The Burlington Watch Company is an interesting figure in the commercial history of watches. In those days, watch companies sold movements, dials, and hands to middlemen, who sold them to jewelers. Movements and cases were sold separately, so the buyer picked a case and a watch, and the jeweler cased up the watch on site. Naturally, many mark-ups happened along the way, and small-town buyers were at the mercy of their town’s only jeweler.

Burlington, in contrast, was a direct-to-consumer company that sold complete watches—movement, dial, and case—by mail order. Their sales pitch emphasized value, and they had a good point: the traditional sales channel could not have been buyer friendly. Many of their early ads railed against the “trusts” and positioned the company as a “trust buster.” Watch companies like Waltham, Hamilton, and Illinois were clearly not “trusts” (monopolies)—they didn’t own the wholesalers or the point-of-sale outlets—but “trust busting” language must have resonated with price-conscious consumers during the gilded age, when monopoly power was a major cultural issue.

Burlington purchased movements from other companies, so they are considered a “private label,” much like Food Lion’s house brand of Cheerios or Costco’s Kirkland brand of everything. Nearly all of their watches were made by the Illinois Watch Company. Illinois made excellent watches, and Burlington focused on their middle range, so they are typically adjusted watches with good jewel counts (17 to 21). As a result, Burlington watches were a great value then and now: you can get a great watch for a better price.

We have a Burlington watch on the bench today. It is a 16-size, 21-jewel pocket watch from 1918. Like most of their watches, it was made by Illinois.

It’s a nice watch with some high-grade features: it is adjusted to temperature and 5 positions, and some of the balance wheel screws are gold. But it otherwise lacks the high level of finish that a 21-jewel Illinois from that period would have, which is part of the value equation for Burlington watches.

Reading the Omens

I got this watch from a local consignment store. They had a big batch of non-running pocket watches—including a factory-cased 12s Hamilton 912 that turned out well—at an uncommonly good price. The watch got a new white alloy mainspring and a thorough cleaning and oiling. It started right up, so after running for day and visiting the demagnetizer, it was ready for fine-tuning.

The omens for this watch are mostly good:

• Illinois watches from this period are excellent. It was originally adjusted to 5 positions, and nothing seems obviously wrong with it.
• The rates are clean and stable.

But on the other hand:

• This watch has been through many hands. A Burlington watch should have a Burlington dial and a Burlington case—these were sold as complete watches. This watch has a cheap metal Illinois dial and an unusually nice (but not original) swing-out case.
• There are a few bad signs. A couple dial screws were missing, and one screw fits but is clearly wrong (above the escape wheel). Missing dial screws are a good sign that the watch was once worked on by someone who couldn’t be bothered to fix something so fundamental and basic.
• Oddly, one of the mean-time screws is stuck—you rarely see this. This makes adjusting a bit trickier.

As an adjusting target, I’d like to get all the rates within +/-7 seconds from zero, with no two positions more than 10 seconds apart. That seems realistic for this watch.

For this watch, I’ll describe all the steps and changes in more detail than usual. A few readers have remarked that it would help to see how to reason about each change, so we’ll march through this watch step-by-step.

Baseline Readings

So what are we looking at here? These are the “Time 0” baseline readings. Keep in mind that pendant left (PL) and right (PR) are defined from the dial side, but we are working from the movement side. I set the regulator to the center of the fast/slow index and measured the rates:

Overall, this is not bad. The watch is running fast, but most of the vertical rates are similar to the DU and DD rates. And the DU and DD rates are basically the same, and aligning those is usually the most vexing part of adjusting, so this ought to be straightforward. But there’s clearly a poise error somewhere, given the big difference between the PR and PL rates.

The next step, of course, is to identify the heavy spot in the wheel. We do this with dynamic poising. As we have demonstrated and explained elsewhere, you can isolate the heavy spot by running the watch at a low amplitude, around 140 to 160 degrees. We want to find the fastest position: the heavy spot is directly below the balance staff when the watch is in that position.

Here are the readings at a low amplitude. The dot (.) means I didn’t assess that rate. You can measure the four main positions and then read the diagonals that flank the fastest rate.

Our fastest rate is shown in red—PR, which is this position.

Specifically, when the watch is PR, the heavy spot is directly below the balance staff, which is the red spot in the picture. The green spot, directly opposite, is the light spot.

So what change should we make? We need to align the vertical rates while also aligning the vertical and horizontal rates. Because the watch is running fast, we should add weight to the light spot (in green). This will improve poise and bring the horizontal and vertical rates closer.

So we raid our big stash of timing washers. Poise seems fairly good, so I added a relatively light washer. It was rated for a change of 60 seconds for a 12-size watch, so it should have a smaller effect on this 16-size watch. I placed it under the screw at the green spot.

While the wheel was out, I also set it in beat. The beat deviation was around 2.5 ms; with one tweak, I got it to .2 ms, an excellent outcome. This was part luck and part karma, I suspect.

Next Round

Afterward, I fully wound the watch to see what we hath wrought. Here are the new readings.

Clearly, they are much better. The 5 major positions (all but pendant down, PD) are fairly close. But we’re not quite to our goal (no deviation more than 10 seconds).

At this point, we’ll be making only minor changes, and it is easier to remove tiny amounts than add tiny amounts. And I do mean tiny. I thus adjust the regulator to make the watch run around -2 DU. If we remove a tiny amount of weight from a heavy spot, it will improve poise and make the watch run a tiny bit faster. This keeps our horizontal and vertical rates close.

Here are the low amplitude readings.

The heavy spot is between PL and PD, shown here.

Conveniently, there is a screw basically at that spot.

If we choose to remove weight when adjusting—not always the right choice—we want it to be imperceptible. For a watch of this size, I prefer screw slot files, which can slightly widen or deepen a screw slot.

This method was used by factory adjusters at the time, and it doesn’t appreciably change a screw’s appearance. And it is easy to remove a very small amount of weight with such a file.

And Another Round

After some very light filing, I fully wound the watch and set the regulator so the watch ran close to zero when DU. Are we done yet?

These readings are great, but we’re not quite done. Some positions remain more than 10 seconds apart. Note, though, that the 5 main positions are much closer than the 6th (PD) is. It would be okay to stop here, but we’ll do one more round. This will be a tiny change, so I set the watch to -1 DU and take the low amplitude readings.

Here they are—notice how close they are, which reflects the excellent poise of the wheel by this point. I doubt that static poising on an old-time poising tool would uncover a heavy spot for this wheel.

The heavy spot is below the staff when the watch is in this position.

Specifically, here it is, in red.

Notice that there is no screw at that spot. The heavy spot is directly between a mean-time screw (the one that could turn) and a regular timing screw. In this case, you can make a very tiny tweak to these spots, shown in the dotted circles. I very lightly filed the screw slot of the regular screw, and I very slightly rotated the mean-time screw inward.

As an aside, filing a screw slot should be light and imperceptible. Here are two screws, one of which was filed and has a slightly wider slot. Even at this magnification, it is hard to tell.

And the End

And now I think we’ll call it a day. Here are the final readings.

The change was minor, but it got us within our targets—no rate is more than +/- 7 seconds from zero, and no two positions are more than 10 seconds apart. The 5 main positions are very close. I am tempted to mess with it more, but way too many watches are waiting in the wings.

Our final step is to case it up, wear it for a few days, and then fine-tune it if needed. Casing a watch often makes adjustments drift a bit—adding the hour wheel and hands can reduce the amplitude, and the mainspring and lubricants might not have fully “bed in”—so it is worth revisiting the watch after a week to see what kind of time it keeps when cased and worn.

Overall, this watch is looking sharp. Perhaps I’ll find a period-appropriate Burlington dial someday for it.

Burlington offers some interesting historical parallels for modern watch collectors. The major Swiss and German brands are moving more of their retail sales into their own boutiques, which allows them better control over prices. Customers are right to wonder what value brand boutiques add to the watch. And Burlington’s business model—pair someone else’s movement with your case and dial, and market ferociously—absolutely dominates, especially among the lower and mid-range companies that use ETA or Sellita movements (e.g., Rado, Mido, Oris, Tissot, and the like).

The more things change, the more they stay the same.