LACHENAL CONCERTINA PRODUCTION
(Dating of Lachenal Concertinas)
By David Aumann
The story goes that Louis Lachenal, who had been employed by the Wheatstone
company, left to set up his own business (possibly taking some of Wheatstone's
tools, and even some employees) around 1848-1850. He is thought to have started
up manufacture around 1850. With his Swiss toolmaking background, he was able
to machine up to out-produce Wheatstones. For example, whereas each of the Wheatstone
fulcrums for the button levers was painstakingly hammered in by hand, Lachenal
devised a machine to push an entire set into the action board in one press. It
is thought that concertinas numbered up to about 20000 are labelled Louis Lachenal,
but after that they are labelled Lachenal and Co.
Records of Lachenal serial numbers are reputed to have been lost during World
War II bombing raids. In their absence, a 'rule of thumb' has been devised (by
Geoff Wooff, Concertina Magazine, #6, Spring 1983), based on the following premises:
(i) Lachenal began production in 1850, and ceased in 1936 (i.e. he was in
business for 86 years),
(ii) in that time he made 350000 anglo concertinas,
(iii) his output was constant.
Mathematically, 350000 concertinas in 86 years represents 4070 p.a. Therefore
you should be able to estimate an age from the formula
Date of Manufacture = 1850 + (serial number)/4070
How reliable is this? We could start by asking how reliable the starting premises
are. The 1850 commencing date seems reasonable, based on the reminiscences of
George Jones, which have been published in Free Reed and Concertina and Squeezebox,
and are available on the WWW. The 1936 finishing date is documented in an interview
which Neil Wayne did in the 1970s with Lachenal employee Tommy Williams, so they
are not limiting factors in our formula. The production total of 350000 was unjustified
in the original article, and is probably something of a guestimate. Does anyone
out there own a Lachenal with a serial number of over 300000 ? Over 250000 ? I
would be interested to hear from you. The constant output assumption is obviously
rather dicey.
To test the formula, I have been trying to collect known date of manufacture/serial
number combinations. I have written dozens of letters with precious little information
coming to light. My idea was that concertinas were sometimes sold with a retailer's
stamp and a date inside them. All of us who own concertinas get to see the insides
of our instruments more than we would sometimes like, so if yours has any such
markings, please let me know. If your instrument has a repairer's label and date
inside, please tell me ... this would put an upper limit on the date of manufacture.
After all my research, I have come up with a few combinations, from Chris Algar
(Barleycorn Concertinas) and Richard Evans (Australian maker/repairer). These
combinations are:
Serial Number
|
Date of Manufacture
|
|
|
1
|
1850 (likely starting date)
|
11653
|
1860
|
18197
|
1868
|
51480
|
1895
|
140375
|
1908
|
350000 (possible total production)
|
1936 (known finishing date)
|
These combinations can be graphed against the dates predicted from the formula
(using Excel spreadsheet) with the following results:
A few observations are readily made:
1. Lachenal's production began slowly but seems to have really accelerated
in the 1890s,
2. There is clearly a need for more data, but I'd be guessing that the curve
of Lachenal production would be S-shaped, and would NOT reach 350000. (To reach
350000 after our last known point would necessitate an annual production of almost
11000 concertinas, at a higher rate than any time in the company's history. The
environment of post-war re-establishment of industry and the depression of the
1930s would not have been conducive to this. Additionally, the gradient of the
production graph would probably be tending to zero near 1936 ... after all, if
your company is in full swing, you probably wouldn't be shutting down! The invention
of the gramophone around 1930 must have had its effect on instrument makers, too.
3. The length of a horizontal line between the two graphs would represent
the error in the formula's prediction. Note that, for serial numbers of about
50000, the formula says ca. 1865 but known dates indicate ca. 1895.
4. A slight rearrangement of data gives the following graph, which will allow
you to find your serial number on the horizontal axis and read your date from
the vertical axis.
The trendline which has been fitted to this data may be a little misleading,
since it implies extremely high production between about 1905 and 1915,
but with present data, it may be the best way of estimating your Lachenal's age.
The only other dating info on Lachenals that I can think of is the actual
specifications of the instruments, which seem to have changed from time to time.
For example, my earliest 20-key has medium diameter buttons (it's a Louis Lachenal,
as opposed to Lachenal and Co., #12113). My next, #25219 has larger diameter buttons,
and later models, e.g #124339, have buttons of smaller diameter than either. The
two early ones have numbers on the keys. Looking at rosewood ended models, the
fretting is identical on #74681 and #57778, and these differ only minorly from
#12113 (only around the fretting around the serial no. in the left hand end.)
By #165981 the fretting is a totally different design. There are other differences
I could document if others are interested. If dates of specification changes could
be pinpointed or narrowed down, then you could perhaps date a concertina by it's
combination of features. Anyway, that's just a thought, and I doubt if enough
info would ever surface to allow dating a concertina by its specifications.
This sort of investigation is limited by the availability of data. In particular,
if your Lachenal has a high serial number (say, around 250000 or higher) could
you let me know, even if you don't
have a date for it. And if you DO have a serial number/date combination, please
email me so we can include it in our analysis. It's all a bit academic, but lots
of us would like to know how old our box really is!