Some other Swiss watch makers (eg Longines) have detailed
records of every movement they produced and a watch can be accurately
dated from this alone. The West End Watch Co. were also meticulous about
numbering their watches, though the number is usually on the case back
rather than the movement.
Unfortunately they don't seem to have
been so meticulous in keeping the records matching the serial numbers
so it's not possible to directly date using them. However, with some
assumptions, a lot of watches and some luck it will be possible to build
up a database and make it possible to get rough dates for serial
numbers.
The Assumptions
- The serial numbers are sequential. (That's a biggie)
- As the number is on the caseback there is a small risk that over time, as dead watches are used for spares, that a caseback can end up on a different watch to it's original mate. I make the assumption that this swap rate is low enough that, on any given watch, the chances of it being swapped are very very low.
The Watches
While I'd like to use only watches I owned, or at least had handled, realistically that's not going to produce a large enough sample to get a meaningful result. Luckily, in the days of the interweb, seeing high quality pictures of lots of watches is very easy, thanks almost entirely to eBay...
The Luck
Even with lots of pictures, you still need some external references to match a date to a number. There are several obvious methods we can use: inscriptions, catalogues, adverts, receipts and other documentation, perhaps even old photos.
In lieu of the above you can also attempt to roughly date a movement from it's style and features, though this is unlikely to be more accurate than to within about a decade.
Oh well, lets carry on regardless and see what emerges...
The Bad Bit
I'm going to come immediately unstuck here. It actually appears that serial numbering was somewhat haphazard and at least the watches with Longines movements follow a completely different numbering scheme. Furthermore, some watches have a stamp that appears at first glance to be a serial (two sets of four digit numbers, sometimes prefixed with a letter) but actually is more like a model number and/or import mark and the watch may or may not have a serial number also stamped beside it or elsewhere.Oh well, lets carry on regardless and see what emerges...
Hi Nick,
ReplyDeleteCan you help with some information on westend and longines connection? I have a Westend Imperator pocket watch which has longines stamped on the case.
I know this is well over a decade old, but I got myself a West End Matchless and been looking for info and found your post. I have just opened it and I think it finishes off something on your Methodology page.
ReplyDeleteMine is case 6657 165, serial 3436359.
Your own Matchless, the one in the 4 August 2012 post, is case 6657 2067, serial 3438261.
3436359 minus 165 = 3436194
3438261 minus 2067 = 3436194
The same base number on both. So on case reference 6657 the second group stamped on the back is the watch’s position in the batch, and the serial runs in step with it:
serial = 3436194 + the case sequence number
Mine is the 165th of the batch, yours is the 2067th, and the two watches are 1902 apart on both numbers. So the two group case number is not just a model number or an import mark. The first group identifies the case and the second counts the watches built to it.
It also gives a quick test for a swapped back. If the serial does not equal 3436194 plus the sequence number, the back does not belong to the watch.
Your spreadsheet may already hold more of this than you realised. Two rows carry the same short number in both the case number and the serial: case 2997 12 with serial 12 6252, and case 6656 50 with serial 22822 50. Worth checking those against the actual watches, as I am only reading your transcription.
One other thing. Mine has two caseback covers. The first is stamped EVERBRIGHT STEELBACK and the serial. The second carries the serial again and C.S.(I). Everbright is West End’s own name for stainless steel and the usual account is that they began using it in the 1930s, so if our two watches really are from one batch that would put yours later than the 1925 you have against it. The case references either side of ours in your own data point the same way, with 6654, 6656 and 6667 all carrying 1940s presentation engravings.
There are also watchmakers’ service marks scratched inside, 34144JB and 35545JB, and what looks like ww5646/64 M270.