The Model 22 had a traditional main spring, available in a traditional double box and also in a deck watch.

The Model 21 and 22 had a two-day power reserve and the movements of both were marked U.S.

Navy Bureau of Ships.

The model 22 was also used by the U.S.

Army and on the back of some it is marked U.S.

Army, but all the model 22 movements are marked U.S.

Navy Bureau of Ships.

As the Second World War loomed, the United States realised that its navy was going to need a great many chronometers.

The United States scarcely had a chronometer industry, and such as were made used parts imported from overseas, mainly from Great Britain.

However, the belligerents had no spare capacity to produce chronometers for others, and Switzerland, the other main maker of chronometers apart from Germany and Britain, was in a difficult position: their German neighbour in effect forbade them to sell chronometers to the Allies.

The Hamilton Watch Company stepped up to the mark and began delivery of their Model 21 box chronometer with detent escapement in April 1942; By the end of the war, they had delivered nearly 9000 of these very fine instruments.

However there was also a great shortage of deck watches, both for larger ships for transferring time from the box chronometer(s) and for small vessels where a chronometer-rated watch had to serve as the principal timekeeper.

The Hamilton watch company began delivery of their Model 22 watch in June 1942.

By war’s end, 13,531 gimballed watches and 9780 non-gimballed watchers had been delivered.

The non-gimballed watch was in the form of a large pocket watch in a rectangular padded wooden case, while the gimballed watch was contained in a small three-part cubical case, like the M21 chronometer’s but smaller at about 51/2 inches on side.

A total of 9,780 non-gimballed (Figure 1) and 13,531 gimballed watches (Figure 2) were made.

At first sight, the M 22 watch mechanism looks like a large ordinary lever-escapement watch of high quality (Figure 3).

Its pillar plate is 2 1/4 inches (57 mm) in diameter.

However several features distinguish it from previous navigating watches: its motive power is an exceptionally long mainspring in a going barrel; it is jewelled back to the centre wheel; it has Hamilton’s ovalising balance; it has an Elinvar balance spring, whose elasticity varies very little with temperature; it has a safety setting button so that the hands cannot be accidentally set while winding and it has a hand to indicate its winding state.

^[0]

At first sight, the M 22 watch mechanism looks like a large ordinary lever-escapement watch of high quality (Figure 3).

Its pillar plate is 2 1/4 inches (57 mm) in diameter.

However several features distinguish it from previous navigating watches: its motive power is an exceptionally long mainspring in a going barrel; it is jewelled back to the centre wheel; it has Hamilton’s ovalising balance; it has an Elinvar balance spring, whose elasticity varies very little with temperature; it has a safety setting button so that the hands cannot be accidentally set while winding and it has a hand to indicate its winding state.

Motive power

Box chronometers are fitted with a fusee so that a more-or-less constant power is delivered to the escapement.

If the balance is isochronous, that is to say it takes as long to make a full swing whether the arc of the swing is large or smaller, then constancy of power delivery to the escapement becomes of less importance.

As will be seen later, the balance has design features that make for isochronism and this, combined with the advantage of the watch being maintained face-up at all times, meant that a going barrel was fitted, despite its non-uniform delivery of power.

Nevertheless, the M22 was fitted with a mainspring five feet (1524 mm) long which gave a power reserve of over 56 hours, though usually navigating timepieces were wound every 24 hours at the same time of the day.

Figure 4 shows in blue how the power declines in, say, a pocket watch going for 36 hours.

In the first few hours, there is a fairly steep decline in power which then tends to level off, followed by a steep decline in the last few hours.

The effect of lengthening the mainspring so that the watch runs longer is shown in red, with a less steep initial decline, and a lesser rate of loss of power thereafter.

Another advantage of a going barrel is that there is no need for maintaining power during winding.

The spring fitted to the M22 was 4 mm wide and 0.195 mm thick, with thicknesses of 0.190 and 0.20 also being available as required.

The train

With the exception of the going barrel and its integral great wheel, the whole train was jewelled with the escape wheel having endstones (Figure 3).

The purpose of increasing the jewelling beyond the usual seventeen is to reduce frictional losses in the train and to further enhance constancy of power delivered to the escapement.

The escapement

This is a pallet lever escapement, jewelled with endstones and otherwise unremarkable in design.

The balance

It is in the balance that the M22 differs from usual practice of the time.

It uses an ovalising balance of a similar design to that of the M21 box chronometer with a Hamilton Elinvar balance spring with Breguet over-coil.

Figure 5 shows on the left a typical pocket watch balance wheel and on the right the M22 balance wheel and spring (the scale is in millimeters).

Unlike an ordinary steel balance spring, one made of Elinvar has an elasticity that decreases only very little with temperature, so that correspondingly less compensation is demanded of the balance wheel.

In the ovalising balance, the arms are made of Invar, which expands scarcely at all with rise in temperature and the rim is made of 18/80 stainless steel.

At some temperature, the rim will be circular.

With a rise in temperature, expansion of the rim will force it into a slightly oval shape, with the long axis at right angles to the arms, and the moment of inertia, together with the period of oscillation, will increase.

Conversely, a fall in temperature will cause the balance to become oval with the long axis aligned with the arms and the period of oscillation will fall, since the distribution of compensation and timing weights tends to be concentrated in the rim.

In a normal bimetallic balance with a steel spring, the stiffness of the spring does not decrease in a linear manner with increasing temperature, while its change in diameter is linear, giving the so-called “middle temperature error, but in the M22 balance, such small changes in elasticity and moment of inertia as do occur are practically linear, so that the middle temperature error is reduced to an almost undetectable fiftieth of a normal compensated balance and is opposite in sign.

There are 28 holes for temperature compensation screws around the rim.

Hamilton provided a procedure for adjusting these screws, by checking the rates at 55 and 90 degrees and moving screws around the rim according to a table which gave the changes in position necessary to correct a given plus or minus change in rate in seconds per day per 35 degrees Fahrenheit.

There are also four nuts for poising after such a change and the daily rate was coarsely adjusted by the two pairs of timing nuts, with final rating being carried out using the regulator.

The balance spring

The Elinvar balance spring itself was pre-formed and heat-treated on a former so that all springs were pretty well identical and no tedious (and perhaps somewhat intuitive) hand adjustment of the overcoil was necessary for isochronism.

The central collet was counter-poised to allow for the weight of the pin and the asymmetry of the first coil.

The form of the counter-poise is visible in Figure 5.

The balance stud was pentagonal, with a hole of matching shape for it in the balance cock.

The Regulator [4]

As in most fine watches, there is a micro-regulator to allow very small movements of the regulator for final setting of the rate.

This is shown in Figure 6.

The regulator cam and index plate move together about a common axis and the end of the regulator is held against the cam by the regulator spring.

Each division of the index plate represents a change in rate of about 2 seconds per day, so very precise regulation is possible.

After I overhauled the watch shown in most of the figures, it had a mean daily rate at room temperature in summer of -1.96 seconds per day with a mean deviation from the mean of ±0.93 seconds

Safety setting

When it was important never to “lose the time”, as for example on railroads and for navigation out of sight of land it was usual to provide some way of preventing accidental re-setting of the hands when re-winding.

The American railroad watches were most often lever-setting, that is to say the front glass and bezel had to be unscrewed to expose a tiny lever, which had to be pulled out before the hands could be set (Figure 7), while box chronometer hands were usually reset, to Greenwich Mean Time, at the conclusion of a voyage.

The M22 had a safety setting button to the left of the winding button (Figure 8). ^[0]

The mechanism of the safety setting is perhaps of some interest, but to help in understanding it a few words about winding and setting generally may be useful.

When the winding stem is in its normal position (Figure 9) and is rotated, a square formed on it and passing through the clutch causes the latter to rotate and, since the clutch is engaged with the winding pinion, it too rotates with the winding stem.

Its teeth are engaged with the winding wheel which in turn causes the mainspring arbor to rotate and the spring is wound up.

A pin in the setting lever engages with a groove in the stem.

When the stem is pulled outwards, the pin causes the setting lever to rotate, so that a shoulder formed on it pushes the clutch lever inwards and the clutch is disengaged, while a pinion formed on the other end of the clutch engages with an intermediate setting wheel through which the minute wheel and canon pinion are caused to rotate, thus resetting the hands.

The winding pinion, being disengaged from the clutch of course remains stationary during setting.

Note that when the setting lever moves inwards, it slips from a groove in the setting cap spring and that the spring has a projection that nearly closes a gap between it and the mounting base of the spring.

Referring now to Figure 10, which has the safety setting lever in place, I have indicated with a red disc the position of a pin on the underside of the lever, which pin normally occupies the gap just mentioned.

While the pin is in position, the setting cap spring cannot move and so the setting lever cannot rotate prior to the clutch being disengaged and the hands turning, and the stem cannot be pulled outwards..

When the safety setting pin is depressed, the “red” pin moves out of place, and the setting lever becomes free to rotate as the stem is pulled outwards

Figure 11, taken from the Bureau of Shipping overhaul manual may help to make matters clearer.

Winding indicator [4]

In an ordinary box chronometer the mechanism to drive the winding indicator is simple: a pinion which rotates with the fusee as the clock runs down, is simply geared to the indicator hand.

As the clock is wound it moves one way and as it runs down it moves the other.

However, matters become more complicated when there is a going barrel, for when being wound, it is the arbor that rotates, but when running down it is the barrel that rotates, so that no simple gear train will suffice.

Figure 12 shows the solution adopted by Hamilton.

As the winding wheel rotates during windup it causes the windup gear of the planetary gear cluster to rotate clockwise and with it the differential sun gear (lower left of diagram).

Since the carrier gear is held stationary by the near-motionless barrel gear, the upper planetary pinion rotates together with the lower planetary pinion which then causes the alternating pinion to rotate clockwise.

The alternating pinion is geared to the windup indicator via the reduction gear and the wind indicator also moves clockwise (when viewed from behind, as in the diagram) to indicate “Up”.

As the clock runs down (lower right of diagram), this time it is the windup and differential sun gear that remain stationary while the barrel gear rotates clockwise.

This causes the carrier gear to rotate anticlockwise and, as the upper planetary gear rotates around the sun gear it causes the lower planetary gear also to rotate.

This tends to drive the alternating pinion clockwise, while rotation of the lower planetary pinion about the axis of the sun pinion tends to move the alternating pinion anticlockwise.

The latter dominates with the result that the alternating pinion moves anticlockwise and the winding indicator moves towards “Down”.

Figure 13 shows this tiny gear cluster from two viewpoints.

Case and gimbals [4]

The three-part case is of the usual mahogany finish but without the brass corners and bindings.

The corners are rebated and a lock is provided (Figure 14).

Note that the low serial number of 525-1941 places it in 1941, while the 1941 on the face of its larger brother refers to the date of the contract for design rather than date of manufacture.

The unit price for the M22 in 1941 was $92.33.

Figure 15 shows the gimbals.

Because of the winding stem, the ring has to be suspended fore and aft rather than the more usual athwartships.

The reason for the large piece of brass at the front rather than the more usual screw, as at the back, is presumably due to the presence of the lock, which had to be bridged.

The gimbals lock is conventional.

The back of the gimballed watch is a substantial lump of brass which screws into the body of the case (Figure 16) while the front bezel screws on to hold the movement in the case.

Except on the repair bench, it is unwise to remove bezel since inadvertent inversion may decant the movement on to the floor!

^[0]

The Model 22 was also available in a deck watch (a bit more than an oversized pocket watch).

^{[2] [1]}