[Bruce Reding]

Folks,

Fellow HEQ forumner David Johnson (dwjquest) has been compiling serial numbers for Omega 2400 Marine Chronometers. I've attached his results to date. If you have a 2400, or have a picture of a 2400, that has a serial number other than those listed, we would appreciate it if you could post a picture of it with the serial number clearly showing or stated. If possible, it would also be nice to know if it has a 1511 or 1516 movement, and, if you have the Besancon certificate, the certificate number and date. (These are the first two entries on the certificate.) If you would prefer to remain anonymous, you may e-mail the pics and information to me, and I will see that David gets it.

The 2400 is a remarkable instrument that is unique in the annals of horology. Introduced in 1974, it was the first (and remains one of only three) wristwatch to use a megahertz range AT cut quartz crystal. This extremely demanding and cutting edge technology made it by far the most accurate wristwatch ever made to that time. Even today, few watches can outperform it. By compiling serial number information, we can make inferences about the production of this remarkable piece. For instance, from what David has already collected, we can infer that there were at least two distinct batches made. (Probably separated by the 1511 and 1516 versions.)

We appreciate your help in piecing together an intriguing bit of horo-history! :-!

 

[Bruce Reding]

Three new ones collected so far. Folks have been shy about posting so far, but good results. |>

 

[andymlow]

My father, Frank Marlowe has #37057767. He does not have the certificate. I believe it has the 1516 movement.

 

[CFR]

I think Omega changed from cal 1511 to 1516 somewhere between SN 37057440 and 37058156...

.... so I'd be curious to know if your dad's is a 1516. I attached some pics showing how to tell the difference without removing the caseback. It's subtle, but the 1511 case is indeed larger and more sloped than the 1516, which is slightly "snubbed." In the 3rd photo, the 1516 is the smaller one on the right (34mm long, vs. 38.5mm long for the 1511). They look identical from a top view (not shown), though.

 

[CFR]

And here are the movement differences...

...can't remember if I posted these pics before. In the first photo, the cal. 1516 movement is on top, and the 1511 is on the bottom.

 

[Stanford]

I recently obtained this one. This is my first time on this forum so I hope both the text and pictures come out as intended.

One positive thing to come out of taking the pictures is that I can clearly see (from the pictures of the movement) that the bracelet pins are very bent and in urgent need of replacement!

 

[Bruce Reding]

I recently obtained this one. This is my first time on this forum so I hope both the text and pictures come out as intended.

The pics turned out well. A nice, complete set! :-!

 

[Bruce Reding]

My father, Frank Marlowe has #37057767. He does not have the certificate. I believe it has the 1516 movement.

Thanks Andy! Does he still wear it?

 

[Bruce Reding]

Re: I think Omega changed from cal 1511 to 1516 somewhere between SN 37057440 and 37058156

.... so I'd be curious to know if your dad's is a 1516. I attached some pics showing how to tell the difference without removing the caseback. It's subtle, but the 1511 case is indeed larger and more sloped than the 1516, which is slightly "snubbed." In the 3rd photo, the 1516 is the smaller one on the right (34mm long, vs. 38.5mm long for the 1511). They look identical from a top view (not shown), though.

Interesting, Craig! Note that in David's compilation (the first post in this thread) that the case shape apparently transitions from the long to the snubbed after 37057517 and before 37058065. If one can assume that the transition to the snubbed case shape corresponds to the transition to the 1516 movement (which seems reasonable to me), then these provide tighter bounds.

Note also that watch number 37958522 in Sanford's post is clearly a 1516, and the Besancon certificate is number 1007. This would seem to support the conjecture that the upper limit on the number of 1511s made is around 1000.

 

[CFR]

Re: I think Omega changed from cal 1511 to 1516 somewhere between SN 37057440 and 37058156

Interesting, Craig! Note that in David's compilation (the first post in this thread) that the case shape apparently transitions from the long to the snubbed after 37057517 and before 37058065. If one can assume that the transition to the snubbed case shape corresponds to the transition to the 1516 movement (which seems reasonable to me), then these provide tighter bounds.

Note also that watch number 37958522 in Sanford's post is clearly a 1516, and the Besancon certificate is number 1007. This would seem to support the conjecture that the upper limit on the number of 1511s made is around 1000.

Yes, thanks for the clarification, Bruce. Regarding numbers for the 2 flavors of Marine Chronometer movement -- the cal 1511 and cal 1516 -- Omega's website says that there were 1000 of the cal 1511 produced and 8000 of the cal 1516 produced. I've always doubted the figure for the 1511 -- I think it's too low -- because it seems like the cal 1511 pieces are far too common in today's market for there to have been only 1000 of them. And I question the ratio of 1511's to 1516's as well, because when I've looked on eBay, I'd say that about 30% of the Marine Chronometers I've seen are cal 1511, and the rest are cal 1516 (certainly not a 1:8 ratio!). Intuitively, it doesn't seem like there are tons of cal 1516 pieces out there -- 8000 seems quite high.

So assuming our MCs were cased sequentially (i.e., all cal 1511's first, followed by all cal 1516's), then here's what I figure, using rough numbers based upon the serial plates that we've seen:

Cal 1511 = 34914700-34917000 (2300 pieces) + 37057000-37057800 (800 pieces) for a total of about 3100 pieces

Cal 1516 = 37057800-37061900 = 4100 pieces

That would put total production at about 7200 pieces, with about 25% being cal 1511 and 75% being cal 1516. Somehow that seems more plausible than the 1000/8000 figures, although I'd love to know the answer for sure.

 

[Bruce Reding]

I agree with your hypothesis, Craig, that the reported 1000/8000 split is probably erroneous. When I first saw David's compilation, it was very obvious that the ratio was nowhere near one to eight. In fact, in his first compilation posted below, the split is 15 to 17. (He's since collected six or so more, but I don't know the new results yet.) While this is a small sample set, it is statistically unlikely to yield this ratio if the underlying population is 1000 and 8000. (An interesting challenge calculating the actual statistics. I won't do it this morning, but I'd bet the odds are less than 1 percent.) This leaves three possibilities in my mind:

1. The transition from longer to shorter cases does not correspond to the transition from the 1511 to the 1516. This doesn't seem likely, but should be examined. Do we have any known examples of a 1516 movement in one of the longer cases?

2. There's biased selection phenomenon present. (A "Dewey Wins" scenario.) Hard to imagine what this would be, though. Also, you'd still have to postulate 1516s in older style cases or non continuous serial numbers.

3. The ratio was in fact much more even.

The last feels the most likely to me. I generally agree with you serial number analysis. Two slight worrisome notes, however. The first is the existence in David's list of a 35XXXXXX serial number. (Interestingly, per your analysis, this would only further even the ratio.) The second, more worrisome note is that watch number 37058522 posted in this thread by Stanford, which is clearly a 1516 movement, holds Besancon certificate number 1007. While I would not expect a lockstep relationship between manufacturing order and certification order, I would generally expect them to be within a few hundred of each other, corresponding to sizes of the batches sent off to Besancon.

 

[Bruce Reding]

The second, more worrisome note is that watch number 37058522 posted in this thread by Stanford, which is clearly a 1516 movement, holds Besancon certificate number 1007. While I would not expect a lockstep relationship between manufacturing order and certification order, I would generally expect them to be within a few hundred of each other, corresponding to sizes of the batches sent off to Besancon.

I found this on eBay this AM:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=330069211772&fromMakeTrack=true

It's a 1516 with a Besancon certificate number of 212. Puts my second concern above to rest. It makes my wonder, though, about the scheme for sending watches to Besancon and/or how the resulting certificates were numbered.

 

[CFR]

I found this on eBay this AM:

http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=330069211772&fromMakeTrack=true

It's a 1516 with a Besancon certificate number of 212. Puts my second concern above to rest. It makes my wonder, though, about the scheme for sending watches to Besancon and/or how the resulting certificates were numbered.

I saw the "Buy It Now" price and thought it was $200-$300 on the low side . . . until I read that the watch doesn't work!! This translates to: "Add 890 Swiss Francs plus insured overseas shipping (about $800 USD) to the price, because that's the flat fee that Omega charges to fix it, AND wait at least 5 months for Bienne to fix it IF all goes well." But I digress....

Back to the point at hand . . . here are my caseback numbers/Bescancon certificate numbers/Bescancon certificate dates:

caseback 34916232, cal. 1511, certificate 588, dated 2/24/76
caseback 37057395, cal. 1511, certificate 1165, dated 6/2/76
caseback 37058157, cal. 1516, certificate 642, dated 11/9/76

Do you think Bescancon assigned certificate numbers by movement type (e.g., certificates for cal 1511 probably started from #1 or #100, and certificates for cal 1516 also probably started from #1 or #100)? If not, wouldn't we probably see higher and more random certificate numbers for at least some Omega Marine Chronometers, especially if Bescancon also issued certificates for non-Omega movements during this timeframe?

I wondered 2 other things:

(1) What are the highest and lowest known Bescancon certificate numbers for the cal 1511 and 1516 movements, along with their dates?

(2) An unrelated question . . . can David (or you) double-check the existence of that 35xxxxxx serial number? And do you know if it was taken from an advertisement or from a "bona fide" existing watch? That 350xxxxx number made me really curious. I tried to magnify the posted image but it's too small.

It would be interesting if Bescancon could tell us how many certificates they issued for cal 1511 and 1516 movements, along with issue dates, movement numbers, and Bescancon certificate numbers. I'll bet Bescancon would know more than Omega about production numbers for the Omega Marine Chronometers! I'd specifically want to know the issue date/movement number/certificate number for: (1) The first cal 1511 that Bescancon received in the 349xxxxx series; (2) the last cal 1511 they received in the 349xxxxx series; (3) EVERYTHING about cal 1511 movements in the 350xxxxx series; (4) the first cal 1511 they received in the 370xxxxx series; (5) the last cal 1511 they received in the 370xxxxx series; (6) the first cal 1516 they received; and (7) the last cal 1516 they received. I think those 7 pieces of information would clear up a lot. If I were French, I'd take up this inquiry task!

 

[Bruce Reding]

That's an excellent set of questions. I also have found difficulty communicating with the French speakers at Omega (as did my wife, who speaks French quite well, which made me wonder if it was just the individual we were dealing with.) I wonder if Gino feels up to the task? I'll e-mail him.

David would have to answer the question about the 35xxxxxx.

FWIW, for my watch, which is serial number 34916529, my certificate is numbered 132, and is dated 19 Septembre 1975. I had a theory that they started over each new year, but your series of serial numbers put the lie to that one.

 

[dwjquest]

Bruce and CFR are too sharp for me!! The serial number was listed incorrectly. It should be 37057280 and it looks like a calibre 1511 from the case design.

 

[Bruce Reding]

Re: I think Omega changed from cal 1511 to 1516 somewhere between SN 37057440 and 37058156

Omega's website says that there were 1000 of the cal 1511 produced and 8000 of the cal 1516 produced.

Craig -- I went to Omega's website, and here is the quote that I was able to find:

With its derivatives 1510 (calibre standard) and 1511/1516 (marine chronometers), around 10 0000 Megaquartz 2400 units would be produced, of which 8000 units were of the 1516 calibre.

It doesn't state how many were 1510 vs. 1511, but, per their info., the two added together numbered around 2000. I agree that this feels doubtful.

 

[Bruce Reding]

Bruce and CFR are too sharp for me!! The serial number was listed incorrectly. It should be 37057280 and it looks like a calibre 1511 from the case design.

Thanks for the clarification, David! :-!

As a general question, do you feel that your pics are good enough that all the serial numbers you've collected so far are unambiguous?

 

[CFR]

Re: I think Omega changed from cal 1511 to 1516 somewhere between SN 37057440 and 37058156

Craig -- I went to Omega's website, and here is the quote that I was able to find:

With its derivatives 1510 (calibre standard) and 1511/1516 (marine chronometers), around 10 0000 Megaquartz 2400 units would be produced, of which 8000 units were of the 1516 calibre.

It doesn't state how many were 1510 vs. 1511, but, per their info., the two added together numbered around 2000. I agree that this feels doubtful.

Ah, the cal 1510 . . . you HAD to mention that, didn't you, Bruce? :-s I've REALLY wondered about the number of those produced. Based upon what you've seen for sale, doesn't there appear to be even fewer 1510's than 1511's and 1516's out there? I rarely see 1510's for sale anywhere.

I've also wondered about cal 1510 dial options. We know Omega made this model with a sparkling blue aventurine (lapis) dial. But I also recall seeing a matte blue dial, perhaps in 2 variations -- I think one might've had a waffle texture and one might've been smooth, but I'm not positive. I thought those variations were genuine Omega dials but of course I'm not certain; because the aventurine dials were very breakable, these might've been aftermarket dials, although I recall doubting that at the time I saw them.

Presumably Bescancon would know nothing about the cal 1510's, because those movements weren't submitted for certification.

Back to production numbers . . . an obscure section of Omega's vintage website does specify that the cal 1511 was "produced in a total quantity of 1'000 pieces only." This means 1000 of the cal 1510 pieces were allegedly produced, because as you said before, Omega's website specifies 10,000 pieces total for the cal 151x movements (8000 of which are supposedly cal 1516).

 

[Bruce Reding]

Re: I think Omega changed from cal 1511 to 1516 somewhere between SN 37057440 and 37058156

Ah, the cal 1510 . . . you HAD to mention that, didn't you, Bruce? :-s I've REALLY wondered about the number of those produced. Based upon what you've seen for sale, doesn't there appear to be even fewer 1510's than 1511's and 1516's out there? I rarely see 1510's for sale anywhere.

When I first became vaguely aware of this movement family, it seemed to me that I saw as many "non gold rimmed" (i.e., non marine chronometers) as "gold rimmed" models. This is hazy recollection, though, and may have been influenced by the places I browsed, such as Gisbert Joseph's site.

 

[Bruce Reding]

Many years ago, I took a short course in data analysis from Stu Hunter, Emeritus Professor at Princeton and coauthor along with Box and Hunter of "Statistics for Experimenters", which is an enduring classic on experimental design and data analysis. There was one lecture I'll never forget. He shared with us a paper that one of his former grad students had proudly sent him. We went through page after page of abstruse, high level data analysis. I was quite impressed. When we finished the review, Stu put the first slide back on the overhead projector, pulled out his red grease pencil, and wrote a large "D minus" on the title page. I was surprised and perplexed. The analysis had been a masterpiece of subtlety with no holes that I could see. Stu's beef, it turned out, was that there was not one plot or graph in the whole paper. His former student had never actually looked at the data.

That lesson has always stuck with me, and the ghost of Stu has been whispering in my ear that the OMC serial number data is ripe for a visual look see, especially since I just got an excellent update from David (dwjquest). I chose to do a simple scatter plot, which I've attached. It's crude, but revealing, as I think you'll see.

Before we get into a discussion about it, let me describe the scatter plot first. The number lines cover the ranges of the two groups of serial numbers (one for the 349xxxxx series and one for the 370xxxxx series), and are broken into major intervals of 500. Note that there is a discontinuity between 17000 and 46000 in the first line. Each point above the line represents one of the serial numbers collected by David. They are from his latest data set of 43 (good progress since this thread was started!), and have been corrected as per the discussions in this thread. In case my most excellent handwriting isn't totally clear roll: ), the first note states that question marks are used to denote numbers for which there is some doubt about their correctness. There are only four of these, and David feels that there's a good chance that they're correct. The second note states that points in close proximity are stacked. I did this to make it clear that there are two distinct serial numbers at these locations. Finally, the hatched area on the 370 number line indicates the transition from 1511 movements to 1516 movements. There are no known 1516s below this region, and no known 1511s above it.

With this as background, here are my observations:

1. There were two major series -- the 349xxxxx and the 370xxxxx. (This, of course, is a total No Duh. Something else I've learned in my career, though, is that one should always start with the obvious.) Clearly, Omega made two major batches, probably waiting for the first one to largely clear out before committing to the second. The interesting thing here is that the first part of the second batch used 1511 movements. Almost certainly these were leftovers from the first 1511 manufacturing run. (More on this later.)

2. In the 349xxxxx series, there is a significant gap at the start. The "dry spell" after the first serial number (34914743) is as large as the span among the next eight (34915904 to 34916963). While there's no guarantee that this gap isn't just due to our sampling, I strongly suspect that it's significant, and indicates an underlying gap in production. If so, then, assuming nonbiased sampling, the batch of which 34914743 is a part must have been quite small. Possibly this batch was made to iron out kinks in the production process and/or to get product to market as early as possible. (Basel or some equivalent, perhaps?) If the gap is real, then the span among the contiguous eight (from 34915904 to 34916963) would imply that there were somewhere around 1000 MCs in the 349 series. (More on this later as well.) It's interesting to speculate about the intervening watches. Could they have been the 1510s? (Do we have serial numbers for any 1510? That could tell us a lot.) Finally, this gap makes the first watch "special" in my mind, sort of like a first edition. Do we know anything about this watch? Does it have special provenance such as residing in Omega's collection or some such?

3. There is a much larger gap at the end of the 349 series, with last serial number being 32,000 after all the previous, which span only 2200. Is this another possible misread? (Not poking at David. It's just the kind of question one must always consider in data analysis.) If not, then evidently this sub-batch was quite small as well.

4. By comparison, the 370xxxxx series has no similarly obvious gaps. (This doesn't mean that there are none, but a lot more data would have to be filled in here to conclude with any confidence that a gap or gaps exist. Also, if they exist, they must be substantially smaller than those seen in the 349xxxxx series.)

5. Assuming that the transition between 1511s and 1516s is a clean one ( i.e., they're all 1511s until the transition, after which they're all 1516s), then the split between the two is 18:25, or 42% vs. 58%. Clearly this is a far cry from Omega's stated 1:8 (11% vs. 89%) split. This leads me to strongly suspect, as discussed before, that the info on Omega's site is wrong. I further conjecture (with less conviction) that the error stems from their confusing the 349/370 split with the 1511/1516 split. Per the gap analysis in point two above, there appears to have been around 1000 in the 349xxxxx series. Also, it is not inconsistent with the data to think that there may have been as many as 8000 in the 370xxxxx series (albeit this latter requies a bit of "sampling luck"). The person compiling the stats may have overlooked the fact that a substantial portion of the 370xxxxx series used 1511s at the beginning.

Well, that's my thinking. I just hope Stu isn't spinning. :-d