Inverted W Notes Intriguing Errors|
February 27, 2013
This article was originally printed in Bank Note Reporter.
>> Subscribe today!
One of the most fun and intriguing serial numbering errors to appear were $1 Series of 1957 Silver Certificates with an inverted W in place of an M in the upper right serial number.
They occurred in the MA serial number block, and, as of this writing, all the reported errors have serial numbers lower than M51840000A.
This fascinating error was created when a W was inserted upside down into a prefix letter wheel in place of an M on one of the numbering registers used to overprint the MA block.
A Little History
Co-author Huntoon first learned about the error from an upstate New York collector named George Killian in 1968. Killian was an engineer with an eye for detail and may have been the first to discover the error. He had a couple of them so sold M04098541A to Huntoon in May 1968.
Shortly thereafter, Huntoon visited the Bureau of Engraving and Printing with Chuck O’Donnell and met Frank Tucci and Morton C. Rice of the BEP staff. He showed them the error and they explained how an inverted W could be mounted in one of the lettering wheels.
Huntoon didn’t understand enough about the error to ask why the inverted W showed up only in the right-hand serials from positions E2 and E4. This unasked question represented a great lost opportunity, because those 32-subject overprinting presses were current then and those men probably could have explained exactly how the error could move between the two positions.
Now we no longer have access to information about the configurations of those presses or access to the men who operated them, so all we can do is wave our arms about what may have been. Such speculation is thoroughly unsatisfactory.
Huntoon has written a number of articles about the error since his first appeared in Paper Money in 1970, but none fully explain it, and some contain faulty speculations that have been discredited. We’ll be careful here to separate what is certain from what is arm waving.
The Series of 1957 Silver Certificates were printed in 32-subject sheets. The sheets were numbered 20,000 at a time on 32-subject overprinting presses. Consequently, it took 156 full runs of 20,000 sheets, and a 157th partial run, to produce the 100,000,000 notes in a given block. The BEP had more than one overprinting press, so different press runs could be and were assigned to different machines.
The 32-subject sheets were laid out in four numbered quadrants, with 1 being the upper left, 2 lower left, 3 upper right, and 4 lower right.
There were eight notes in a quadrant, consisting of two columns of four notes that were respectively lettered vertically from A to D and E to H.
Serial numbering was sequential through the pile of sheets, so serials 1 to 20,000 in the first press run were numbered on notes from the A1 position, 20,001 to 40,000 on B1, and so forth to 620,001 to 640,000 on H4.
The inverted W occurs only in the E2 or E4 plate positions. These are the upper right notes in the two lower quadrants on the sheet.
The inverted W moved between the E2 and E4 plate positions, but stayed in the same position throughout a particular press run. Consequently, all 20,000 notes in the affected position exhibited the error, whereas all the notes from the other possible position are normal.
We have recorded pairs of notes from press runs 41, 55, 63, 69 and 70 where notes with an inverted W come from one of the two possible positions and normal notes from the other. This demonstrates conclusively that the inverted W occurred only on one of the lettering wheels on the press. (See Fig. 2 and Tables 1 & 2.) We have yet to find consecutive press runs where the inverted W occurs in the same position.
We have recorded normal notes from both position E2 and E4 from press runs 26, 39, 48, 50, 112 and 122, so we have definite evidence that the inverted W was not used in selected press runs. (See Fig. 3 and Table 2.)
M51773011A currently is the highest reported serial number bearing an inverted W. It was overprinted in the 81st run of 20,000 sheets.
How the Error Was Made
Each serial numbering register consists of 10 wheels mounted on a common shaft. All 10 wheels are the same diameter. The numbers 0 through 9 occur around the circumference of the eight numbering wheels.
The first and last wheels carry letters. The letter wheels in use when this error was created were made with the first nine letters of the alphabet and a blank position with a receptor for a removable letter or star that was held in place with a set screw.
M is the 13th letter in the alphabet, so it was not on the wheels. Consequently the press operator had to insert 64 Ms into the prefix wheels when he set up his 32-subject numbering press. One W got mixed in with the Ms and was inserted upside down into a wheel. It is obvious that such letters could be inserted into the receptor correctly or rotated 180 degrees.
Once the W was inserted, it remained in that serial number register at least until the notes from press run 81 were numbered.
So why does the error appear in either the E2 or E4 position if there was only one wheel with an inverted W? Here is where we must speculate about the character of the press.
Our working hypothesis is that the serial numbering registers were mounted on two interchangeable frames on the press, one for quadrants 1 and 2, and the other for quadrants 3 and 4.
The operator removed the two frames and dialed in the correct new starting serial numbers prior to the beginning of a press run. He then remounted the frames on the press. It was luck of the draw which frame was used for a given half, so the wheel with the inverted W ended up either over the E2 or E4 position. However, once there, it stayed for the entire 20,000 sheet press run.
Why is the inverted W missing from more than half of the first 81 press runs?
There are two likely explanations. The most plausible is that those press runs were done on a different press. Less likely, but possible, is that the frame containing the register with the inverted W was periodically cycled off its press and replaced with a look-alike. Both of these scenarios could have been going on simultaneously.
We believe that the lack of inverted Ws above run 81 implies that the serial numbering register containing the inverted W ceased to be used for the rest of the MA block. Probably all the numbering was done on another press or presses. It is very unlikely that the inverted W finally was discovered after run 81 and replaced with an M.
How to Calculate the Press Run and Plate Position
To obtain the press run, divide the serial by 640,000, then round the result up to the next integer.
To obtain the plate position, (1) divide the serial by 640,000, (2) keep only the decimal fraction from the result, (3) then multiply the decimal fraction by 32. If the result is between 12 and 13, the position is E2; if between 28 and 29, the position is E4.
Inverted stars have occurred on a few occasions.
There are two $1 Silver Certificate cases: Series of 1935E in the left serial from the *D block and Series of 1935G in the left serial from *G block. An inverted star also occurred on the left side of some $20 1950B Federal Reserve Note Chicago notes and right side of some $5 1950C FRN Philadelphia notes.
The all-time favorites among error collectors involve $100 Series of 1928 Federal Reserve star notes from the Dallas and San Francisco districts. Many of the notes from the first printing of these stars come with either one or two inverted stars, depending on the plate position. Both printings were from press number 14 on Nov. 22, 1929. It is clear that the press operator was unaware that the stars had a preferred orientation, so they were inserted randomly into the suffix wheels.
The same mistake had been made the day before on press No. 2 when they printed a run of 2,000 sheets of Series of 1928 FRN $20 star notes for New York. The right star in at least the H position was upside-down, creating at least 2,000 such inverts.
These errors prove that if it can happen it will, and that the same mistake will be repeated if the opportunity for its occurrence isn’t engineered out of the equation.
Your Reports are Valuable
Please send data on your inverted W and normal notes from the E2 and E4 positions from the Series of 1957 MA block. Every new entry to our census more completely fleshes out our understanding of this great error.
We also are very interested in other instances of inverted characters in serial numbers. Send scans or descriptions of your finds to firstname.lastname@example.org.
Co-author Derek Moffitt and another serious collector who wishes to remain anonymous have diligently collected and recorded the all-important normal notes from the E2 and E4 positions of the $1 SC 1957 MA block in order to better understand the inverted W errors. Their data coupled with supplemental data provided by Billy Baeder, Fred Bart, Gary Brown, Laura Kessler, Steve Sullivan and others made the census presented here possible. Derek Moffitt proposed the idea that the presses probably contained two interchangeable frames that held the serial numbering registers. The personnel of the Bureau of Engraving and Printing Historical Resource Center confirmed that the $1 Series of 1957 notes were numbered on 32-subject bi-color presses. A cut sheet containing the $1 A00000005A Series of 1957 Silver Certificate is part of the Treasury collection now housed in the Division of Numismatics, Smithsonian Institution, and demonstrates that serial numbering advanced by 20,000 between the plate positions. The large 32-subject overprinting presses predated the COPE numbering and separating presses used today, which utilize 16-subject half sheets as feed stock.
More Coin Collecting Resources:
• Get the 2012 Coin of the Year, and begin your collection of the award winning Mint of Israel Biblical Coins Series all at once for one low price!
• Subscribe to our Coin Price Guide, buy Coin Books & Coin Folders and join the NumisMaster VIP Program
• The Essential Guide to Investing in Precious Metals
Add to: del.icio.us digg
With this article: Email to friend Print
Something to add? Notice an error? Comment on this article.