Very recently, I looked at the data in the SNDB for polymer "Frontier" series $5 and $10 notes. Both of these denominations are being printed exclusively by the Canadian Bank Note Company (CBN). I determined that both of these denominations are being printed with the same position number layout as all the other denominations issued by CBN (see the layout at the beginning of
an earlier discussion ). The skip interval, however, is just 200, and the chronology of notes is column-based, meaning that the serial numbers increase by 200 down a column instead of across a row. So it takes just 200 polymer sheets to create a range of sequentially-numbered notes, and at 45 notes per sheet, reams are 9,000 notes.
Here is the chronology of position numbers: 41, 42, 43, 44, 45, 46, 47, 48, 49, 31, 32, 33, 34, 35 ,36, 37, 38 ,39, 21, 22, 23, 24, 25, 26, 27, 28, 29, 11, 12, 13, 14, 15, 16, 17, 18, 19, 01, 02, 03, 04, 05, 06, 07, 08, 09.
With the $5 and $10 denominations being in circulation just 2 months to this point, it would seem like much more data is needed to confirm the skip-200 printing format's systematic use. However, it was simple for me to determine that this format is being used for all notes of these smallest denominations because of patterns in numbering. It takes 5 positions of the layout to print 1,000 notes. After a ream of 200 sheets is produced, the serial numbers jump ahead by 9,000 to start the next ream. Here is a hypothetical example of notes that could be printed at one particular position:
0010000 to 0010199
0019000 to 0019199
0028000 to 0028199
0037000 to 0037199
0045000 to 0045199
Notice that the fifth digit (or third digit from the right) is always a 0 or a 1. For the given position, this NEVER changes. For the next position in the layout, a similar observation is made:
0010200 to 0010399
0019200 to 0019399
0028200 to 0028399
0037200 to 0037399
0045200 to 0045399
In this case, the fifth digit in the serial number is always a 2 or a 3. Every note printed at that position will have the fifth digit as a 2 or a 3.
I was able to sort the data based on the value of the fifth digit of serial numbers and found that position numbers were grouped accordingly. Here is the summary:
- If the fifth digit in the serial # is 0 or 1, the position numbers are 05, 14, 19, 23, 28, 32, 37, 41, 46.
- If the fifth digit in the serial # is 2 or 3, the position numbers are 01, 06, 15, 24, 29, 33, 38, 42, 47.
- If the fifth digit in the serial # is 4 or 5, the position numbers are 02, 07, 11, 16, 25, 34, 39, 43, 48.
- If the fifth digit in the serial # is 6 or 7, the position numbers are 03, 08, 12, 17, 21, 26, 35, 44, 49.
- If the fifth digit in the serial # is 8 or 9, the position numbers are 04, 09, 13, 18, 22, 27, 31, 36, 45.
So, the logical question to ask is... WHY ARE THEY DOING THIS? And are the printers literally making only 200 sheets at a time before they have to stop the numbering machine and reset the numbers?
In a recent discussion, it was mentioned that new technologies are available that would allow computerized control of numbering so that the machines would not have to be stopped and numbers would change as needed on the fly. Of course, it can't be proven or disproven whether these technologies are, in fact, in current use by CBN. Just because a technology exists, that doesn't mean it is in widespread use. Think of it this way... just because the iPhone 5 exists, that does not mean everyone has one.
I believe that CBN can apply serial numbers to more than 200 sheets at a time, but I don't think a computerized system of controlling the numbering reels is necessary. In fact, I think the decision to use a skip interval of 200 was not at all arbitrary. The reels in the numbering machine are very much like the odometer in your car. In a system where a base of 10 is used, there are 10 digits on every reel, from 0 to 9. When a 9 flips over to a 0, the reel to the left also moves ahead by 1. That one reel can affect the reel immediately to the left of it is achieved simply by physical means, and this is something anyone can learn about by researching mechanical odometers on the Internet.
In the hypothetical example I gave earlier, you will notice that the fourth digit in the serial number does not increase with every group of 200 notes printed. It decreases by 1. So, instead of having conventional reels all numbered 0 through 9 in ascending order, it's possible to put in a reel numbered in reverse order, from 9 though 0, and it behaves exactly the way a normal reel would.
But what about the reel controlling the fifth digit? It has only two values. To function purely by physical controls, it would need to be built in such a way that it has either only two positions that change over rapidly, or it has the two values repeated five times and a different set of gear teeth that make the reel to the left roll over five times as frequently. Here is a schematic diagram of what the numbering reels could look like:
| 0 | 0 | 0 | 0 | x | 0 | 0 |
| 1 | 1 | 1 | 9 | y | 1 | 1 |
| 2 | 2 | 2 | 8 | x | 2 | 2 |
| 3 | 3 | 3 | 7 | y | 3 | 3 |
| 4 | 4 | 4 | 6 | x | 4 | 4 |
| 5 | 5 | 5 | 5 | y | 5 | 5 |
| 6 | 6 | 6 | 4 | x | 6 | 6 |
| 7 | 7 | 7 | 3 | y | 7 | 7 |
| 8 | 8 | 8 | 2 | x | 8 | 8 |
| 9 | 9 | 9 | 1 | y | 9 | 9 |
In the above diagram, x= 0, 2, 4, 6, or 8, and y = x + 1.
I'm not a mechanic, so I don't know if "special reels" would be easy to install. I do believe, however, that computerizing a numbering machine with 90 sets of mechanical reels would be very difficult. It would be easier for someone to design a new machine than to adapt technology to an older mechanical machine. Even if you could adapt an old machine and design software that controls it and train CBN workers to use the software, you still have to tell the machine exactly what to do, and with 45 positions in the layout and a numbering scheme that jumps every 200 sheets... it could be an absolute nightmare. And I actually have experience using old scientific equipment where you could analyze 20 separate samples from 20 separate manifolds, but the software interface required you to enter the analysis type, the sample identity, the output file format, etc, for every single sample without any shortcuts.
Anyhow, it should be remembered that putting serial numbers on bank notes is an old-fashioned physical process. By tweaking a numbering machine so that it needs a minimal amount of human input before and during runs makes a heck of a lot more sense than trying to use technology when there is no guarantee that the new technology will help to make the job faster and/or easier.
Regardless of the technology in use, what is very clear is that the concept of "reams" has changed. We used to think the ream size was matched to the skip-interval of the numbering pattern. With a skip-200 numbering system and machines that don't need to be reset, reams essentially don't exist. This fact appears to be supported by the prevalence of mixed bricks and mixed bundles. There are clearly no attempts by CBN printers to maintain order after serial numbers are printed on sheets. Moreover, insert notes seem to be lacking. From what is being reported, many bundles of notes will have two or more serial number ranges, and even within the different ranges, not all the notes are consecutive. Runs of notes will have gaps in the numbers without inserts in the place of the missing notes or added to the ends of the bundle. With two or more position numbers inside bundles, it is clear that bundles are being made out of loose and seemingly random notes. The days when we thought they would cut up 100 sheets and then bundle all the stacks of 100 notes appear to be over.
There is no explanation I can think of why using insert notes in the current printing process makes any sense. I think skip-200 printing, whether it is here to stay or not, is the beginning of the end of insert notes. Even if the Bank of Canada still opens bricks and bundles and puts in SNRs here and there, they will be very scarce, and scarcity actually discourages brick hunters from trying to find them. And how would you even define the boundaries of insert note ranges given the chaotic nature of the distribution system? A brick of notes set aside by the Bank of Canada for use as SNRs is bound to be just as scrambled as the bricks into which they will be putting SNRs.
I think if skip-200 printing is applied to all denominations in the future, insert note collecting will die. There might be some interest in older issue inserts and replacements, creating a sort of "renaissance market", but I think many collectors will just drop out and take the market down with them.
