by removing the bezel, although a few were able to also set the hands from the back, such as Waltham's Model 70.
Pin (or nail) setting was largely a Swiss innovation and was seldom used on American brands. It was an awkward arrangement, requiring the pin to be held in while turning the crown to set the hands.
All American companies offered full plate designs, with the balance wheel being the only visual part of the gear train. The balance wheel was recessed on some models, but was always above the upper plate.
A 3/4-plate design protected the balance wheel between the plates. Some models split the upper plate between the mainspring bridge and the gear train bridge, while others had a single plate to cover both areas.
A bridge movement also stacked the balance wheel inside the gear train, with as many as four separate upper plate bridges. A "false" bridge model was actually a single piece with shallow milling cuts to mimic the real thing.
Gilded plates were the most common prior to 1880, with a thin layer of gold on bronze.
Nickel-plated steel was an ideal choice to mill patterns into, called damaskeening.
Two-toning required the masking of parts of the nickel plates with wax and then gilding.
Similar to two-toning, except the pattern was milled first before the plates were gilded.
Waltham first pioneered this look, with the pattern milled after bead-blasting the plates.
US Watch of Waltham was the first to etch a pattern with acid on mirror-smooth plates.
American watches used the Lancashire gauge, which is based on the dial plate of a O-size watch measuring precisely a standard inch in diameter as a starting point, with each ascending size adding exactly 1/30th of an inch. The additional 5/30 inches was for the mounting shoulder up to a 16-size watch, when the diameter was increased to 6/30 inches.
Different countries had their own system, adding to the confusion.
The verge dates to the 1200s and used a pair of steel paddles mounted 90 degrees apart on the balance staff to alternately lock and release the crown wheel. It was bulky and the mating surfaces wore quickly, although they could be made very small. Combining the verge with a fusee as a stable power made the first pocket watches possible.
The fusee was a very simple yet effective method of equalizing drive torque as the mainspring unspooled and lost power. The design dates back to the 1400s, with the earliest versions using gut or wire. When the watch is fully wound, the chain is coiled entirely around the fusee cone, and the mainspring is applying the largest force with the smallest amount of leverage. As the mainspring unwinds, the chain begins to wrap around the mainspring barrel, and its weakening pull is applied to the expanding spiral radius toward the bottom of the fusee cone.
Verge fusees inevitably speed up over time because of the design and the wear of the mating surfaces. The brass teeth of the crown wheel (see above) become flattened out against the steel paddles on the balance staff, reducing the travel of the balance wheel. This decreases the amount of time between impulses, causing the watch to run faster.
Accuracy was the most important aspect of any watch, so high-grade movements were adjusted to keep better time. These adjustments offset the effects of friction, gravity, temperature and the fading torque in an unspooling mainspring by fine-tuning the balance assembly, and getting a watch to run accurately took many hours of careful work by a skilled watchmaker.
The word means "same time" in Greek, and is the accuracy of a watch whether the mainspring is fully wound or almost spent. Barrel stopworks, counter-balanced pallet forks, and longer hairsprings with overcoils all helped to equalize mainspring torque in pocket watches.
Changes in temperature affected both the hairspring and the oils in a watch, altering its accuracy. Factory adjustments took days, with the watches kept running in cold boxes near freezing and hot boxes at 90°F. Bi-metallic balance wheels with expansion cuts, new synthetic oils, and the invention and use of alloy metals like Elinvar in hairsprings in the 1920s all helped to offset this.
A watch can be oriented pendant up or down, pendant left or right, or the dial up or down. There are nine total adjustments - the six positional ones, isochronism, plus one each for heat and cold. When a movement was marked with a given number of positions it was understood to mean certain specific physical orientations, although this was not universal, and it was assumed that isochronism and temperature adjustments were included. Some companies simply milled the word "Adjusted" on the plates with no explanation of what that actually meant, while other firms inscribed the total number of adjustments to which the watch was timed against.
The growing country moved nearly everything by rail in the 1880s, so an accurate timetable between freights became increasingly important. For decades the railroad companies had generally accepted any 18-size 15-jewel watch for service, but in 1891 the terrible Kipton, Ohio crash killed several people because an engineer's watch had stopped, and the rules changed.
Webster Clayton Ball, a successful jeweler and store owner, was given authority as Chief Time Inspector, overseeing thousands of miles of track. New standards were adopted, and by the turn of the century a Railroad Grade watch was American-made, open-face, lever-set, either 16 or 18-size, had a minimum of 17 jewels and was adjusted for temperature and positions. It had to be equipped with a steel escape wheel, a micro-regulator, and a bold-font Arabic dial. Railroad workers were required to submit their watches for regular servicing and inspection, and the most important criteria was that any watch had to be accurate to within 30 seconds a week.
Earlier American watches ran on a "slow" gear train, designed for precisely 4.5 beats-per-second (BPS). Multiplying 4.5 BPS by 3,600 seconds in an hour returns 16,200 beats-per-hour, or BPH.
In later years the balance rate would rise to 5 BPS - or 18,000 BPH - to help increase accuracy. That figure translates to 432,000 beats per day, 3,024,000 per week, 12,960,000 per month, and 157,680,000 per year.