Deciphering a 7-Digit Mismatch
A remarkable mismatched serial number error on a 1976 $2 Federal Reserve Note reveals how a mechanical failure created one of currency collecting’s most intriguing mistakes.
My favorite error type, and not coincidentally, the first error note I ever owned, has always been the Mismatched Serial Number. My census lists over 250 varieties, encompassing both large- and small-sized currency, through almost the full range of types/classes, denominations, and series.
Recent discoveries of a handful of LEPE-created errors with multiple mismatched digits, including non-numeric symbols and/or missing digits, and the 7-digit mismatches on Series 1995 $1 Federal Reserve Notes rekindled my interest. I searched through my collection of reference materials, photographs, and handwritten notes for information about a notable $2 FRN 7-digit mismatch that I had heard of more than three decades earlier to facilitate the writing of this article. Unfortunately, I found only a single footnote describing it as merely “$2 FRN—1976 F700A range—7 digits. NOT FOR SALE.” I put the article idea on the back burner until this note resurfaced earlier this year.
The first thing I do when I encounter an obscure, mismatched serial number error is try to determine the root cause. Certain attributes can normally point to the precise malfunction. For instance, a mismatch in the first four (left-to-right) digits often points to an operator setup error, while a mismatch in the last four numbers strongly suggests a mechanical failure.
I then apply a bit of math to determine the plate position for each of the two serial numbers on notes with substantial numeric differences. If these plate positions differ, the one that corresponds to the actual plate position on the note will show which serial number is correct.
In this case, since the highest serial number of the Series 1976 $2 Federal Reserve Note for the Atlanta district was F 608 00000 A, and the left-hand serial number is substantially higher, we know the right-hand serial number was correct. This was further substantiated when the sheet-position math applied to that serial number confirmed the H3 check plate position. Then, some simple subtraction of the serial numbers can show a pattern, especially if more than one note from the run is known.
Subtracting the right serial number (132 70088) from the left (798 36978) resulted in 665 66890. At first glance, the difference seemed arbitrary. However, a digit-by-digit comparison eventually revealed the underlying mechanical skip that occurred during overprinting. By subtracting each number on the left from the corresponding one on the right, the differences were: -6 | -6 | -6 | 4 | -6 | -9 | 1 | 0. As a Quality Control Engineer who relies on statistical analysis daily, these negative numbers immediately stood out to me. To understand their significance, it’s important to analyze the interaction between the independent numbering heads and base-10 serialization logic.
In a base-10 mechanical system like this, negative numbers indicate a “carry the one” failure between two adjoining wheels because one of them is a 0 but actually represents a 10. Adjusting these negatives by adding ten to any results of zero or less changed the right serial number to 1|3|2|7|10|10|8|8, and the numerical offset became 4 | 4 | 4 | 4 | 4 | 1 | 1 | 0.
This discrepancy in serial numbers suggests that the Tens and Hundreds wheels were engaged simultaneously, and that all five wheels from the Thousands to the Ten-Millions place were also locked together. A mechanical jam likely caused each of the two groups of numbering wheels to move or skip together as a single entity rather than advancing independently.
Although the Series 1976 Atlanta district was printed as complex blocks (on both COPE and conventional overprinting presses), the run of 20,000 sheets (Run 95) on which this note was printed was done on a COPE press. While digits could (and did) lock together in conventional presses, the mechanical linkages were simpler than those in the integrated COPE units. These failures would more often than not manifest as a single wheel failing to turn, creating a simple 1-Unit or 10-Unit difference that might only last for a few sheets before correcting itself or being caught by the more frequent manual inspections.
The COPE press utilizes two separate numbering blocks, each containing eight individual wheels, one for each digit in the serial number. These wheels are advanced by a ratchet-and-pawl mechanism. The difference of 44,444,110 on this note provides a clear picture of the malfunction:
• Ganged Wheels: The repeated difference of “4” across the higher-order columns (first five digits) strongly indicates a mechanical binding. If an actuating pawl—the mechanical lever that clicks the next wheel forward—slips, breaks, jams, or becomes misaligned, it can cause a lockstep failure, making multiple wheels move together and turn in unison, jump several numbers at once, or fail to turn.
• The Number 11 and the Ten-Unit Multiplier: When analyzing errors involving the serial numbers, keep in mind that the BEP serializes notes in descending order, from highest to lowest. This ensures that after slicing and stacking the sheets, the note with the lowest serial number ends up at the top of the strap.When two numbering wheels lock together, the number 11 (and its multiples) becomes the mathematical constant of the error. To understand why the 11-logic dominates these errors, you need to understand the difference between how we read a serial number and how the machine counts it.
Because the numbering wheels interact in a 10:1 ratio, the Units wheel must complete 10 rotations to cause the Tens wheel to turn once. If these wheels physically lock, they turn at a 1:1 ratio. Every time the Units wheel moves one position, the Tens wheel also moves one rotation. Since they move together, the distance between the two digits (11) remains fixed.
The mismatch between the correct side and the locked side grows in predictable increments (11, 22, 33, etc.) with each note printed. This logic applies to any cluster of locked wheels. If the Hundreds and Thousands lock, the difference jumps by 1,100 per note. If three wheels lock together, the factor becomes 111.
Occasionally, the expected sequence or numerical differences will vary. The letterpress process involves considerable mechanical force and pressure. The numbering head cylinders exert sufficient force on the paper to induce slight debossing and rotate at several thousand revolutions per hour. Internal components, such as the springs that maintain the pawls and the detents that prevent wheel wobbling, are subjected to constant tension and vibration as they function to ensure precise indexing with each new sheet. This may result in the wheels slipping partially (stuck digit), skipping multiple turns, or jumping backward before re-engaging, leading to an inconsistency and non-repeatability of the number gap. This phenomenon has been observed in various mismatch runs where multiple discrepancies within a serial number range are known to occur.
With only a single note, it’s impossible to determine whether the numbers remained static, skipped at times, or continued to advance together. Also, the relationship between the first five digits and the next two can’t be determined, as they show different variations, even though both follow the Number 11 logic.
While it is possible to hide physical evidence of tampering—such as chemical altering, disruption of paper fibers due to abrasion, or slight differences in ink color—in lower-grade (heavily circulated) notes like this, the base-10 math (especially the carry-the-one negative numbers) of the serial numbers, and the way the differences in the numbers match known instances of confirmed locked-wheel examples, strongly support this note’s validity.
Postscript: After this article was submitted, a second example of this error surfaced. That note, bearing Serial Numbers F 576 14756 A on the left and F 132 70056 A on the right, provides the missing link needed to determine the relationship between the first five digits and the last three.
The right-hand serial numbers, which are known to be correct, differ by 32, confirming that the two notes are 32 notes apart within the run. Since the numbering wheels were locked, the mismatch pattern follows the same Number 11 progression described earlier, with the pattern repeating every 10 notes as the wheels cycle through their base-10 positions. Over a span of 32 notes, this produces three full cycles with an additional offset, which brings the left serial number of the discovery note into alignment and shows that all eight digits advanced together.
To further confirm this, subtracting 576 14756 from 798 36978 yields 222 22222. This represents two full increments of 111 11111 and matches the expected pattern for a fully locked set of eight numbering wheels.
See more serial number error notes:
