A Pantalon Clavichord from the Friderici stable?

In 2005 I was fortunate to be able to purchase from Odd Aanstad in Sweden an unusual Pantalon clavichord, almost certainly of Saxon origin.

The clavichord after restoration. No decision has yet been taken as to the decoration. Evidence points to it having been gilded and patinated. The blue paint is a much later addition

With Odd’s help and my own investigations I have been able to trace the instruments whereabouts back to about 1840. Odd purchased it sometime between 1990 and 1992 from the piano makers Schimmel in Brunswick – at this time the firm was disposing of its collection of antique instruments – but he remembers seeing it in Oslo in 1982. The then owner was at that time unwilling to sell the instrument, and although Odd asked to be notified if the owner changed his mind, the next time he encountered the clavichord was in the Schimmel workshops. How the instrument landed temporarily at Schimmel is itself an extraordinary story, for the details of which I am endebted to Lothar Bemmann. Lothar has had the clavichord on file for some years and in September 2007 had a telephone conversation with a previous owner Wilott Heerde (living in Mönkeberg bei Kiel). Mr.Heerde was in the 1980s in Stavanger, and noticed in the Oslo newspaper Aftenpostenan advertisement that an instrument was for sale in Oslo. He temporarily lost the advertisement but, on finding it again after a year, he phoned the seller, who still had the clavichord, and purchased it for NKr. 4000-5000 (between €465 – 581 at today’s rate). Mr.Heerde’s mother, who at this time lived in Gilzum, south-east of Brunswick, had plans to make an attractive table out of it. It was later wisely decided, however, that the good lady should seek her table elsewhere and that the clavichord should be restored. It was sold for ca. DM 3000.- to the Schimmel workshops, sometime between late 1987 and September 1989. Although Germany would be the natural geographical location for such an instrument, it would seem that this clavichord spent most of its life in Norway, a circumstance which surely contributed strongly to its survival.

In 1982 Odd learned something of its history from its owner in Oslo, a Mr. Veum, who said it came from a farm in the Fyresdal district of south Telemark. He remembered it from his childhood on the farm in the 1930s and remarked that “he found this curious, as the family had been violin players for generations”. The farmhouse also contained some furniture which had been bought at an auction on the demolition in about 1840 of an old “stave” church, and it could well have been that the clavichord was purchased at the same time. This assumption is logical when one remembers that most of the upper, or educated, class in Norway at this time came from Denmark, of which from 1389 to 1814 Norway was a part. Danish influence in the seventeen and early eighteen-hundreds extended to Schleswig-Holstein and to Hamburg and Lübeck. Higher education was centred on Copenhagen. This would have included the clergy, and my theory is that the clavichord, which is definitely not a country instrument, was brought to the area by a clergyman.

The clavichord is a five octave unfretted instrument, with an ebonized walking stick bridge and rose suggesting a Saxon origin. Incorporated within the clavichord is a pantalon mechanism of an unusual if not unique design. A single lever which notches into two positions gives the pantalon two possibilities; in the half position the strings from c upwards would be engaged by elongated “U”s in 0,75mm brass soldered to the tops of the pantalon tangents. My assumption is that these Us make a side contact with the strings enabling the strings to resonate freely when struck by the playing tangents. The notes from B downwards remain damped. With the pantalon in the fully on position, the bass pantalon tangents engage with the strings, and the the clavichord is fully undamped. The theoretical advantage of such a system is that the tops of the pantalon tangents, which are basically flat iron nails set in the pantalon board, do not need to be exactly positioned for the stop to work fully, as the Us allow the stop to function with slightly irregular tangent tops; once set, the tangents are non-adjustable.

Pantalon tangents with ‘U’s of brass soldered to the tops. Also visible, one of the blocks nailed to the hitchpin rail as a support for …what?

For reasons which I shall explain, I wondered briefly whether this system was original, but the escutcheon with its two positions seemed to be original, persuading me that the system was designed to operate with two possibilities. What made me doubt it initially was the following:

At some point the treble to tenor hitchpin rail fixed to the spine of the clavichord had warped upwards,  resulting in a step between the bass and tenor hitchpin rails measuring about 5mm. This means that the stringband would be at two different levels. The playing tangents could be adjusted to accommodate this, but it would have been far more difficult, although not impossible, to reset the pantalon tangents. My initial thoughts were that the U shaped extentions to the pantalon tangents were conceaved to compensate for the stringband being at two levels, but I decided that this was not the case. Apart from anything else the range of the soldered Us does not coincide with the range dictated by the warped hitchpin rail. I think the warp in the hitchpin block occurred early on in the instrument’s history, possibly due to an insufficiently dried piece of wood being used. There is evidence within the instrument to suggest that attempts were made to correct, or compensate for this warping. One, namely a clamp indentation in the veneer of the hitchpin rail, suggests to me a fairly recent attempt to flatten the the rail, surely doomed to failure on account of the sheer mass of the rail. Two earlier attempts to compensate are suggested by;

1: the insertion of a rough piece of wood at the fulcrum of the pantalon lifting lever, which had the effect of raising the entire mechanism.

2: two blocks of wood and a staple fixed to the two hitchpin rails. My thoughts were, that these could possibly have been supports, added later, for some sort of under-damper rail which could also have had the effect of evening out the string band, with or without normal listing. My experimental reconstruction of such a rail proved unsatisfactory, so eventually I evened out the stringband by means of a small additional nut on the bass hitchpin rail. My experimental board I left in place, no doubt causing confusion to future generations, as it was impossible to remove with strings in place.

The pantalon mechanism. The mechanism hinges from two support blocks left and right , glued to the front rail. It is operated from outside the instrument by means of a metal extension fixed to the central arm (missing). Note the two dowelled blocks, left and right, holding the unglued balance rail in place.

Despite its appearance, the instrument was a good candidate to restore to playing condition. The casework was sound and with suprisingly little twist; possibly the clavichord had been unstrung for many years. The keys were in good condition and all the playing tangents were present with the exception of f3. The fronts of the natural keys are covered in ebony, the backs with a fruit wood stained black. The sharps are of a blackened fruit wood capped with ivory. All the key guides were missing and there had been a futile attempt to remedy this with short lengths of steel wire introduced into the guide slots. I found remains of horn in the guide slots. The balance rail is not glued in place, but fixed by means of two blocks of wood at right angles to the rail, pegged to the bottom boards. All but 12 tuning pins were present.

The soundboard presented the worst problem. There were multiple cracks and it was clear that barring was loose. The bridge was still attached but only just. Someone had inserted a makeshift soundpost from underneath, presumably in an attempt to stop a complete collapse. There were no soundboard mouldings present, suggesting that someone, perhaps Schimmel, had began restoration work. On removing the soundboard I discovered that it was in fact a replacement. It was inscribed underneath in blue “copy pen”;

“Laaket til dette Fortopiano er Forarlei det af mig Anund Knudsen Brokke. 1877. Moland I Telemarken”

The translation of the Norwegian text reads;

“Lid (ie. soundboard) for this fortepiano made by me, Anund Knudsen Brokke. 1877 Moland in the Telemark.”

The unrestored soundboard from underneath, with the inscription from the 1877 restoration at the bottom. One small rib across the top right hand corner is not shown as it had fallen off. The line of the bridge (original) marked in blue copy pen is clearly visible. The joint of the two book-matched halves of the soundboard is in the centre, running between the two resin slits on the left.

This soundboard has almost certainly been in the instrument longer than the original. It is composed of only two pieces of wood, book matched, which is a technique regularly used for the backs of string instruments, but as I understand it, unusual for a clavichord soundboard. The Telemark region of Norway is known for violin making, so it is tempting to speculate that Anund Knudsen was an instrument maker and, if what Mr Veum said about his predecessors was correct, a player as well. The ribs and cutoff bar are made of good quality wood and are worked to a high standard. What is still an open question is whether they are original or are also replacements. Interestingly, there is barring in the “dead” area of the soundboard. Traces of markings under the soundboard, also in blue copy pen, suggest that care was given to the positioning, or repositioning, of the ribbing. There are notches cut in the liners,  which I assume are original, to receive the ends of some of the ribs. The soundboard has been thicknessed and is in places as thin as 1.9mm.. A piece has been taken out of the cut-off bar to accommodate the rose.

All in all, the impression is that the 1877 replacement soundboard was executed to a high standard and with care to reproduce what was originally there, assuming that the ribs are also replacements.

A possible reason for the soundboard problems, possibly also the  collapse (sic) of the original soundboard, became apparent during restoration; the tops of the wrestplank were slightly convex. Whether this was again as a result of the wood moving, or if the parts were fitted like this, is unclear. Strips of wood were added to level the tops of the wrestplanks prior to refitting the soundboard.

The inscription on the soundboard corroborated what I had learned of the the instrument’s history to date, so in an idle moment I decided to try to find out more about the man who had replaced the soundboard in this clavichord. On researching the Norwegian archives I was able to establish the following facts about Anund Knudson Brokke. He was born on the farm Brokke in 1850, the son of Knud Ols (1809 – 1885) and Haege Gjermundsdtr. Veum (1810 – 1890). In those days it was normal in Norway for the place of birth (e.g. a farm) to provide the surname. He had two brothers and one sister. In the census of 1865 he is listed as being sixteen years of age and helping his father out on the farm. He married twice, the first time in 1897 to Anne Talleivsdtr. Vadliplassen, who was born in 1867. She seems to have died the following year 1898, possibly in childbirth. Anund is listed in the 1900 census as a farmer, unmarried with no children, living on a farm named Liarheim.  According to the entry in “Fyresdal Gards- og ættesoge” (a documentation of Fyresdal farms and the families who lived on them) the farm Brokke was divided in 1882, with Anund buying the northernmost part from his father for Nkr. 14400.[i] The new farm he named Liarheim after the district. In 1911 he married Sigrid Talleivsdtr. Bergland, who was born in 1876 and died in 1931. There were apparently no children from this marriage either. Anund himself died at the Liarheim farm in 1933. On p. 850 of  the above-mentioned documentation he is he is listed as a farmer and a watchmaker; if he was indeed a music instrument maker, this was not considered of sufficient significance to have been mentioned. He would have been twenty seven or twenty eight when he repaired the clavichord. Anund himself had no children. Mr. Veum, the previous owner, is the grandson of Anund’s sister; his father took over the farm, and the clavichord, on Anund’s death.

My work on the pantalon register remains at this time unfortunately inconclusive. The fine adjustment of the Us of wire soldered to the tops of the pantalon tangents proved daunting, even to someone of my patience and motivation. I managed to set up a few notes and indeed when a reliable side contact with the strings was achieved an undamped effect was produced, proving that the system would work in principle. It is extremely difficult adjusting the bits of wire with the stringband in situ, particularly so as any bending has to be done close to the tangent head, which in turn has repercussions  at the tops of the U. The whole process was rendered more nervewracking by the tendency of the soldered joints to fail, and the impossible task of resoldering them with the stringband in place. The uneven heights of the pantalon tangents (I attempted to correct the worst of them) means that the fully engaged position is also unsatisfactory at present. My own feeling is, that a minute assessment of the situation, now that the instrument has stabilized, would be necessary, followed by complete destringing and restringing a course at a time to achieve a better result.

Most of the pieces of instrument’s cabriole stand survived. Given that a number of small bits were present, including some original dowels, I imagine the stand was more or less in one piece until relatively recently. Curiously, the two major bits missing were the two largest, namely the front and back stretchers, but the remains of the front stretcher attached to the two front legs were sufficient to determine accurately the original positions of the drawers. Amazingly, three of the four drawer supports had survived, giving support to the theory of a recent dismantling of the stand. The remains of the front stretcher also gave tantalizing suggestions as to how the panel was carved. With this information, and the design of the side stretchers as a guide, I remade the front and back stretchers using limewood – a furniture restorer has suggested that the original wood is birch a wood difficult to obtain in large enough pieces. The brass drawer pulls which I assume may be original to the instrument, as no other holes were present, and the blue paint was applied over them, were too big for the drawers, being folded over tops and bottoms of the drawer fronts.

The one-piece lid, which was extremely battered, shows signs of having something attached to it on the inside, possible a form of music desk. The inside of the lid has been painted in an ochre colour, possibly a grounding for a unrealised lid painting; natural wood shows in the places where the suggested music desk had been fitted, so this paintwork may be original.

The instrument at some stage was treated to a coat of blue paint. This was incompetently carried out – the paint had not even been stirred properly – and it seems, mercifully, that it ran out before the deed was completed. On close examination tiny flecks of gold are visible on all parts of the exterior casework and stand. One of the legs, although completely stripped, still shows evidence of two different colour treatments on the carved and uncarved sections[ii]. On the upper edge of the case side a dark band about 4mm. wide, tapering to a point at the back corner, is just discernable, giving the effect of an applied cap moulding  The mouldings on the top of the case sides have been cut directly into the wood, and have  been undercut on the inside, and blackened, again heightening the effect of an applied moulding.

An examination of the legs in particular suggests to me that colours were applied to the case, which was then gilded and patinated, so that the different colours showed through. The quality of the gold could be ascertained, I understand, with the application of Aqua Regia (Königswasser). Until a detailed examination of the surface of the casework can take place, one can only assume that the instrument was sumptuously decorated.

There are traces of a handsome brocade paper (Brokatpapier) lining the inside of the case walls above the soundboard and wrestplanks[iii]. A less successful paper with depictions of saints and prophets has been applied to the inside of the toolbox and the side of the bass hitchpin rail.The sizes of the individual motives are too large for the area treated and seems aesthetically out of place with the decorative approach suggested by the other remains on the instrument’s case. From the motives it is impossible to date this paper with any certainty; similar papers were in circulation from the 1760s to the middle of the nineteenth century, so an application later than the instrument’s manufacture need not be excluded.[iv]

The clavichord is unsigned and undated. Two of the prime candidates for a maker’s signature, namely the original soundboard and the nameboard, were both missing, however it is very tempting to associate the clavichord with one of the most prestigious workshops in Saxony at the time, namely that of Friderici.

The bridge, with its cutouts to guide the strings is very similar to those on the two surviving Friderici clavichords. The rose, presumed original to the instrument, is very similar to those used by Friderici.

A comparison of some basic dimensions is interesting but, of course, not conclusive.

Leipzig Friderici[v] Paris Friderici[vi] Anon. pantalon
Case length 1725mm 1740mm 1729mm
Case width 491mm 500mm (?) 473mm
Case depth 155mm 160mm (?) 167mm
Stichmaß 490mm 483mm 490mm
c2 length 272mm 264mm 273mm


I realise that this article poses more questions than answers. It represents work in progress, and I hope by making this information public we will learn more about this unusual clavichord.

I would like to thank the following people for their help with various aspects of this research: Odd Aanstad, Lothar Bemmann, Eva Helenius and Karin Richter.

[i]Steinar Marvik, Fyresdal Gards- og ættesoge, vol. 2, pub. Fyresdal Kommune, 1992, pp.849 – 850.

[ii]Just visible on the left front leg in the first illustration.

[iii]Very similar if not identical to illustration 37 in E.W. Mick, Altes Buntpapier, pub. Harenberg Kommunikation, Dortmund 1979, where it is listed as South-German, around 1760.

[iv]The paper most resembles a design in use around 1790 from Fürth, see ibid. Illustration 63.

[v]Dimensions from Henkel, Hubert, Clavichorde,Musikinstrumenten-Museum der Karl-Marx-Universität, Leipzig, Katalog, Band 4, Verlag das Musikinstrument, Frankfurt/Main 1981, pp.65-67. Further reference was made to the working drawing of this clavichord by Martin Kather.

[vi]Dimensions from the Paris Museum website.