Collector’s Gallery

Research Center

 

 

“The Bank of Canada stopped identifying replacement notes with a special
 identification in 1996 and has adopted a procedure commonly used among other
 central banks. Notes are now taken from regular production and used as
 replacements to replace notes found during the examination process that did
 not meet the quality standards.  This procedure is less time consuming and
 more practical for both the Bank and its printers.  They enter circulation
 through the normal distribution process.”

Bank of Canada Public Relations Official, November 16th, 2006.

 

 

Journey Replacements: The Black Box!

Part II

CBN $20s

© Hudson Byblow, February 5th, 2007.  This information is free for use for collector research and educational purposes.

 

 

Before you start reading:  Make sure to allot your self a good section of time, grab some water, and take a deep breath. This may seem boring and technical, but if you are into replacements, then you must continue. This will involve modular math, logic built on logic, logic ruling out impossibilities, determining contradictory suppositions, and other dendrite forming exercises. Feed your brain, read this.  You will see exactly how relevant this is, and how complex it can be. I am sorry, but you have reached the point of no return. J  Without further ado, I present my research, current up to this point in time. BEGIN!

 

 

 

Solving the CBN Journey $20 mystery:

Working backwards from the ranges reported from Gilles Pomerleau, I tried to construct the position numbers of the entire CBN Journey $20 sheet.  In my research with notes, I often discovered finding notes from many different positions, often contained in a small number of bricks.  They would all be of close proximity, so I had the opportunity to compare serial numbers to see how they would line up.  From the variety of replacement notes that I got, all belonging to the same sheet, I was able to determine that the skip numbering pattern was in fact 6000.  That is, there would be 6000 notes printed in position A1 before notes 6000-11999 would be printed in A2.

 

Below is a table that is based on the conventional replacement range, starting at 9, 720,000.  The highlighted cells represent positions that were identified from my EZJ replacement findings listed below.  From the EZJ replacements, we know the serial numbers are skip-6000, thus making the full range of the ream encompass 270,000 notes.

 

 

1

2

3

4

5

A

9720000

9726000

9732000

9738000

9744000

B

9750000

9756000

9762000

9768000

9774000

C

9780000

9786000

9792000

9798000

9804000

D

9810000

9816000

9822000

9828000

9834000

E

9840000

9846000

9852000

9858000

9864000

F

9870000

9876000

9882000

9888000

9894000

G

9900000

9906000

9912000

9918000

9924000

H

9930000

9936000

9942000

9948000

9954000

I

9960000

9966000

9972000

9978000

9984000

 

Below is a sample of EZJ replacements that I found in a series of bricks.  Their position is highlighted in yellow the table above.
9954958 48/11 = H5
9948958 25/46 = H4
9942958 43/38 = H3
9936958 54/12 = H2
9906958 26/26 = G2
9876958 18/53 = F2
9870958 41/30 = F1
9834958 30/41 = D5
9828958 53/18 = D4

 

I also had a finding of EZI replacements, and they confirm the skip-6000 serial numbering pattern.  The also map properly onto the table above, containing the EZJ replacement sample.  Note that there are two common locations, G2 and F2 between the two prefixes. That means that the notes are of the same relative proximity on a sheet, thus the two prefixes (in this number range) can be combined into one table.
9960xxx 27/27 = I1

9984xxx 12/54 = I5   
9822xxx 16/50 = D3  
9864xxx 49/35 = E5
9930xxx 32/51 = H1
9876xxx 18/53 = F2   This conforms to the EZJ layout
9852xxx 24/24 = E3

9888xxx 15/37 = F4  
9894xxx 29/21 = F5  
9780xxx 10/14 = C1  
9732xxx 42/28 = A3  
9756xxx 28/42 = B2

9906xxx 26/26 = G2  This conforms to the EZJ layout.

 

At this point we can draw a consolidated table combining both EZI and EZJ replacements into one sheet layout.  Yellow is EZJ, and orange is EZI.  F2 and G2 are green, because they are represented by both prefixes.

 

 

1

2

3

4

5

A

9720000

9726000

9732000

9738000

9744000

B

9750000

9756000

9762000

9768000

9774000

C

9780000

9786000

9792000

9798000

9804000

D

9810000

9816000

9822000

9828000

9834000

E

9840000

9846000

9852000

9858000

9864000

F

9870000

9876000

9882000

9888000

9894000

G

9900000

9906000

9912000

9918000

9924000

H

9930000

9936000

9942000

9948000

9954000

I

9960000

9966000

9972000

9978000

9984000

 

 

What is the replacement range then?

One other important thing to note is this: from one sheet of EZI replacements, there is a number range spanning from A3 to E5, which is 6000* 43 = 258,000 notes.  The ream itself is 270,000 notes, so this makes perfect sense. The ream would be the first note of A1 to the last note of I5.  The published ranges of 270,000 are correct. In fact, it cannot be otherwise, unless not all 6000 sheets were set aside for replacements.   If that is the case, then the ranges get trickier. I will discuss that later. I will return to CBN $20 replacement ranges later.


Problem posed: Can we test this layout to a known error note?
EYL 8634987 12/54 is a known error, with a butterfly fold at the bottom left (facing the front); therefore when face down, in the bottom right of the sheet.  This implies that it should be in position I5.  This validation will help prove that the method used to determine the replacement note positions (and subsequently, the replacement ranges themselves) are correct.

 

Consider position A1.  If we subtract 270,000 from it, we will arrive at another position A1, but one that is part of a prior ream of notes.  Since we are hypothesizing that notes are 6000 skip, then a ream would encompass 270,000 notes.  Consider a ream to be a multi-brick stack that is 6 bricks high, with 45 bricks making up the total family.  Therefore, if we subtract any multiple of 270,000 from A1, then we should end up with another A1.  Similarly, if we subtract 270,000 from position E4, we would end up back at E4 for example.  This should work for any positions.

 

Let 9720000 = A1

If I subtract 270,000 * 4 I get 8,640,000 as a position A1.  If I subtract 6000 from 8,640,000 to bump it back one more position (to I5), then my starting point for I5 is 8,634,000.  The butterfly error EYL 8634987 falls directly into the hypothesized position based on the serial number.  I went ahead with this method, but soon found out that many numbers were not working properly. The logic I had been using had a flaw of some sort.  Perhaps the EZL error matching up was purely coincidental?

 

To figure this problem of notes not matching to my partially built sheet layout, I took another sample of notes.  They were EZD non-replacement notes I had recorded from prior bricks. The first one I tried was EZD 1950886 53/18, which had to be D4, since it was mapped directly onto a prior EZJ replacement with the same position number.  From there when I counted back by 6000 to A1, and then reduced the A1 position to its lowest term by subtracting multiples of 270,000 from it (modulo 270,000), the A1 position did not comply with a range that should be in that position is reams were 270,000.   I was anticipating on having an A1 value of 1890000, but instead it was actually 1842000 when I skip counted back by 6000 from position D4 to A1 for confirmation. It confirmed that I was wrong. Thus, there was a mystery to solve in regards to skip numbering once again.  Perhaps there was a change, or some notes missing.  What I found out was very interesting.

 

After long and hard research, I made a discovery that these EZD notes I was testing were actually skip numbered by 8000.  I was able to determine this from the large number of EZD notes that I had recorded from a few very mixed bricks.  The samplings that are bold and highlighted yellow are where I found notes with the same FP/BP combination that were in a range of over 6000 notes (but under 8000).  The proximity of other FP/BP combinations define this ream to be skip-8000.  This sample was necessary to show that CBN $20s are not all skip-6000, but some are skip 8000.  Through further research, I found out almost all are skip-8000.  

 

 

Here is a table of the notes that I recorded and the ranges found within the same position: Note: 1801xxx will be written as 1801. 

 

Recorded serial numbers

FP/BP

Implied Block Range

(Six bricks of one position)

1801-1806

52/19

1800000-1807999

1810-1815

36/31

1808000-1815999

1819-1823

42/28

1816000-1823999

1829-1831

33/47

1824000-1831999

1839-1839

45/45

1832000-1839999

1846-1847

47/33

1840000-1847999

1848-1855

28/42

1848000-1855999

1856-1862

31/36

1856000-1863999

1864-1871

19/52

1864000-1871999

1877-1878

50/16

1872000-1879999

1880-1887

10/14

1880000-1887999

1892-1893

13/22

1888000-1895999

1903-1903

17/23

1896000-1903999

1908-1908

40/40

1904000-1911999

1919-1919

23/17

1912000-1919999

1920-1927

22/13

1920000-1927999

1935-1935

14/10

1928000-1935999

1936-1943

16/50

1936000-1943999

1944-1950

53/18

1944000-1951999

1957-1959

30/41

1952000-1959999

1964-1964

35/49

1960000-1967999

1970-1974

34/44

1968000-1975999

1976-1976

24/24

1976000-1983999

1984-1991

44/34

1984000-1991999

1992-1999

49/35

1992000-1999999

None

??/??

2000000-2079999

2008-2014

18/53

2008000-2015999

2021-2021

51/32

2016000-2023999

2024-2029

15/37

2024000-2031999

2037-2038

29/21

2032000-2039999

2043-2047

20/39

2040000-2047999

2048-2054

26/26

2048000-2055999

2059-2063

39/20

2056000-2063999

2070-2070

21/29

2064000-2071999

None

??/??

2072000-2079999

2080-2086

32/51

2080000-2087999

2095-2095

54/12

2088000-2095999

2101-2101

43/38

2096000-2103999

2107-2109

25/46

2104000-2111999

2114-2119

48/11

2112000-2119999

None

??/??

2120000-2127999

None

??/??

2128000-2135999

2136-2139

46/25

2136000-2143999

None

??/??

2144000-2151999

2152-2159

12/54

2152000-2159999



After many hours of research, and by carefully placing the information in the correct pigeon hole according to the data collected, I was able to develop a near complete table of FP/BP combinations, as seen below:

 

 

1

2

3

4

5

A

52/19

36/31

42/28

33/47

45/45

B

47/33

28/42

31/36

19/52

50/16

C

10/14

13/22

17/23

40/40

23/17

D

22/13

14/10

16/50

53/18

30/41

E

35/49

34/44

24/24

44/34

49/35

F

41/30

18/53

51/32

15/37

29/21

G

20/39

26/26

39/20

21/29

??/??

H

32/51

54/12

43/38

25/46

48/11

I

27/27

??/??

46/25

??/??

12/54

 

I was also able to map EZJ 9870958 41/30 to F1, and EZI 9960xxx 27/27 to I1, because of the correctness of the note mapping of the others.  If the others would not have worked, this would not have worked.  So, at this point, F1 and I1 are also mapped.  This is shown above in orange.  At this point, we can see that there are only 3 positions remaining.

 

Conclusions thus far:

CBN $20s are 45/on

Notes from 9.72M to 9.99M appear to be skip-6000.  Therefore, in this sections a ream is 270,000 notes.

Notes below that serial number range are skip-8000.  Therefore, in this section a ream is 360,000 notes.

Notes above 9.99M have an unknown skip pattern, if they exist at all. Some have been found.

We can ascertain that CBN $20 replacements are full sheet replacements.  All evidence that I am aware of at this time plus my findings point to this likelihood.

 

How on earth did the butterfly error correspond to the current position when we were counting down by 6000?

It was the fourth ream down, and 4 * 6000 is 24,000.   24,000 is attainable by adding groups of 8000 also.  The reason the crossover is because the Least Common Multiple of 6000 and 8000 is 24,000.  It was just a coincidence.

 

 

Continuing to finish the matrix:

I had four bricks of EZM that were in the high 9M range.  They were skipped by 8000, unlike the replacement ranges that were skipped by 6000. From getting four bricks of them, I was pretty sure that they were not replacements. I found it very interesting that they were skip-8000.  Perhaps that is some sort of implicit way to tell with CBN $20s in the 9M range are replacements?  That is something we may not know for sure. But from it, I managed to map out the remaining positions.

EZM 9687xxx 27/27:  This note checks out perfectly to the EZI and EZJ position of this particular FP/BP combination.

EZM 9695xxx 11/48:  When in skip-8000, this falls perfectly into I2.

EZM 9639xxx 37/15:  When in skip-8000, this falls perfectly into G5.

 

This leaves only one position left: I4.

There is only one FP/BP combination that has not yet been employed. By process of elimination, the missing FP/BP is 38/43, in I4. From this, we can put together the Standard Matrix for CBN $20s.  It is seen below.

 

 

The CBN $20 Standard Matrix

 

 

1

2

3

4

5

A

52/19

36/31

42/28

33/47

45/45

B

47/33

28/42

31/36

19/52

50/16

C

10/14

13/22

17/23

40/40

23/17

D

22/13

14/10

16/50

53/18

30/41

E

35/49

34/44

24/24

44/34

49/35

F

41/30

18/53

51/32

15/37

29/21

G

20/39

26/26

39/20

21/29

37/15

H

32/51

54/12

43/38

25/46

48/11

I

27/27

11/48

46/25

38/43

12/54

 

 

Conclusions about printing:

With certainty, we know that replacements that are skip-6000 are from a ream of 270,000.

With certainty, ANY other range of notes that are skip-8000 must be from a ream of 360,000.

The ream size is dependant on the skip-numbering, since the 45/on format is used continuously.

From this matrix, we can easily find out the exact position of a single banknote within a ream.

 

 

 

 

How this applies to Replacement Notes:

Now, with this matrix, let us analyze the sample:

Information reported to the Canadian Paper Money Forum on January 10th, 2007:

 

Found more EZR replacements (very few) in EZS brick. Don't know if this goes in EZR or EZS thread.
EZS $20s bricks x 4  Jan 8th, 2007
 
EZS 7913000
EZR 9984364
replacement
EZS 7913200
All 12/54
 
EZS 7841000
EZR 9930364
replacement
EZS 7841200
All 32/51
 
EZS 7769000
EZR 9876364
replacement
EZS 7769200
All 18/53
 
EZS 7665000
EZR 9798364
replacement
EZS 7665200
All 40/40
 
EZS 7632601
EZR 9774325-324
replacements
EZS 7632600
All 50/16
 
EZS 7664601
EZR 9798325-324
replacements
EZS 7664600
All 40/40  -- End source quote.

 

From the replacement notes found in this group of 4 EZS bricks, we have only 4 unique position numbers.  The table below will show where the replacement notes came from, and what the full ream would be.  Neighboring “Super-Brick” packages (groups of four bricks) likely would have had more EZR replacements.

 

 

1

2

3

4

5

A

9720000-9725999

52/19

9726000-9731999

36/31

9732000-9737999

42/28

9738000-9743999

33/47

9744000-9749999

45/45

 

 

 

B

9750000-9755999

47/33

9756000-9761999

28/42

9762000-9767999

31/36

97680000-9773999

19/52

9774000-9779999

EZR 9774324

EZR 9774325

50/16

 

C

9780000-9785999

10/14

9786000-9791999

13/22

9792000-9797999

17/23

9798000-9803999

EZR 9798324

EZR 9798325

EZR 9798364

40/40

9804000-9809999

23/17

D

9810000-9815999

22/13

9816000-9821999

14/10

9822000-9827999

16/50

9828000-9833999

53/18

9834000-9839999

30/41

 

 

 

E

9840000-9845999

35/49

9846000-9851999

34/44

9852000-9857999

24/24

9858000-9863999

44/34

9864000-9869999

49/35

 

 

 

F

9870000-9875999

41/30

9876000-9881999

EZR 9876364

18/53

9882000-9887999

51/32

9888000-9893999

15/37

9894000-9899999

29/21

 

 

 

G

9900000-9905999

20/39

9906000-9911999

9926/26

9912000-9917999

39/20

9918000-9923999

21/29

9924000-9929999

37/15

 

 

 

H

9930000-9935999

EZR 9930364

32/51

9936000-9941999

54/12

9942000-9947999

43/38

9948000-9953999

25/46

9954000-9959999

48/11

 

 

 

I

9960000-9965999

27/27

9966000-9971999

11/48

9972000-9977999

46/25

9978000-9983999

38/43

9984000-9989999

12/54

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

From this table, we know with certainty that the replacement range MUST include the notes fall into the other positions of the sheet as well.  That is, there is an replacement note that ends with 364, 325, and 324 in every position.  The lowest notes from this combination of three replacement sheets found would be 9720324, 324, 364, all from position A1, with the FP/BP of 52/19.  The highest notes from this trio of sheets would be 9984324, 325, 364, all from position I5.

 

From a single replacement sheet, we can see that the range spans between 9720324 to 9984324.  It is impossible for a sheet replacement to contain a spread of any less, provided that the whole ream is set aside for replacement notes.  If the whole ream is set aside for replacements, then every serial number from 9.72M to 9.9M (less one) is an replacement note.

 

Question:

Is it possible to have a replacement range that is only 11250 consecutive notes?  (Note: CBN $20s are full sheet replacements)

No.  The only way for this to be possible is if:

a)     The skip numbering was such that there were a total of 11250 notes in the entire ream (this would mean notes are skip-250).

b)     The notes are printed skip-11250, and one position of 11250 notes in height (out of the 45 positions) was removed and set aside for replacements.

 

The first possibility fails because we know notes are skip-6000 in the 9.72-9.99M replacement range, and not skip-250 which would be needed in order for the first possibility to be valid.

The second possibility is disproved by the fact that replacements are found in full sheets.  They are not a consecutive run of notes all coming from one position.  We know this because of the varying FP/BP combinations which do in fact match up to the notes that are neighboring when they are removed from the brick.  Also, the fact that same sheet cousin notes are often found (having the same last three digits).

 

 

Is it possible to have 11250 total notes as replacements? YES.

Suppose that the bottom 250 sheets have been removed (after being serial numbered) for replacement purposes.  That means, notes that end in 0000-0249 would be the replacement range for position A1. That is a range of 250. In A1, notes 0250-5999 would be regular issue.  A2 would follow the same pattern, since it is a sheet we are removing.  In A2, 6000-6249 are replacements, and then 6250-11999 are regular issue. 

If the case is that a full ream of notes is NOT used as replacement notes, the replacement ranges must be as follows. There is no other possibility other than this:

(Using 9.72-9989999 as the basis for the example)

 

 

Position on Sheet

Serial Number Ranges

Replacement or Non-replacement (Regular issue)

A1

9720000-9720249

Replacement

A1

9720250-9725999

Non-Replacement (Regular issue)

A2

9726000-9726249

Replacement

A2

9726250-9731999

Non-Replacement (Regular issue)

A3

9732000-9732249

Replacement

A3

9732250-9737999

Non-Replacement (Regular issue)

A4

9738000-9738249

Replacement

A4

9738250-9743999

Non-Replacement (Regular issue)

A5

9744000-9744249

Replacement

A5

9744250-9749999

Non-Replacement (Regular issue)

B1

9750000-9750249

Replacement

B1

9750250-9755999

Non-Replacement (Regular issue)

B2

9756000-9756249

Replacement

B2

9756250-9761999

Non-Replacement (Regular issue)

B3

9762000-9762249

Replacement

B3

9762250-9767999

Non-Replacement (Regular issue)

B4

9768000-9768249

Replacement

B4

9768250-9773999

Non-Replacement (Regular issue)

B5

9774000-9774249

Replacement

B5

9774250-9779999

Non-Replacement (Regular issue)

C1

9780000-9780249

Replacement

C1

9780250-9785999

Non-Replacement (Regular issue)

C2

9786000-9786249

Replacement

C2

9786250-9791999

Non-Replacement (Regular issue)

C3

9792000-9792249

Replacement

C3

9792250-9797999

Non-Replacement (Regular issue)

C4

9798000-9798249

Replacement

C4

9798250-9803999

Non-Replacement (Regular issue)

C5

9804000-9804249

Replacement

C5

9804250-9809999

Non-Replacement (Regular issue)

D1

9810000-9810249

Replacement

D1

9810250-9815999

Non-Replacement (Regular issue)

D2

9816000-9816249

Replacement

D2

9816250-9821999

Non-Replacement (Regular issue)

D3

9822000-9822249

Replacement

D3

9822250-9827999

Non-Replacement (Regular issue)

D4

9828000-9828249

Replacement

D4

9828250-9833999

Non-Replacement (Regular issue)

D5

9834000-9834249

Replacement

D5

9834250-9839999

Non-Replacement (Regular issue)

E1

9840000-9840249

Replacement

E1

9840250-9845999

Non-Replacement (Regular issue)

E2

9846000-9846249

Replacement

E2

9846250-9851999

Non-Replacement (Regular issue)

E3

9852000-9852249

Replacement

E3

9852250-9857999

Non-Replacement (Regular issue)

E4

9858000-9858249

Replacement

E4

9858250-9863999

Non-Replacement (Regular issue)

E5

9864000-9864249

Replacement

E5

9864250-9869999

Non-Replacement (Regular issue)

F1

9870000-9870249

Replacement

F1

9870250-9875999

Non-Replacement (Regular issue)

F2

9876000-9876249

Replacement

F2

9876250-9881999

Non-Replacement (Regular issue)

F3

9882000-9882249

Replacement

F3

9882250-9887999

Non-Replacement (Regular issue)

F4

9888000-9888249

Replacement

F4

9888250-9893999

Non-Replacement (Regular issue)

F5

9894000-9894249

Replacement

F5

9894250-9899999

Non-Replacement (Regular issue)

G1

9900000-9900249

Replacement

G1

9900250-9905999

Non-Replacement (Regular issue)

G2

9906000-9906249

Replacement

G2

9906250-9911999

Non-Replacement (Regular issue)

G3

9912000-9912249

Replacement

G3

9912250-9917999

Non-Replacement (Regular issue)

G4

9918000-9918249

Replacement

G4

9918250-9923999

Non-Replacement (Regular issue)

G5

9924000-9924249

Replacement

G5

9924250-9929999

Non-Replacement (Regular issue)

H1

9930000-9930249

Replacement

H1

9930250-9935999

Non-Replacement (Regular issue)

H2

9936000-9936249

Replacement

H2

9936250-9941999

Non-Replacement (Regular issue)

H3

9942000-9942249

Replacement

H3

9942250-9947999

Non-Replacement (Regular issue)

H4

9948000-9948249

Replacement

H4

9948250-9953999

Non-Replacement (Regular issue)

H5

9954000-9954249

Replacement

H5

9954250-9959999

Non-Replacement (Regular issue)

I1

9960000-9960249

Replacement

I1

9960250-9965999

Non-Replacement (Regular issue)

I2

9966000-9966249

Replacement

I2

9966250-9971999

Non-Replacement (Regular issue)

I3

9972000-9972249

Replacement

I3

9972250-9977999

Non-Replacement (Regular issue)

I4

9978000-9978249

Replacement

I4

9978250-9983999

Non-Replacement (Regular issue)

I5

9984000-9984249

Replacement

I5

9984250-9989999

Non-Replacement (Regular issue)

 

 

From the evidence that I have gathered from other CBN replacement prefixes, it is unlikely that the above case is what is happening. Thank goodness, otherwise that would make replacement notes possibly too much to handle.  I have found EZD that, per position, are near the end of the 6000, in the middle of the 6000 and even near the start of the 6000.  I have witnessed this in other prefixes as well.   If the above case is a possibility, I would expect the CBN to remove a group of 1000 sheets, or some other round number, thus making the interlaced replacement ranges all 1000, totally 45,000.  This is just my opinion of how they might do it, if this was the case.

 

 

 

 

Conclusion:

It is most likely and logical that the replacements are set aside per full ream.  The evidence with EZD and other prefixes prove this. It makes no sense for the CBN to set aside a portion of a ream for replacements, because then they would have to compensate for missing notes from that ream by robbing notes from another ream, thus causing a potentially never ending cycle.  By finding replacement notes that fluctuate throughout the entire section of 6000 notes per position, we know that the replacements are not restricted to only a small section of the 6000 sheets (such as the bottom 250).  Because of this fact, this table with the many small replacement ranges interlaced within non-replacement notes, would in all likelihood be a non-possibility.

The only remaining logical possibility is to have replacements set aside in full reams.  That is, if this is the case, then replacement ranges MUST be either 270,000 or 360,000, depending on whether they are skip-6000 or skip-8000 respectively.

 

 

 

 

What about EZL replacements found in the 8 million range?

http://www.cdnpapermoney.com/cgi-bin/yabb/YaBB.pl?num=1166492031

On or about December 22nd, 2006, I was lucky enough to search 12 bricks of EZN $20s; 3 full packages of four bricks each. Of all 12 bricks, I found only three with replacements. They were as follows:

EZN 2000201-200 
EZL 8652977 Replacement
EZL8660601-600 TWO Replacements
EZN1968219 and into the EZN again
Jumping around in the SAME bundle,
EZN2159871
EZL 8796617-16 TWO Replacements
EZN 2031999-8
EZL 8868986 Replacement
EZL 8876991
Replacement

EZN 2064xxx and on....
Too much of a hurry to record the FP/BP of this bundle, because there was a lot of jumping around, and not very much time.
This was ONE bundle by the way.
 
Second bundle of significance:
EZN 1084xxx
EZN 1084660 replaced by EZL 8644957
EZN 1084667 replaced by EZL 8644954
ALL 52/19
 
Third Bundle of significance
EZN 1100xxx
EZN 1100660 replaced by EZL 8660957
EZN 1100667 replaced by EZL 8660954
ALL 42/28   - End source quote.

 

The first thing to figure out with these EZL replacements is the size of the ream.  By applying the Standard Matrix to the replacements, I know that EZL 8644957 must be two positions away from EZL 8660957.  There is also a numerical difference between the serial numbers of 16,000.  This means that these EZL replacements are skip-8000.  Because they are skip-8000, which means that the ream is 360,000 notes large.  Other notes confirmed this statement.  The find indicated that these were full sheet replacements, which is normal for CBN $20s.  In other words, there are 43 other replacement notes that all end in 957, each one representing a different position of the sheet.

 

The first bundle had many jumps within it.  The replacement notes found were not specifically replacing notes removed in the exact spot from what I could tell.  It is likely that they were inserted in those “common” positions after other notes were removed, to compensate for the missing notes – much like the older style CBN $5s.  However, in the next two bundles that had information, it was clear that the EZL replacements were replacing a certain identifiable note.  This was important because we could determine that these EZL replacements had common FP/BP combinations with their surrounding non-replacement EZN notes.   From that, we can accurately assume that they were full sheet replacements, just like the other CBN $20 replacements.

 

Getting back to the first bundle, we can work backwards from the Standard Matrix to figure out the position numbers of the replacements and the notes surrounding them.  Keep in mind that often times within a CBN bundle of $20s, there are different runs of notes, and sometimes even singles.  The construction of the brick seems to be quite random at times.  Let us examine the notes from the first bundle:

Note:

The EZN are all from the 6th ream of the prefix, with 1800000 being the first note of A1. 

The EZL are from the 24th ream of the prefix, 8640000 being the first note of A1.

 

EZN 2000201 is from the 26th position of its sheet.  From the Standard Matrix, we know then that EZN 2000201 is 41/30.

EZL 8652977 is from the 2nd position of its sheet.  From the Standard Matrix, we know then that EZL 8652977 is 36/31.

  • The fact that the EZL FP/BP differs from the notes on either side is irrelevant in this case, because we can see that it is found amongst a factory shuffled position point.  Again, many times the CBN has sent bricks out that have been pieced together with one or more single notes.  How they select notes to be put into bundles is not like any other Journey note denomination I have witnessed before, with patterns resembling a shuffled deck of cards at times.

EZL 8660601-600 are from the 3rd position of their sheets.  From the Standard Matrix, we know then that EZL 86660601-600 are 42/28.

EZN 1968219 is from the 22nd position of its sheet.  From the Standard Matrix, we know then that EZN 1968219 is 34/44.

EZN 2159871 is from the 45th position of its sheet.  From the Standard Matrix, we know then that EZN 2159871 is 12/54.

EZL 8796617-616 are from the 20th position of their sheets.  From the Standard Matrix, we know then that EZL 8796617-616 are 30/41.

EZN 2031999-998 are from the 29th position of their sheets.  From the Standard Matrix, we know then that EZN 2031999-998 are 15/37.

EZL 8868986 is from the 29th position of its sheet.  From the Standard Matrix, we know then that EZL 8868986 also is 15/37.

EZL 8876991 is from the 30th position of its sheet.  From the Standard Matrix, we know then that EZL 8876991 is 29/21.

EZN 2064xxx and downward (to 2064000) are from the 34th position of their sheets.  From the Standard Matrix, we know then that these notes are 21/29.

 

 

Now, combining these precise numbers with the other EZL replacements found in the other two bundles with the FP/BP combinations of 52/19 and 42/28, and we can draw out a new matrix, highlighting the positions that EZL replacements were found in.  I will spare the detail, and provide a much simpler table than the EZR one designed earlier.  The position numbers that are bolded are the positions in which EZL replacements were found:

 

 

1

2

3

4

5

A

52/19

36/31

42/28

33/47

45/45

B

47/33

28/42

31/36

19/52

50/16

C

10/14

13/22

17/23

40/40

23/17

D

22/13

14/10

16/50

53/18

30/41

E

35/49

34/44

24/24

44/34

49/35

F

41/30

18/53

51/32

15/37

29/21

G

20/39

26/26

39/20

21/29

37/15

H

32/51

54/12

43/38

25/46

48/11

I

27/27

11/48

46/25

38/43

12/54

 

 

For the same reasons described earlier, any range less than the full ream is impossible, excluding the possibility of the 45 small interlaced replacement ranges. If these EZL were single note replacements, which CBN $20s have shown not to be over and over, then they would have likely come from a source with serial numbers in much closer proximity.  The second and third relevant bundles in the EZL replacement finding point to the use of full sheet replacements.  Had I found more bundles of EZN that crossed the same serial number range (relative position out of 6000), then they too would have had echoing finds.  The problem is that it is difficult for many people to access $20 bricks to discover this information.  Thus many of the echoing replacement finds (same finds but from other sheet positions, or other finds within the ream) are not found.  That should not mean that the replacement range is pared down though to a smaller range, it just means that less representations of the sheet were found.

 

If the EZL replacements were sheet replacements of an amount that was less than the full ream, then we would have to develop another large table, interlacing the replacement ranges within the larger non-replacement ranges, for each position.  Since this is a possibility that has not been logically ruled out completely, I will consider it with these EZL skip-8000 replacements.  These replacements were confined to a range of the 4601st note to the 5000th note (ending in 4600-4999) of the 8000 per position.  This needs to be compared with other finds. If ALL other finds fall within a tiny range like this, then yes by all means a new 90 line table must be drawn to accurately show the serial numbers of each position that are not replacements.  Based on the other prefixes I would expect the replacements to be full ream, and thus be a range going from EZL 8640000 – 89999999, or 360,000 notes.  There is no possible way, however, for these replacements to have a range exceeding 8000 consecutive notes, unless the range is in fact the full ream.  Anything below 8000 and we are talking about a large table with 45 sub-ranges.  If the range(s) were 8000 notes exactly, then that would likely imply single note replacements from one or more specific bricks.  Since CBN uses full sheet replacements we can rule out the prior statement.

 

 

The cautionary perspective:

1. EZL 9.72-9.99M replacements have been found, and so have these 8.64-9M EZL replacements.  Can we assume that all notes in between are replacements? NO!  There are two full reams in between these two found ranges.  We cannot make that assumption.

 

2. EZL 8.644 and 8.868 replacements have been found.  Can we assume that all notes in between are replacements?  According to my research regarding FP/BP combinations, skip numbering patterns, and sheet size, then we can answer this question with a “yes”, with a very high degree of certainty.  Yes to the rest of the ream being replacements as well.  The FP/BP combination placement and skip numbering pattern show us that it is a pure contradiction to assume that we can determine the replacement range by taking x amount of consecutive notes on either side of a discovered replacement note when dealing with full sheet replacements.  That is, just a few replacement notes can DEFINE the whole range if they come from the same ream (excluding the possibility of the 45 small replacement ranges interlaced within non-replacements).  I use bold and capitalization as emphasis for the word “Define” because it is separate from “confirmation”.  A range may be defined through the use of the Standard Matrix, but it takes multiple findings of notes from that ream in order for it to be officially confirmed as per protocol. The important thing to remember here is now we can pinpoint the position of a note within a prefix, a ream, and a brick, thus allowing us to define replacement ranges with much more ease and accuracy.

 

 

Final Conclusion:

The logic and data point to full ream replacements, just as they are with the BABN $5s.  That is, by using the Standard Matrix, we can pinpoint the exact position of any note, replacement or not, and determine from that the exact replacement range with certainty. 

  • If skip-6000, then the ream is 270,000 notes
  • If skip-8000, then the ream is 360,000 notes
  • Replacement notes (for CBN $20s) are inserted in full sheets (45/on)
  • Evidence points to replacement notes being full reams
  • Ranges of 11250, 22500, or even 45000 are physically impossible based on how CBN prints the notes.

 

 

 

 

As you can see from the pure necessity of details, it is imperative that the information gets sent to the proper collaborators.  At this point I would like to thank all those that came before me, for their leadership and support, and for their confidence in me in this endeavor.  I would also like to thank those who came after me, for asking the questions that needed to be answered.  I hope this paper has raised awareness on this topic.

 

Yours in Collecting,

 

 

Hudson A. Byblow