I already made the attempt of a number system in the

serial number thread. This attempt is a little bit outdated, because after this many numbers were added. Therefore I will it update here for an alternative estimation (the E2N 71000021 was not included, it is one of mine):

51:

5100025 E2

5100035 E2S

5100055 E2

52:

5200007 E2S

5200011 E2S

5200014 DS-515

5200015 DS-515

5200022 DS-505

5200028 E2S

5200028 DS-515

5200044 DS-505

5200057 DS-505

5200067 E2

5200072 E2

5200075 E2

5200107 E2

53:

5300026 E2

5300053 E2

54:

5400001 E2S

5400006 E2

5400013 DS-515

5400016 DS-515

5400040 E2S

5400062 E2

5400063 E2

61:

6100034 DS-505

62:

none

63:

6300024 E2N

6300030 DS-505A

6300031 DS-505A

64:

6400009 E2N

6400006 DS-505A

6400009 DS-505A

6400034 E2N

6400045 E2N

71:

7100006 E2N

7100007 E2Ns

7100015 DS-505A

7100021 E2N

7100030 E2N

7100064 E2N

7100067 E2N

7100076 E2N

7100103 E2N

72:

7200003 DS-505A

7200004 DS-505A

72000024 DS-505A (typing error? -> 7200024)

7200031 E2N

7200035 E2N

7200071 E2Ns

7200080 E2N

73:

7300049 DS-505A

7300102 E2N

74:

7400023 E2N

7400029 E2N

74.1:

7410003 DS-515A

7410005 DS-515A

7410006 DS-515A

7410010 E2N

7410058 E2N

7410073 E2N

81:

none

82:

8200045 E3

8200046 E3

8200067 E3

83:

8300008 E3

8300015 E3S

8300022 E3

8300031 DS-565

8300032 DS-565

8300035 E3S

8300036 E3S

8300053 E3S

8300058 E3

8300090 E3

8300100 E3

83.1:

8310090 E3

84:

8400025 DS-560

8400043 E3

91:

9100012 E3S

9100024 E3S

9100024 DS-565

9100032 E3S

9100065 E3

9100092 E3

92:

9200003 DS-565

9200005 DS-565

9200006 E3S

9200015 E3S

9200037 DS-560

My assumptions from the first attempt are confirmed: The first digit is the last digit of the year (e. g. 5 -> 1995), the second digit will be the quarter of the year (no number higher than 4). With the last 3 digits the cameras are counted within the quarter.

This means, that there is no limit within the assumed "batches". The real batch size will be much smaller. We have a few numbers higher than 100 in the last digits. No problem, sometimes there were more than 100 cameras within a quarter.

I don't think, that each model was counted separately. This means, an assumption of 100 cameras of each known "batch" (-> quarter) will be wrong. We still have 2 numbers, which were found on different cameras. At least one of these cases was confirmed. I cannot explain this.

i will try to summarize the production according to time:

1995, Jan - Mar (51): Start of production (this means, that the official release year of 1994 could be wrong, until now there is no camera of 1994). At least 55 cameras were made.

1995, Apr - Jun (52): At least 107 cameras were made.

1995, Jul - Sep (53): At least 53 cameras were made.

1995, Oct - Dec (54): At least 63 cameras were made.

1996, Jan - Mar (61): At least 34 cameras were made. End of production of the 1st generation.

1996, Apr - Jun (62): Gap between 1st and 2nd generation, no production.

1996, Jul - Sep (63): Start of production of the 2nd generation, at least 31 cameras were made.

1996, Oct - Dec (64): At least 45 cameras were made.

1997, Jan - Mar (71): At least 103 cameras were made.

1997, Apr - Jun (72): At least 80 cameras were made.

1997, Jul - Sep (73): At least 102 were made (Although we have only 2 of them!) .

1997, Oct - Dec (74): First part: at least 29 cameras. Then there was a change in manufacturing (whatever this could have been), Afterwards there were again at least 73 cameras (numbers beginning with 741). End of production of the 2nd generation.

1998, Jan - Mar (81): Gap between 2nd and 3rd generation, no production.

1998, Apr - Jun (82): Start of production of the 3rd generation, at least 67 cameras were made.

1998, Jul - Sep (83): At least 100 cameras of the first part, afterwards at least 90 cameras of the second part (numbers beginning with 831).

1998, Oct - Dec (84): At least 43 cameras were made.

1999, Jan - Mar (91): At least 92 cameras were made.

1999, Apr - Jun (92): Still after the announcement of the D1 at least 37 cameras were made. End of production.

If we add these numbers, we will not get the real total number. But it is a hint for the distribution of the total number on the generations:

1st generation (E2, E2S, DS-505, DS-515): at least 312 cameras

2nd generation(E2N, E2NS, DS-505A, DS-515A): at least 463 cameras

3rd generation(E3, E3S, DS-560, DS-565): at least 429 cameras

total: at least 1204 cameras

There a two possibilities to estimate the distribution on the generations: From the known numbers in the list and from the estimated number, if we assume, that no number was omitted.

From known numbers:

1st generation: 26 cameras, 30%

2nd generation: 34 cameras, 39%

3rd generation: 28 cameras, 32%

From assumed numbers, which were at least produced in each quarter:

1st generation: 26%

2nd generation: 38%

3rd generation: 36%

The real numbers will be anywhere between this. I am sure, that the 2nd generation is the most numerous (but is not new, we knew it already

). The 1st generation seems to be the rarest.

If we consider, that the numbers found "at least" for each quarter are lower than the real, a total number of 2000 seems to be likely. This means, that it is unlikely, that the Nikons and Fujis were counted in parallel rows.

Of course we need more serial numbers for better statistics. This would allow to estimate the numbers of each model (which would be the most interesting). And perhaps the problem with the doubled numbers can be clarified.