Intriguing duet

[Myrtle's cook]

The following duet is being offered for sale by a Cumbria auctioneer (I'm not a party to this sale):

https://www.the-saleroom.com/en-gb/auction-catalogues/1818-auctioneers/catalogue-id-sr1810458/lot-94ff6842-2284-47b2-8349-acfb00ede64c

 

It looks to be a McCann duet.

 

Some unusual aspects.

-Firstly, to state the obvious, it is rectangular in form. Not unknown I realise, and I have a postcard of the Paget trio with six such rectangular instruments, although these appear based on the New Model or Wheatstone equivalent (i.e. raised ends and metal labels). Quite a few earlier rectangular instruments in the Concertina museum etc

-Its numbering -57473 - seems to follow the series used for ECs (or perhaps early Anglos??), not the 4 digit serial numbers one usually encounters on Lachenal duets

-It has an intriguing inner inscription - a photo of this is shown, and appears to read ' LACHNAL. & Co . MAKERS./[FO]R INVENTOR. J.H. MACCANN_/LONDON_ 1883. PLYMOUTH'. (sic)

 

Some thoughts:

As Lachenal is misspelt in the internal inscription one has to wonder if it was really applied by the makers or MacCann himself.

As it doesn't use 4 digit Lachenal serial numbers is it a very early duet?

The patent was awarded for Maccann's duet in 1884, so 1883 could fit for a prototype - if the inscription is treated literally and we make the link that the date actually applies to the instrument, rather than just being a sort of commentary on the instrument inscribed by an owner

If this is the case then isn't 57473 rather later than 1883 (and the patent was awarded for Maccann's duet in 1884 if I recall correctly) or an early duet? If following the EC sequence this would be second decade of C20th (or later?). Or is it part of the Anglo sequence of numbering, which I confess I know little about.

 

What do people think?

[Bill Crossland]

Dowright has an anglo attributed to 1883 with serial number 74693, so the anglo sequence doesn't fit....

[Jake Middleton-Metcalfe]

It looks like it was played quite a lot - the buttons are quite worn on the sides, it must have been well appreciated at some stage.

[Stephen Chambers]

Intriguing indeed!

 

The Provisional Specification for Maccann's Patent, dated 11th March, 1884, states "They are made in either shape hexagonal or square with 39 to 58 keys." Whilst the Patent drawing shows a hexagonal instrument with the same button layout as this one, including the two offset buttons (one each side) which are given no note values - so quite likely they were for a whistle and squeaker on both.

 

But the claimed serial number makes no sense for 1883 - perhaps there has been a misreading?

 

 

Edited April 5, 2021 by Stephen Chambers
Edited typo.

[wes williams]

This non-standard duet serial number jogged a distant memory from around 15 years ago - see Lachenal 43438

[Stephen Chambers]

5 hours ago, wes williams said:

This non-standard duet serial number jogged a distant memory from around 15 years ago - see Lachenal 43438

 

Hmmm, so now we seem to have two anomalous Maccann duet serial numbers that look like they might belong to the English system number sequence:

 

   57473 (circa 1919) on a supposedly 1883 "Prototype" instrument

 

   and

 

   43438 (circa 1905) on an instrument in the Wayne Collection at the Horniman Museum...

 

However, the catalogue description of the latter contains a further anomaly in that it states that the instrument has a "Circular reed pan label: 'Lachenal & Co, 8 Lit James St, Grays Inn Road, WC, London'. Serial number: 43438." - and such reed pan labels are only found in English system concertinas... So could we perhaps be looking at a duet that has been cobbled together with the reed pans of an English inside it? 😲

[wes williams]

Another Horniman oddity to be considered, which is definitely an English->Crane and shows that conversion isn't simple: 50584

[Stephen Chambers]

On 4/1/2021 at 9:22 PM, Stephen Chambers said:

... the claimed serial number [57473] makes no sense for 1883 - perhaps there has been a misreading?

 

I wonder if the lead digit might be a poorly stamped 2 that has been misread as a 5? Though (according to Dowright's calculations) 27473 would suggest a date of c.1887