The Single-manual Italian harpsichord in the Royal College of Music, London, Cat. No. 175:

An Organological Analysis

Grant O’Brien, Edinburgh, 2008

 

Figure 1 - The single-manual Italian harpsichord, attributed here to Onofrio Guarracino, Naples, c.1675 Royal College of Music, London, Catalogue No. RCM 175.

 

Introduction and brief description:

         The instrument that is the subject of this paper is a small single-manual harpsichord which has many of the usual characteristics of Italian instruments[1].  Interestingly, unlike many other Italian instruments, this harpsichord has case sides of sycamore and not of cypress.  But like many other Italian instruments it was originally intended to be placed inside an external outer case.  The original outer case is, however, now missing and only the inner instrument remains.  The outside of most of the case of the instrument itself is plain, but the insides of the keywell and soundwell are beautifully decorated.  Indeed, those parts of the instrument that are decorated are those that would normally be exposed with the instrument enclosed inside its outer case.  This strongly suggests therefore that those parts not visible with the instrument in its outer case were not decorated, whereas the decorated parts are those which would be visible.  This, in turn, implies that there must have been an outer case to hold and protect the surviving inner instrument.  The present outer case and the inelegant stand[2] for the instrument are, however, modern and were made by Christopher Stevens in 1974.  They will not be included in the description here.

         The instrument is neither signed nor dated.  But the style of the instrument, the materials and methods used in its construction, the mouldings, its decoration and its original compass all suggest that it is by Onofrio Guarracino.  These feature taken by themselves would make a convincing argument that it is by this maker.  However, the most important feature of the instrument and the most convincing argument in its attribution to Guarracino is the use in its design and construction of Guarracino’s unique workshop unit of measurement found to have been used in many of his other instruments.  Further details of the attribution are given below on page 18, and biographical details of Guarracino are given in Appendix 3. 

         The present altered state, although also unsigned and undated, most probably dates from the first half of the eighteenth century.  It seems likely from the characteristics of the workmanship of this alteration that the author of the most recent historical state is either Bartolomeo Cristofori or his pupil Giovanni Ferrini[3].  But the instrument has been the subject of at least two intermediate alterations between the original and the present state.  There now seem to be no remaining clues as to who might have carried out these intermediate modifications.  Each of the different final and intermediate states altered some or all of the compass, number of registers, string scalings, pitch and possibly the stringing materials without, however, in any way changing the basic appearance and decoration of the instrument in a major way.

         Among the many altered features of the instrument is the keyboard (see Figure 15).  Although much of the keyboard is original, the present balance rail is not and so the original compass of the instrument cannot be determined directly from any putative plugged balance pin holes in it.  However, the attribution of the instrument to Guarracino enables a comparison of this instrument with the other harpsichords which are either signed or can be attributed to him.  I will attempt to show here that Guarracino’s standard building practice, his use of his own unique and characteristic workshop unit of measurement near to that normally used in Naples, and numerous clues still left on the instrument enable the original compass to be determined with a high degree of certainty.  The process of the attribution and the consequent determination of the original compass, the original string scalings and the original pitch of the instrument will be elaborated below.

 

InscriptionS:

         The only inscription now left on the instrument is a small label pasted to the lower surface of the baseboard near the tail.  This says:  “ON LOAN FROM / G Donaldson / December 1890”.  The keys and jacks are numbered in pen and ink but, as these are not original, there is therefore no signature nor date of the original maker on them, nor is there a signature of the builder who last re-made the instrument (see the section in the biography of Onofrio Guarracino at the end of this paper for a possible reason that the signature, and possibly the date, have been lost from this instrument).  There is neither a signature nor date anywhere else on the instrument.

 

The soundboard rosette:

         There are the remains of an elaborately-made soundboard rosette in the normal position in the soundboard (see Figure 2 in the colour section on page xxx and the drawing shown in Figure 6).  The inside diameter of the hole in the wood of the soundboard is 95mm.  Remains of the rosette are now found only on the upper surface of the soundboard.  But the rosette was clearly originally in two parts: the existing upper part was glued to the top, and a second part must have been glued underneath the soundboard.  The remaining upper portion is in the style found on numerous other Guarracino harpsichords, virginals and bentside spinets.  What remains of the rosette consists of two layers of parchment with indications of a third top layer.  Originally the lower layer of parchment that now appears to be white was gilded all over and the remains of this gilding can be seen in various places, but especially where bits, now lost, of the layer immediately above it have protected the bottom layer leaving the undamaged gilded surface of the lower layer exposed.  The second layer appears in some places to be black.  However the interior section of this layer, which was originally covered over with a third top layer probably of wood, now appears pink or red.  This part, which was not exposed to light over a long period, may therefore retain the original colour of the part that now, through discolouration and ageing, appears black.  In places there is a thin additional third layer of parchment glued over the middle part of this central red band.  This is almost certainly the layer of parchment which would originally have been glued to the lower surface of the wooden ring that was glued over the whole of this central section of the top level of the rosette.  The use of one or more thin layers of wood is almost universal in the soundboard rosettes used in the instruments of Guarracino.  Rosettes made wholly or partly of wooden layers must have a layer of parchment underneath each layer of wood in order to support the wood while it is being cut to form the rosette pattern and prevent it from splitting and cracking during cutting.  The wood, glued to the lower supporting layer of parchment, is scraped very thin and therefore lacks the necessary strength and integrity to be supported on its own without this additional lower layer of parchment glued to it.  However, when seen normally the layer of wood, with the parchment underneath it, seems to be much thicker and made solely of wood.  Here, although the wood has now disappeared completely, parts of the layer of parchment originally used underneath the wood still remain glued to the rosette in places.  The rosette was therefore highly coloured compared to many Italian soundboard rosettes suggesting that the missing top layer of thin wood might also have been dyed or coloured.  However, it seems now impossible to tell if the top ring of wood was also originally coloured to contrast with the red layer and gilt layer below it.

         Originally there would have been an inverted ‘wedding cake’ type of rosette below the soundboard similar in style to the numerous original soundboard rosettes found on other Guarracino harpsichords and virginals.  It is likely that the lower portion would, in this instrument, also have made use of brightly-coloured layers of parchment similar to those in the upper part of the rosette.  Most, but not all, of the other instruments signed by or attributed to Guarracino have survived with this portion of soundboard rosette (see my website http://www.claviantica.com/Characteristics.htm (section 9) and particularly ..\..\Mozart_harpsichord_files\Mozart_harps_rosette_files\Mozart_harps_rosette_historical_models.html).  There are faint signs of the glue that originally held the lower part of the rosette to the underside of the soundboard when this is viewed through a mirror, but nothing more now remains of the lower portion of the rosette.

 

Figure 2a.jpg

 

Figure 2b.jpg

Figure 2 - Soundboard rosette, scale 1:1 (above) and a detail of the cutting and colours of the soundboard rosette (below)[4].


The Decoration: 

         The outside of the case appears untreated and, unlike the usual treatment of the cypress cases in normal Neapolitan instruments, was probably not even oiled.  Although the lower case moulding is decorated with paintings at the front of the case, it is unpainted along the cheek, bentside, tail and spine (see below).  The inner, outer and cap mouldings along the upper edge of the cheek, bentside, tail, spine and nameboard are gilded as are the soundboard mouldings, the jackrail mouldings, the moulded parts on the front of the nameboard and keyboard, and the carved female figures at either side of the keywell (see Figure 5 in the colour section on page xxx).

         The flat surfaces in the soundwell above the soundboard and wrestplank and those on the nameboard, jackrail, jackrail support as well as the lower case-front moulding are painted a dark bluish-green.  Decorating these green surfaces are vine- and leaf-work with birds, insects, rabbits and flowers in gold (probably bronze powder, although this has not been confirmed analytically) over the green ground.  Detailing of the birds and animals is done in ink over the gold/bronze powder (see Figure 3 and Figure 4 in the colour section on page xxx).

 

Figure 3.jpg

Figure 3 - The gold decoration on the rear of the nameboard.

 

Figure 4.jpg

Figure 4 - A detail of the gold decoration on the spine part of the soundwell.  The inked fine detailing of the figures can be clearly seen.

 

         The soundboard rosette, which is also to be considered as part of the decoration, has already been discussed in detail previously. 

Figure 5.jpg

Figure 5 - The carved figure of a siren at the treble end of the keywell, possibly carved by Aniello or Michele Perrone.

 

         Naples is considered to be the land of the sirens.  The carved figures at the sides of the keywell are certainly intended to represent sirens[5] and may even symbolize Partenope (Greek for ‘maiden-face’) who is sometimes called ‘The Queen of Naples’.  Partenope is normally depicted as a double-tailed siren or mermaid and is one of Naples’ most important icons.  According to Greek myth the siren Parthenope fell in love with Ulysses but he blatantly ignored her attentions.  She swam ashore in the Bay of Naples and died there of a broken heart.  Because of this Naples, in an attempt to live down this sad legend, has always been known as a place of safe and friendly harbour to anyone coming ashore there, and especially to anyone who may have fallen victim to misfortune and hard times and might therefore be in need of nurture and sustenance.  The figure of Partenope has thus become a symbol of Naples and is, after Vesuvius, one of the most important images and icons for Naples and its harbour.  The image of Partenope is visible on practically every fountain, doorway, gate and monument in Naples and so it is no surprise to see this image repeated here.  It is therefore also a clear link between this harpsichord and its origin in Naples.

         It is relatively rare that the decoration of a harpsichord can be attributed to a specific artist or workman.  This kind of decoration is almost never signed by the artist and therefore the decorator normally remains anonymous.  However it is known with certainty that Aniello and Michele Perrone were Guarracino’s brothers-in-law and it is also known that they worked as wood carvers (see Appendix 3).  It is therefore highly likely that they were best placed to do this sort of work for Guarracino and that there is at least a high probability that they are indeed the artists who did at least this part of the decoration.

         The carved keywell scrolls, the gilt decorations in the keywell and soundwell, and the moulding on the outside of the top edge of the case are all very similar to those on an instrument which has been attributed to Vincentius Pratensis in Leipzig[6].  However, the nameboard which carries the signature of the Leipzig harpsichord does not belong to the instrument and has mouldings found nowhere else on the instrument.  Indeed the moulding along the top of the nameboard does not at all match the corresponding mouldings on the rest of the case.  Therefore because of the similarity of the mouldings of the case of the instrument, the similarity of the decoration, the similarity of the top part of the rosette, the carved keywell scrolls and the methods and materials of its construction the same arguments for the attribution of RCM 175 could be used to suggest that Leipzig harpsichord Cat. No. 69 is also by Guarracino[7].

 

Figure 6.jpg

Figure 6  - Drawing of the plan view (above) and treble section (below) made by the author in 1974.  The original position of the bridge was measured from Guarracino's scribed marks on the baseboard and is indicated above the present bridge and the position of the original nut (with the pronounced ‘nick’) is shown to the near side of the present (only slightly curved) nut.

 

The Case Structure - A Novel Construction method:

         The case sides and the bottom outside moulding are of quartered sycamore or maple.  There is a great deal of confusion about the way these two words are used in that they are often used interchangeably and indiscriminately.  The two woods are incredibly difficult to distinguish macroscopically, and this has contributed to a great deal of the confusion in the use of the two non-technical terms.  Technically sycamore is Acer pseudoplatanus (in Italian acero montano).  The leaves of the sycamore have five distinct points and are strongly ribbed.  The two ‘wings’ of the seeds lie almost parallel to one another and the bark in older specimens tends to be scaly and peels away at the bottom of the trunk.  Maple – Acero platanoides in Italian is called acero riccio – has a smooth bark for most of the trunk of the tree, it has leaves that are much more pointed and less ribbed and the leaves have two small extra points near the stem of the leaf.  The two trees grow throughout most of Europe and are often found together with beech.  Violin makers who usually use sycamore call it variously ‘curly maple’, ‘figured maple’, etc. but in fact it is really technically sycamore.  To be fair, in Italy at least, violin makers specify acero montano which is technically sycamore, for the wood of their instruments.

         Since a microscopic examination of the woods of the instrument has never been made it is impossible to distinguish these two in order to say with certainty that the wood is either maple or sycamore.  Underneath the bottom moulding and hidden by it, the sycamore case side is in contact with a strip of chestnut of the same thickness as the case sides and, at first, these seem to be joined together along the length of the two pieces (see Figure 8, Figure 10 and Figure 14).  Both the lower outside moulding and this chestnut strip project about 6mm below the level of the baseboard.  Normally Italian instruments sit on their baseboard both when in their outer cases and when placed without their cases on their stand.  The type of construction found here with a raised baseboard means that the instrument sits, not on the baseboard surface, but on the combined lower edges of the outside moulding and the chestnut strip.  This projection can be seen at the bottom of the section view at the bottom of Figure 8 and in Figure 14.  The fact that the joins in the chestnut strip, when seen from underneath the harpsichord, are lap-jointed and not mitred like the case-side joins above the strip, and the fact that the chestnut strip was clearly not butt-glued to the rest of the top section of the case side, both indicate that the chestnut strip must have been glued in place first and that the top part of the case with its accurate mitre joins was glued in place in a separate later glueing operation (see Figure 7 below).

 

 

Figure 7.wmf

Figure 7 – The bentside-tail join greatly enlarged showing the lap joints in the chestnut strip (left) and the mitre joints in the case sides (right) immediately above the chestnut strip.

 

Figure 8 - Spine section of the case.  The chestnut strip is shaded-in at the lower left.  Scale 1:2.

 

         The existence of these lower chestnut strips and the way the joins between them are made suggest a novel and, so far as I know, an as yet unpublished construction method used by Guarracino and also some of the other contemporary Neapolitan makers.  The method can be outlined as follows:

1.     After the wrestplank support blocks, the lower belly rail and internal framing were glued to the baseboard and the knees were dovetailed into the top surface of the baseboard along the outer edges, the wrestplank support blocks, the wrestplank, upper belly rail and the liners were all glued into position.  This part of the construction procedure all follows normal Italian practice.

2.     The lower strips of chestnut were then glued and nailed in place along the outer edges of the baseboard with the lower edge of the chestnut strip projecting about 6mm below the baseboard.  It seems likely, although this cannot now be confirmed, that the bentside chestnut strip was probably pre-bent using heat before it was glued in place.  The joins at the corners of the baseboard were made with simple lap joints at the ends of the chestnut strips.  The nails were positioned so that they were driven into the edge of the baseboard in the usual way.  These nails provided the temporary clamping of the chestnut strips while the hot animal glue dried and solidified.  The chestnut strip was both thin and narrow and was easily bendable to the bentside shape, and would have presented no difficulty while being glued to the bentside edge of the baseboard.  The heads of the nails were driven into the wood of the chestnut strip so that they did not project up above the surface of the chestnut strips.

3.     The outer lower mouldings, which are higher than the chestnut strips, were then glued over the full height of the chestnut strips.  This was probably done using a large number of small clamps which were of the correct size and configuration to carry out this specific glueing job (see Figure 9 below).  These clamps held the lower moulding against the chestnut strips using the lower projection of the chestnut strips for clamping along the lower edge, and the readily-accessible upper edge at the top could be used for clamping att he top edge.  The lower outer moulding is relatively quite thick and therefore quite stiff, and even with a large number of clamps it would be difficult to bend it to the shape of the bentside where the bentside was sharply curved in the treble without heating and pre-bending[8]. Therefore at least the bentside lower outer moulding was probably also slightly pre-bent using heat. 

4.     Because the lower outer mouldings were higher than the chestnut strip, when they were glued to the chestnut strips a gap was left at the top of the chestnut strips between the top of the lower outer mouldings and the knees and the other internal framing pieces (this gap can be seen clearly above the shaded strip in Figure 8).  The main upper portion of the case side could then simply be slipped into this gap behind the lower moulding and wedged between this and the knees and internal framing placed along the edges of the baseboard.  This would provide a kind of self-clamping mechanism for the whole of the lower edges of the case sides during gluing and this meant that no clamping was required along the lower edge of the case sides during the process of glueing them in place

5.     Glue was applied in the gap above the chestnut strip and to the outside edge of the liners to glue the case sides in place.  The only clamping that was then necessary would be at the top using the soundboard liners themselves when glueing the case sides to the liners.  The lower edges of the case side would be self-clamped between the top edge of the lower moulding and the knees and would require no clamping along their lower edge at all!  This would enable the top part of the case sides to be glued rapidly and accurately before the hot gelatine in the animal glue used during the operation began to start to cool down, set and solidify.

6.     Once the case sides were mitred and glued in place as just described, the top inner and outer mouldings used to strengthen and reinforce the thin case sides could then be applied and glued in place easily to the top edges of the case.  Both the inner and outer top mouldings are light and easy to bend.  The bending of the wide top cap moulding along the bentside would present problems in glueing and clamping, but this problem may have been avoided by cutting the cap moulding down the middle of the strongly curved portion of the bentside to facilitate the bending process.  However, I have never been able to confirm this by direct observation on any instrument known to have been made in Naples.

 

Figure 9 – Theory put into practice – a clamping extravaganza during the construction of an instrument based on the construction method outlined above!  Here the chestnut strips have already been glued in place along the lower outside edge of the baseboard.  The lower outside case mouldings are being glued to the chestnut strip along the lower left (spine side) with specially-made clamps, and along the front of the case with simple spring clamps.  The bentside is being glued in place by slipping it into the gap above the bentside chestnut strip and then clamping it only to the bentside liner.  The bentside chestnut strip is clearly visible just above the level of the baseboard.

 

         Once the bentside was glued in place the rest of the case sides, all of which are straight, were then relatively easy to clamp and glue in place.  This ingenious clamping system is somewhat non intuitive but works well in practice[9].  The discovery of this clamping system used by Guarracino and some of his Neapolitan contemporaries, gives a rare insight into the daily workshop practices of one of the most important and active centres of harpsichord building in the second half of the seventeenth century.

 

Case Structure - Description:

         The upper mouldings on the case sides are of a gilt and painted fruitwood (pear?).  The baseboard, the internal buttresses, baseboard stiffeners, keybed and the lower part of the bellyrail, are of a coniferous wood, which appears to be fir and not spruce or pine.  The bentside liner, cheek liner and knees are of beech.  The spine liner and tail liner are both of a coniferous softwood which is also probably fir.  The top part of the bellyrail and the baseboard frame below the keyframe in the keywell are of poplar.

         The outside case dimensions, excluding the thickness of the upper and lower case mouldings, are:

 

Element measured Length in mm Thickness in mm

Spine

1834 4.2
Cheek 484 4.4
Bentside - - 4.3
Tail 404 4.2
Baseboard - - 13.8
Case height 213[10] - -

Table 1 – The external case dimension, not including the mouldings

         The internal construction (see the preceding section) is typically Italianate with baseboard stiffners, knees between the baseboard and case sides which are dovetailed into the baseboard, and flying buttresses lodged between the bentside liner and every second bentside knee (see Figure 6).  There is also a large knee brace between the bellyrail and the baseboard positioned slightly towards the spine side of the centre of the bellyrail.  This type of construction is typical of all of the other harpsichords known to be by Guarracino or which can be attributed to him (see Table 4).

         The present wrestplank is of walnut, 46mm thick, but it is not original.  The original wrestplank was removed to allow access to the gap and to make space for an additional register.  The bellyrail is composed of two parts, the upper part of which, along with a small strip of the front of the soundboard, has been partially cut away to make room for the additional register.

         The soundboard is of quartered fir.  There is only one fir soundbar placed diagonally under the soundboard in a position roughly parallel to the central section of the bridge as in Flemish practice.  The bridge is of sycamore or maple, and the nut, which belongs to the most recent state, is of cypress (see Figure 28).  A small piece carrying the bridge pins for the top note has been added to the original bridge[11] (see Figure 6) and this and the rest of the treble section of the bridge has been moved towards the gap shortening the scalings for the top octave of notes.  The pinning of the bridge and nut, and the hitchpins reflect the various states of the instrument's history.  The larger set of hitchpins placed nearer to the case sides have a diameter of 1.6mm and probably belong to the original state.  These alternate with later hitchpins of diameter 1.5mm.  The bridge pins also have alternate diameters of 1.4mm and 1.18mm and correspond to the original and later states respectively.  The nut pins are all of diameter 1.18mm and belong to the most recent state.  The tuning pins of both rows have a diameter of 4.4mm and are probably all replacements but of the historical period.  The soundboard mouldings, although painted over, appear to be of a fruitwood (probably pear).

 

Figure 10.jpg

Figure 10 - The interior photographed by lowering a small camera into the soundboard rosette hole.  This shows the spine side with the lower strip of chestnut (darker wood with a coarse grain) hidden from the outside because of the wide lower moulding, but clearly visible on the inside.

 

Figure 11.jpg

Figure 11 - The interior viewed towards the tail.  The spine side is on the left and the bentside is on the right.

  

         Like many of the other instruments attributed by me to Guarracino the plan of the instrument is marked with scribed lines on the baseboard.  Most of these scribed lines are hidden inside the instrument or below the keyframe support rails inside the keywell.  However the scribed lines used to mark out the nut position are partially visible on the top of the baseboard in the keywell.  There are also punched marks between the two lines scribed on the baseboard which give the position of the edges of the nut for the notes tenor c and for middle c1.  The latter also indicates the point of the ‘nick’ in the bridge where the angle of the bridge changes abruptly[12] (see Figure 12 below).  In those instruments with the plan marked on the baseboard Guarracino’s normal practice was to place this ‘nick’ in the bridge at the note middle c1.  The distance of this mark from the spine baseboard edge is 343mm.

Figure 12.jpg

Figure 12 - The punched mark (indicated with an arrow) on the top surface of the baseboard in the keywell looking toward the tail.  This punch mark gives the position of the nut pin for the note middle c1

Also clearly visible is the change in the direction of the nut where it is bent abruptly between the bass/tenor part of the compass (on the left) and the alto/treble part of the compass (on the right).

 

The original dimensions of the baseboard without the case sides:

         I have been able to show that many of the harpsichord makers working in the historical period both north and south of the Alps used their local unit of measurement to design and build their instruments[13].  In order to determine the unit of measurement used to design and construct this instrument the baseboard and case height were measured.  These are shown below in Table 2 below:

 

Length

1817

 

Width

770

 

Cheek

478

 

Tail

389 originally[14]

387 now

Component of tail perpendicular to spine

231 originally

228 now

Component of tail parallel to spine

313

 

Tail angle

36½º

 

Maximum case height

213½

 

Soundboard to the top of the case

61

 

Table 2 – Some case and baseboard measurements without the case sides

         These measurements are given in both millimetres and in units of the oncia used by Guarracino in Figure 13 below

 

 

    the Scientific determination of the unit of measurement used to design and construct this instrument:

         The city or region of construction of an anonymous instrument can be determined by analysing the size of the unit of measurement used in its construction.  Until the time of the Napoleonic invasions of the Italian peninsula each of the major cities in Italy used its own unit of measurement and the size of these units varied from place to place.  The braccio, piede, canna, palmo, etc. and their subdivision into the oncia or pollice were therefore characteristic of each of the centres in which instruments and virtually all other manufactured items were built.  Therefore if the unit of measurement used in the design and construction of an instrument can be determined, this can be used in turn to establish the centre of its origin.

         The Italian makers began the construction of their instruments with the baseboard.  The various braces, knees, wrestplank supports, belly rail, etc. were glued to the baseboard and eventually the case side were applied around the internal structure to the outside edges of the baseboard and liners.  It is therefore the baseboard that the makers laid out first, often even drawing the plan of the instrument onto the baseboard before the structural part of the construction began.  The baseboard without the case sides is therefore the first point of departure when trying to determine the unit of measurement used in the construction of an instrument. 

         The procedure of determining the unit of measurement used to construct this harpsichord begins most easily with the measurement of the angle of the tail.  The tangent of this angle is tan 36½º = 0.74 @ 0.7414 = .  This suggests that, to within an error of only 0.6%, the two sides of the triangle that form the tail angle are 10¾ once and 14½ once.  Although these are rather ‘untidy’ numbers, they are the only ones that are consistent with one another and that yield estimates for the lengths in once of all of the other measurements of the instrument.[15]  The measured lengths of the two perpendicular components of the tail in millimetres divided in turn by these numbers give an approximate estimate of the size of the oncia that can then be applied to the other measurements of the baseboard, wrestplank, the original string scalings, and all of the other parts and design features of the instrument.  A summary of the measurements of the baseboard and case height in once is given in the following table:

 

                                                                         Measurement     Local                                                                           Length of

                                                                                in mm         unit                                                                                   oncia

                 Component of tail parallel the spine:           313    =   14½ once                                                           Þ       21.586

  Component of tail perpendicular to the spine:           231    =   10¾ once                                                            Þ       21.499

                                               Spine (long side):         1817    =   84 once = 8 palmi                                               Þ       21.631

                                              Baseboard width:           770    =   35½ once                                                            Þ       21.690

                            Baseboard cheek (short side):           478    =   22 once                                                              Þ       21.727

                               Maximum case side height:        213½    =   10 once                                                              Þ        21.35   .

                                                             Total      3822½    =   176¾ once                                          Þ     Average: 21.627mm

Table 3 - Calculation of the unit of measurement used by Guarracino in the design of the case.

 

 

         These measurements are shown in the diagram of Figure 13 where the actual measurements of the baseboard in millimetres are shown on the left, and the measurements in Guarracino’s unit of measurement are shown on the right.  A look at a table of the units of measurement used in the various centres in Italy during the historical period shows that in Naples the palmo, divided into 12 units, had a length of 262.01mm[16].  Hence the oncia had a length of:

The difference between this and the oncia found here is only of the order of 1% indicating a reasonably good agreement.  However, Angelo Martini[17], usually one of the most reliable sources on metrology, gives a value of 263.67mm for the palmo making the oncia = 21.973mm.  This value of the oncia differs by about 1.6% from the value found here, and this is an amount which is significantly different from the unit found here.  Indeed when I first calculated the unit of measurement used by Guarracino I was tempted to conclude that he was probably from Palermo (Sicily) rather than Naples because the oncia in Palermo and used throughout Sicily had a value of 21.51mm – a value much closer to that used by Guarracino than Martini’s value.  The overwhelming archival evidence of Guarracino’s activity in Naples during the second half of the seventeenth century meant that I had clearly come to the incorrect conclusion based solely on his use of the ‘incorrect’ unit of measurement!  The consistency in the results of the later calculations that I made of Guarracino’s unit of measurements found for the considerable number of other instruments clearly meant only one thing: Guarracino used rulers in his workshop divided into units that were notably smaller than the various ‘textbook’ values that I had otherwise found.  As an example of the difference involved, the length of this harpsichord would have been some 29mm longer if Guarracino had used rulers divided using the palmo and oncia given by Martini instead of the value found here.  Obviously this is a considerable difference and one that would show up clearly with the measurements taken here.

 

Figure 13.wmf

Figure 13 - Baseboard measurements without the case sides.  The measured angles and lengths are on the left  and the nominal angles and measurements in units of  Guarracino’s workshop oncia are on the right.

 

         I have, however, measured quite a large number of other instruments made by Onofrio Guarracino (see Table 4 below).  From these measurements it seems clear that Guarracino had his own workshop unit which is somewhat different from the ‘textbook’ value of the Neapolitan oncia given above.  I have been able to calculate an average value of the oncia used by Guarracino in his workshop as 21.622mm (see Table 4 below and Table 9 ) so that the palmo as used by Guarracino would be:

1 palmo = 21.622mm/once x 12 once/palmo = 259.46mm

This value of the oncia and palmo are only 0.07% different from that found here!!  No other maker is known who used this particular value of the Neapolitan oncia so that this is one of the best indications that this instrument is indeed by Guarracino!

 

 

 

 

oncia

Error[18]

Error

 

Instrument

mm

%

mm

Method

1663 Guarracino virginal, Tagliavini, Bologna

21.61

0.07

0.02

Case measurements

1663 Guarracino virginal, Tagliavini, Bologna

21.61

0.60

0.13

Lateral string spacing

1668 Guarracino spinet, Museo Naz., Rome

21.61

0.09

0.02

Case measurements

1668 Guarracino spinet, Museo Naz., Rome

21.58

0.60

0.13

Lateral string spacing

1677 Guarracino virginal, Museo Naz., Rome

21.66

0.07

0.02

Case measurements

1689 Guarracino virginal, Roger Mirrey Collection, Edinburgh

21.62

0.07

0.02

Case measurements

1692 Guarracino virginal, Museo Naz., Rome

21.61

0.07

0.02

Case measurements

1692 Guarracino virginal, Museo Naz., Rome

21.58

0.23

0.05

Lateral string spacing

         

1651 Guarracino harpsichord, Private,  Rome

21.61

0.09

0.02

Case measurements

1651 Guarracino harpsichord, Private,  Rome

21.61

0.07

0.02

Lateral string spacing

 n.d. ‘Guarracino’ harpsichord, Gemeentemuseum

21.65

0.18

0.04

Case measurements

n.d. ‘Guarracino’ harpsichord, RCM 175, London

21.63

0.09

0.02

Case measurements

n.d. ‘Guarracino’ harpsichord, RCM 175, London

21.64

0.05

0.01

Lateral jackslot spacing

n.d. ‘Guarracino’ harpsichord, RCM 175, London

21.63

0.09

0.02

Lateral string spacing

n.d. ‘Guarracino’ harpsichord, Giulini, Milan

21.61

0.05

0.01

Case measurements

n.d. ‘Guarracino’ harpsichord, Giulini, Milan

21.63

0.11

0.02

Lateral string spacing

n.d. ‘Guarracino’ harpsichord, MMA, New York

21.65

0.14

0.03

Case measurements

n.d. ‘Guarracino’ harpsichord, BMFA, Boston

21.60

0.14

0.03

Case measurements

n.d. ‘Guarracino’ harpsichord Beurmann, Cat. Nº 8[19]

21.64

0.14

0.03

Case measurements

Weighted average:    

21.622

0.002

0.01

 

Table 4 - Calculation of Guarracino’s averaged workshop unit of measurement.

All of the measurements, calculations and statistical analyses for all of the instruments in this table are by the author.

 

The Keyboard:

         The compass of the present keyboard is C,D to d3 with 50 notes (see Figure 15 below).  The three-octave span is now 506mm, and the total width of the keyboard is 721mm.  The natural touchplates are of boxwood, the heads of the natural touchplates are 33mm long; the sharps are 69mm long and are of a fruitwood topped with a slip of ebony 1-3mm thick.  The naturals have boxwood arcades.  The present keylevers are of chestnut, 320mm long in the bass and 2mm longer in the treble.  The distance from the balance point to the front of the naturals is 141mm throughout the compass.  The balance line of the accidentals is 13.8mm behind the naturals.  The natural balance of the keylevers for the naturals is about 10-15mm behind their respective balance pin holes[31]

         The keyboard has a keyframe with side edges that project out sideways from the top and bottom keylevers, and these projections enable the keyboard to slide in and out of the instrument in slots in the keyblocks along the cheek and spine.  Indeed this is one of the characteristics of the instrument that points to its Neapolitan origins (see point 7 on page 21).  There is now nothing else to retain the keyboard in position in the instrument.  Originally, however, or at least before the second set of jacks was added, the keyboard was probably provided with the usual diapason rack.  If it was attempted to remove the keyboard without removing the jacks, the diapason rack would then have interfered with the jacks, and this would therefore normally have prevented the keyboard from being withdrawn.  The addition of the second register entailed removing most of the upper belly rail which would normally have overhung the rack.  Removal of part of this upper belly rail meant that the rack had to be abandoned and the keylevers were therefore guided by wooden pegs inserted in the back rail of the keyframe between the keylevers and just in towards the player side of the first row of jacks.  These pegs are covered with leather, and semicircular cut-outs in the sides of the tails of the keylevers match the size of the leather-covered pegs.  This guiding system (seen clearly at the top of the photograph in Figure 15 below) and the moulding of the natural arcades are both the same as that sometimes used by Cristofori and his pupil Ferrini.

 

Figure 15.jpg

Figure 15 – A photograph of the keyboard taken by the author during the restoration of the instrument in 1974.

 

         As a result of the numerous alterations to which the instrument has been subjected over time, the keyframe is of a very complicated structure.  Indeed the alterations to the keyboard are among the many features which show that the instrument has gone through a number of intermediate states between the original and the present.  Originally the keyframe was probably made entirely of a coniferous wood, probably fir.  But there are now additions of poplar and chestnut whose presence and dimensions can only be explained by postulating that the instrument went through a series of intermediate states.  These will be outlined later, but a detailed analysis of them is outwith the scope of this report[32].  The present balance rail is not original and is of poplar.  Marks on the keyframe show that it is now positioned about 20-21mm further towards the tails of the keylevers than the original balance rail.  Because, in the original and the two subsequent states there was only one register of jacks, the balance rail could have been placed further forward than in the altered state with two registers of jacks which would have been uncomfortable to play with the balance rail in its original position[33].  

         Because the original balance rail is now missing there is now no definitive indication of the original compass.  The original compass as determined by John Barnes in his 1974 restoration report[34] was said there to be C/E to d3,e3 with only 48 notes.  However, there is a number of indications that this was not actually the original compass.  Since the original balance rail is missing, the original compass and number of notes will therefore have to be deduced from other features of Guarracino’s normal building practice including the position of the nut markings on the baseboard for the notes tenor c and middle c1.

 

The Jacks, Registers and the lateral bridge-pin and jackslot spacings:

         At the time of the 1974 restoration a number of different styles of jack was found in the instrument.  One group appeared to date from the original state and was of fruitwood (probably pear) with two damper slots and tongues of quartered beech.  Indeed most of the signed instruments by Guarracino which retain their original jacks are made in this way.  The position of the sides and bottom of the damper slots as well as that of the axle pin were scribed on the sides of the jacks.  Another group was of walnut with one damper slot and tongues of beech (mostly but not always quartered).  Various other jacks of poor-quality workmanship were also found.  The restoration of 1974 reinstated only the fruitwood jacks and modern copies based on them.

         The present jackslides are of a fairly complicated construction.  Evidence revealed during the 1974 restoration shows that the present slides are made up using the original single boxslide register.  This was sawn apart into strips to form the basis for the two upper and two lower guides (see Figure 16 below).  The two upper guides are separate and consist of the central original boxslide sections (double-shaded in Figure 16) with added outer mouldings that cover part of the soundboard and wrestplank.  The lower guides are a single unit consisting of six parts:  the central original boxslide sections each with two outer sections on either side of the original parts of the boxslide.  The upper guides are relieved on their lower surface with a knife so as to open outwards in order that the jacks don't bind in their slots should there be a misalignment of the slides.  For a similar reason, and to facilitate the insertion of the jacks, the slots in the lower guides are relieved on their upper surface.

         Sections of the upper registers in this harpsichord are shown in the diagram below (right-hand drawing) and, for comparison, the section of the registers on the Anonymous Neapolitan harpsichord in the Russell Collection (on the left) which is also thought to have been modified by either Cristofori or Ferrini[35].

Russell Collection                                             RCM 175

Figure 16.wmf

Figure 16 - Sections of the upper registers in the Russell Collection anonymous Neapolitan harpsichord (left) and RCM 175 (right).  The heavily-shaded central section represents the original boxslide from which the new registers were made in both cases.  Scale 1:1.

 

         The position of the present long 8' bridge-pin was measured from the spine side of the instrument.  And since the present composite registers are, in fact, made from the original Guarracino boxslide register sliced into layers, the spacing of the jackslots should in some way reflect the use of Guarracino’s workshop unit of measurement which is close to the textbook Neapolitan unit found in books on metrology.  In order to try to discover how Guarracino designed the spacing of the present long 8' bridge pins and of his jackslots in the original boxslide register, the lateral spacing of thesewas measured.  This was done simply by measuring the distance from the spine of the instrument near the gap to the treble side of each register slot[36].  These measurements are given in Table 5 below.

 

  Note 8' br.   Reg. slot     Note 8' br.   Reg. slot
Note number position   position   Note number position   position
d3 49 713.7   721.6   cT1 24 372   378.1
cT3 48 698.7   707.4   c1 23 359   363.6
c3 47 685.7   694.1   b 22 344.5   350
b2 46 671.7   680   bI 21 331   336
bI2 45 658.7   666   a 20 317.5   322.7
a2 44 644.7   651.9   gT 19 305   309.3
gT2 43 632.2   637.9   g 18 290   295
g2 42 618.2   624   fT 17 276.5   281.2
fT2 41 604.7   610.6   f 16 263   267.7
f2 40 591.7   598.2   e 15 248   253.3
e2 39 576   583.9   eI 14 235.5   240
eI2 38 564   570.4   d 13 221.5   226.6
d2 37 552   556.7   cT 12 207.5   211.9
cT2 36 538.5   542.3   c 11 193.5   198
c2 35 524.5   528.6   B 10 179   187.8
b1 34 511.5   515.1   BI 9 166.5   170.4
bI1 33 496.5   501.2   A 8 152   156.6
a1 32 483.5   487.5   GT 7 138.5   143
gT1 31 469   474   G 6 125   129.3
g1 30 455   459.8   FT 5 111.5   115.8
fT1 29 442   446.2   F 4 98   101.7
f1 28 427.5   432.4   E 3 83.5   87.8
e1 27 414   419.2   EI 2 70   73.9
eI1 26 400   405.5   D 1 56.5   59
d1 25 385   391.4   C 0 42.5   45.2

Table 5 - Distance from the spine to the long 8' bridge pins and to the treble edge of each of the jackslots.  Note that CT is missing in the bass and is not plotted in either Figure 17 or Figure 27 below.

 

         These measurements were plotted in Figure 17 below.  Clearly the points all lie on a straight line which is simply an indication of the fact that the register slots are quite accurately equally spaced with the same regular distance between each successive jackslot.  The points on the straight line should therefore be represented by a straight-line equation of the form:

y = mx + b

Here y is the perpendicular distance of each pin or jackslot from the spine.  If the first note has a value x = 0, then each note above the first has a value x of one whole integer higher than the previous note.  In this case there are 50 notes so that the value of x runs from 0 to 49.  The value b is the distance of the first note (with a value of x=0) from the spine.  The value of m is called the slope of the line.  In this case the slope m is just the average value of the regular spacing of the jackslots.

 

Figure 17.wmf

Figure 17 - Graph of the positions of the left-hand edges of the register slots measured from the spine.

 

 

         In order to determine the value of m the numbers in Table 5 above were therefore subjected to analysis to work out the average value of the regular spacing of the jackslots.  Calculation using a regression analysis by the method of least squares gives a correlation coefficient for this data of r = 0.9999936.  The closer this value is to 1.000, the greater the confidence is that the data are represented graphically by a straight line.  Because this value is, indeed, very close to 1.0000, the indication of the analysis is that the there is a very good fit of the measured data to a straight line.  This is confirmed by the small deviation of the points on the graph shown in Figure 17 from the indicated straight line.  The calculated y intercept is b = 33.12mm gives the distance of the first jackslot from the spine, and the slope of the line found in the regression analysis is m = 13.768mm/jackslot.  The standard deviation or error in this value is only 0.007mm which, in itself, is also a good indication that the jackslots are very regularly spaced.  This, in turn, is an indication of the accurate and precise working methods of Guarracino[37].

         Therefore the slope = 13.768 = 0.6368 @ 0.6364=  if the Neapolitan oncia as used by Guarracino already determined above using the case dimensions (see Table 4) is used as the basis of the measurement system used to design the register and jackslot spacing.  This therefore appears to be a spacing of very close to 11 jackslots in 7 once (the difference between the two is only 0.08%[38]). 

         These numbers can, in turn, be used to determine another value for Guarracino’s workshop unit of measurement.  If it is assumed that Guarracino spaced the jackslots with a spacing of exactly 11 jackslots in 7 once, then the averaged spacing determined from the regression analysis in millimetres is equal to this spacing so that:

13.7675 =

Re-arranging the terms in this equations gives the value of Guarracino’s workshop oncia:

 = 21.631mm

This compares with the value given determined from the analysis of the case measurements on page 15 of 21.637mm.  The difference between these two determinations is only 0.03%, indicating both an extremely small error and the fact that both the case and the register spacings were designed and built by Guarracino using his usual workshop unit of measurement.  It is also an indication of the extremely precise working methods of Guarracino.

         The numbers 11 jackslots in 7 once might seem at first like very strange numbers for Guarracino to have used to space out the jackslots.  Why didn’t he use, say, 12 jackslots in 7½ once which gives roughly the same spacing?  An octave normally contains 12 notes, so this might seem a much more reasonable spacing to have used.  The reason for this seemingly odd 7/11 spacing is that each octave did not have 12 notes in it as one might at first expect.  The spacing of the jackslots in this instrument is notably atypical compared to all of the other harpsichords of Guarracino simply because the compass of this instrument is unique among Guarracino’s oeuvre.  The compass C/E to c3 has, in the case of this instrument, 50 notes instead of the usual 45 because of the broken short octave in the bass and because of the split accidentals in the treble.  This will be demonstrated and discussed below.  It had fewer notes in its enharmonic compass than, say the Giulini Collection harpsichord[39] with a compass of C to f3 with an extended bass octave (ie. not a bass short-octave compass) and an enharmonic compass with 19 notes to the octave (see the reconstruction of part of the compass of this instrument in Figure 24).  It also has fewer notes than the ‘Guarracino’ harpsichord in Metropolitan Museum of Art in New York (see the re-construction of its compass in Figure 26).  Hence it is bound to have a different jackslot spacing than either the usual chromatic C/E to c3 model of harpsichord or the two enharmonic instruments just mentioned.  Basically Guarracino has, across the C/E to c3 enharmonic compass he has chosen for this instrument, to get the 50 notes of this compass into the space occupied by the 27 naturals of the C/E to c3 compass.  Each octave of naturals in this compass has a spacing which, in turn, is determined by the width of the human hand.  The result is Guarracino’s pragmatic solution to this problem:  he chose simply to put exactly 11 notes into the space occupied by 7 once!

 

The Present Disposition:

         The jackslots for all of the c notes except for the lowest ones are marked in ink with a mark shaped like a “T lying on its side to indicate the plucking direction.  This indicates that the rear register should pluck the longer string on the left and that the front register should pluck the short string on the right.  The order of the registers cannot be interchanged because of the mouldings attached to their front and rear edges and so the position and plucking direction of these two registers is original and unambiguous[40].  These therefore correspond to the classic disposition with the rear row of jacks (principale) plucking towards the left (the long choir of strings) and the front row (secondo) plucking towards the right (the short choir of strings).

 

 

Plucking

 

 

 

direction

Pitch

Details

Back row (secondo):

¬

8'

This register is fixed and cannot be engaged or disengaged.

Front row (principale):

®

8'

Can be moved for tuning by sliding the jackrail back.

 

         Blocks which limit the sideways motion of the upper guides are glued to the upper surface near the end of each register.  These blocks allow for no movement of the rear guide and a movement of 1.5mm of the front register.  There are no stop levers, and the only way of engaging and disengaging the near register is to push against the sides of the blocks with the jackrail pushed back somewhat from its normal position directly above the jacks and registers.  Clearly, because of the difficulty of engaging and disengaging the front register, both registers must have been meant to be engaged during performance, and it was only during tuning that the front register was disengaged while the rear register of strings was tuned.  The front register was then engaged to tune the short choir of strings in unison with the rear principale register of strings.  The indication is therefore that the maker responsible for the last state of the instrument (and presumably his musician client) did not expect or require the dynamic possibility of using either one register or two to give forte and piano.  Neither was there any expectation of any tonal variation resulting from the different plucking points of the two registers.  It would have been a simple matter for the maker re-working the instrument to have provided it with stop levers to provide both a dynamic and a tonal variation from the two registers, but this seems clearly not to have been wanted.

 

The jackrail:

         The jackrail has a core made of sycamore or maple decorated with gilt mouldings probably of fruitwood.  The jackrail supports are unusual compared to those of most Italian harpsichords, but typical of many instruments of the Neapolitan school.  Instead of the usual forks on either side of the inside of the case which support the narrowed ends of the jackrail, there are only two small rectangular bars one above each of the ends of the registers.  These bars are slightly thicker and carved at their near ends (see Figure 18 below) to prevent the jackrail sliding off the bar towards the player, and they also serve to position the jackrail correctly above the registers and jacks.  The ends of the jackrail have recessed horizontal cutouts placed centrally at the ends of the jackrail, and these cutouts slide onto the bars mounted inside the case (see Figure 19 below which shows the same arrangement on the only signed and dated harpsichord by Onofrio Guarracino).  Indeed this feature is one of the characteristics of Neapolitan harpsichords in general and of Guarracino in particular (see item 6 on page 21 of this report listing the characteristics of Guarracino’s harpsichords).

         The jackrail still retains old (if not original) damping cloth on its lower surface.  This is red wool velour, tabby woven and felted.  It is folded in 3 layers which were sewn together and then glued to the jackrail.  At some stage this cloth has been removed and re-attached using bent-over modern wire nails.

Figure 18.jpg

Figure 18 - The jackrail and jackrail support block

 

 

The end of the jackrail typical of the harpsichords of Guarracino and many other Neapolitan buildings is shown in Figure 19 below.  This photo was taken of the jackrail of the only signed and dated Guarracino harpsichord which is in a private collection in Rome[41]

 

Figure 19 - The end of the jackrail showing how it is recessed for the bar-section jackrail support normally found on Guarracino harpsichords

Single-manual Italian harpsichord, Onofrio Guarracino, Naples, 1651

Private collection, Rome

 

 

The String Scalings and Plucking Points:

         As mentioned previously, an examination during the 1974 restoration by John Barnes[42], showed that the compass, disposition, and scalings had been changed a number of times and that the instrument originally had only one register of jacks which plucked towards the left. 

         The present state has two unison choirs of strings and two registers.  As will be discussed on page 52 it seems likely that the present state was carried out in the Florentine workshops of either Bartolomeo Cristofori or Giovanni Ferrini.  The original scalings and plucking points which were estimated in the 1974 report, and the present scalings and plucking points for the long set of strings are shown in Table 6 below:

 

 

Original State*

Present State

Note

String

Length

Plucking

Point

String

Length

Plucking

Point

e3

143

57

d3=111

52

c3

167

64

129

58

f2

239

86

195

74

c2

306

102

255

86

f1

430

123

362

103

c1

550

137

465

115

F

805

139

687

124

C

1104

141

949

130

F

1362

144

1305

138

C/E

1380

144

C=1385

144

*as estimated in the 1974 report.

Table 6 - The scalings estimated in the 1974 report by John Barnes and the present string scalings from State 4, the final state.

 

 

 

         The present scalings are plotted using the lengths from Table 6 and are shown in the graph of Figure 20 below.

Figure 20.wmf

Figure 20 – Graph of the present String Scalings

 

 

         A thin straight line has been drawn on this graph representing Pythagorean scalings based on a c2 string length of 9 Florentine soldi[43] or 256mm.  Such a scaling would be suitable for brass stringing at a pitch about a whole tone above a1 = 415.3Hz.  The notable feature of the scalings of this instrument in its present state is that they are not at all Pythagorean and exhibit octave doubling for only about one octave of notes in the treble.  Indeed, the instrument has Pythagorean scalings in the treble only because the treble section of the bridge has been moved towards the gap.  Otherwise the scalings would not be Pythagorean in the treble either.  Normally the scalings found in Italian instruments are Pythagorean for most of their compass above tenor c, and normally Guarracino uses Pythagorean scalings throughout the whole of the compass above the tail bridge mitre.  These scalings are therefore to be considered unusual and atypical of both Italian practice in general and of Guarracino in particular.

         The original scalings as determined in the 1974 restoration report by John Barnes are those shown in the graph drawn in Figure 21 below and are discussed following it.

Figure 21.wmf

Figure 21 – Graph of the string scalings based on the ‘original’ compass and lengths from John Barnes’ 1974 restoration report.

  

         Many of the harpsichords I have attributed to Guarracino and listed in Table 8 have Pythagorean scalings based on c2 = 14 Neapolitan once.  In order to attempt to make this harpsichord fit into this pattern, I have therefore drawn a thin line in the graph of Figure 21 which represents Pythagorean scalings which are based on c2 = 14 Neapolitan once.  However, the measured lengths clearly do not follow this straight line.  I have also drawn a thin line on this graph that represents scalings that double in length every 14 notes instead of every 12, and that are also based on c2 = 14 Neapolitan once.  This would seem to indicate that, for over two octaves of the treble part of the compass at least, octave doubling must have occurred every 14 notes and not every 12 notes as is normal with a chromatic compass.  Such octave doubling could occur, for example, if the instrument were originally designed with an enharmonic compass that had two split accidentals per octave.  Normally these would have been the notes dT/eI and gT/aI.  It would therefore appear that the original compass was therefore not that deduced by John Barnes, but some kind of enharmonic compass with some split accidentals, and that the most important evidence for this type of compass is the octave doubling every 14 notes instead of every 12.

         In the bass, the compass implied by Barnes is neither Pythagorean nor does it follow the 14-note per octave doubling either.  Something totally different must be happening which results in the irregular shape of the scaling curve there.  If the compass was not C/E to d3,e3 as suggested by Barnes, then what was it and what clues other than the enharmonic treble compass are there to the original bass part of the compass? 

         The most important clue to answering this question is the punch mark between the two scribed lines scratched on the baseboard indicating the position of the ‘nick’ in the nut which, for all other Guarracino harpsichords, marks the position of the note middle c1 (see Figure 12).  This punch mark fixes both the position of the original string for the note c1 and its length.  The second clue is that, in the reconstructed compass deduced by John Barnes, the top and bottom strings are unusually far from the case sides suggesting that there were additional top and bottom strings in the original compass.  Allied to this is the fact, also noted by Barnes, that the original register re-used by the re-builder of the instrument once had 50 notes, like the present compass, and not the 48-note C/E to d3,e3 compass suggested by Barnes and used in the 1974 reconstruction of the original compass.

         In the reconstruction of the instrument that I made in the 1974 drawing[44] I followed John Barnes thinking and also assumed that the original compass was the 48-note C/E to d3,e3 suggested by him.  This would have put the note middle c1 one note below the nick in the bridge and not at the nick.  If the original compass had 50 notes and not 48, then there must have been one additional note below the bottom note indicated in my 1974 reconstruction of the original compass and one additional note at the top.  This would position the string band centrally between the spine and cheek and leave a space of about 2 once = about 44mm between the top and bottom strings and the case sides.  This is a normal spacing for Guarracino[45] and for a number of other Neapolitan and Italian builders.  Indeed, as mentioned above, the 1974 reconstruction of the original state and compass leaves an unreasonably large space between the bottom and top strings and the case sides and a long unused section of bridge on the mitred tail section which could easily (and more reasonably) have carried an additional note in the bass.  The hole for this original bridge pin must now be filled with the large pin for the long string of the present lowest note C[46].  In the treble, the end of the bridge has been cut away by the re-builder when he moved the bridge and lengthened it to carry the additional short top string required by the addition of a second register.  Therefore the evidence of the original bridge pin was removed in the alteration of the instrument to its present state.  John Barnes incorrectly assumed that there never was an original pin located in this, now missing, portion of bridge.

         If there were one additional note and string in the bass more than that indicated on my 1974 drawing then there must have been 22 notes below c1 (at the nick in the nut) not counting c1 itself.  Without going into all of the details of all of the possible compasses that there could have been to meet this requirement, the one that seems most likely is a C/E broken octave with split a D/FT and a split E/GT, with notes running chromatically without any further split notes to c1.  This bass compass has the requisite 22 notes indicated above.  The presence of a broken octave in the bass is consistent with a treble enharmonic compass since it provides roots in meantone tuning to some of the root chords and first inversions of the additional major and minor tonalities available with the extra split accidentals provided by the enharmonic compass available in the treble.

              In a compass with 50 notes 22 of which are below middle c1, there then remain 27 notes above c1 (making a total of 50 notes including c1 itself).  As implied above the two octaves above c1 should have 2 split notes per octave so that octave doubling occurs with 14 notes and not the usual 12.  This would, however, give 28 notes above c1 and would therefore add one additional note to the compass not allowed by the original number of notes indicated by the number of original slots in the registers and by the string band layout.  It therefore seems likely that the top gT2/aI2 was not split since it would be used only melodically and not harmonically at such a high point in the compass (see a similar state of affairs in the compass of the Guarracino harpsichord illustrated in Figure 26[47]).  A putative split gT2/aI2 would therefore need to be used probably only as a gT leading note in the key of A major.  Hence only the gT2 part would be necessary and useful.  The most likely compass is therefore 4 octaves C/E to c3 50 notes with a broken octave in the bass and dT1/eI1 and gT1/aI1 in the third octave, and only dT2/eI2 in the top octave.  A digital re-construction of the original keyboard layout is shown in Figure 22 below.

 

Figure 22.jpg

Figure 22 - A digital re-construction of the layout of the original keyboard.  The original naturals were of boxwood and not of ivory as implied here.

 

              The string lengths corresponding to this scheme, which have been measured from the drawing, are given in Table 7 below.  Here the string lengths of the notes GT and A on either side of the bridge mitre have also been given.  See Table 12 for a complete list of the original string lengths of all of the notes throughout the entire compass.

 

Note

String length

Plucking point

c3

134

51

f2

195

73

c2

264

92

f1

391

117

c1

525

136

f

760

139

c

1038

141

A

1253

142

GT

1285

142

F

1380

144

E

1304

142

D

1362

143

C/E

1398

145

Table 7 - The original string scalings and plucking points as suggested here

 

         The scalings from Table 7 are plotted below in Figure 23 below.

 

Figure 23.wmf

Figure 23 – Graph of the original string scalings as established in this report assuming an enharmonic original compass.

 

         This graph is markedly different from the scaling graph of Figure 21 above which resulted from the 1974 determination of the original compass.  Most noticeably this graph gives scalings, like all of the other instruments by Guarracino, that are accurately Pythagorean throughout the whole of the compass above the bridge mitre.  Even the note A, the lowest note on the main section of bridge just below the bridge mitre, has a string length very close to the straight line representing Pythagorean scalings.  It should be noted that only by assuming a broken octave in the bass is this so.  If an ordinary C/E short octave is assumed the points corresponding to the notes at the bridge mitre lie well to the right-hand side of the straight line corresponding to Pythagorean scalings with c2-equivalent scalings that are greater than the treble scalings.  This condition would be very unusual and is not allowed by the design of most Italian makers.  Indeed this is the main reason for assuming that this instrument originally had a bass broken short octave.

         The enharmonic C/E to c3 50-note compass with a broken short octave suggested by this analysis can thus be seen to represent normal practice in Naples in the second half of the seventeenth century, and normal practice for Onofrio Guarracino who made some other instruments with enharmonic keyboards.  For example, the harpsichord in the Collection of Fernanda Giulini in Milan[48] originally had an enharmonic compass of C to f3 chromatic with 19 notes to the octave for most of the compass and a grand total of 84 notes.  Part of the compass is re-constructed digitally in Figure 24 below:

 

Figure 24.jpg

Figure 24 - A digital re-construction of part of the original 19-note-to-the-octave keyboard.  There is now no way of knowing whether the raised naturals were originally white or black.

Undated single-manual harpsichord attributed by me to Onofrio Guarracino, Naples, c.1680.

Collection of Fernanda Giulini, Milan

 

         The single-manual harpsichord in the Metropolitan Museum of New York, Cat. No. 89.4.1231, also attributed by me to Guarracino originally had an enharmonic compass of an unusual type.  Although the keyboard compass has been changed[49], like the Guarracino harpsichord in Milan, it retains its original balance rail from which the original compass can be read unambiguously from the plugged balance-pin holes.  The treble and bass ends of the balance rail is shown below in Figure 25 below in the top and bottom photographs respectively.  The plugged holes of the original balance pins are indicated with arrows in each case.

 


 

Figure 25a.jpg

 

Figure 25b.jpg

Figure 25 - The bass end of the balance rail (top) and the treble end of the balance rail (below) showing the present balance pins placed slightly towards the rear of the balance rail and the plugged holes for the original balance pins placed several millimetres further forward and each indicated with an arrow.  A number of the plugged holes for the accidentals can be seen to be doubled indicating that the original keyboard had an enharmonic keyboard with several split accidentals.

Single-manual Italian harpsichord attributed here to Onofrio Guarracino, Naples, c.1670

Metropolitan Museum of Art, New York, Cat. No. 89.4.1231

 

 

The plugged holes in the balance rail indicate that it originally had a highly unusual compass in the bass and reached to f3 in the treble.  From tenor c upwards there are no two octaves with the same number or combination of split notes and notably it had no split notes in the top octave.

 

Figure 26 - Digital re-construction of the original keyboard.

Undated single-manual harpsichord attributed by me to Onofrio Guarracino, Naples, c.1675

Metropolitan Museum of New York, Cat. No. 89.4.1231

 

         The compass of the Metropolitan Museum Guarracino harpsichord is intriguing.  Close attention to the bass end of the keyboard reveals two consecutive sets of three accidentals in a row!  The second set of three accidentals must begin with an FT and the notes then follow logically after that since this second set of three accidentals clearly must belong to the normal notes between F and B in the light of the notes immediately above them.  But how then is the first bottom set of three accidentals to be interpreted?  Part of the answer to this question lies, I feel, in noting that none of the bI accidentals is split anywhere in the compass so that there was never the possibility of playing an aT at any point throughout the compass.  An aT would be necessary in the remote key of B major with 5 sharps in the key signature or, closer to home, in b minor.  Any further modulation in the sharp direction requires bT’s or double sharps which are also not provided.  However it is clearly impossible to play in B major as well because of the lack of an aT, so that a lower B1 is not required as a root to a B major chord.  I suggest therefore that in this first group of notes with three consecutive accidentals, there is no B1 although BI1, required as a root to a BI major chord with only two flats could be provided.  According to my interpretation therefore the bottom notes on this keyboard might have been tuned as follows:

G1, A1, BI1, C, CT, D, EI, E, F, FT, G, GT, A, BI, B, c, cT/dI, d, dT/eI, e, f, etc.

Here the boxed notes represent the raised accidentals.  This compass would provide roots in meantone tuning to all of the chords (including a 16' G1 in g major) playable with the split accidental available higher up in the compass.  However, another possibility might have been:

F1, G1, A1, C, CT, D, EI, E, F, FT, G, GT, A, BI, B, c, cT/dI, d, dT/eI, e, f, etc.

In this configuration, if the left hand of the player reached out to play successive ‘octaves’ on the bottom two naturals, the player would sound F1 and A, and G1 and B, which would provide a kind of 16' for final tenths in F major and G major.  The logical, but perhaps fanciful, development of this idea would then be to play an octave on the raised A1 to the cT at the front of the latter key to give a final tenth in A major as well.  Neither of these two possible solutions would provide a lower B1, although the first does provide for a lower BI1.  The lower B1 would, I think, never be required although a bottom BI1 might occasionally be a useful provision as a root to a chord in tonalities of ex-tempore music performed, and also for music written, in the period contemporary with this instrument.

         One of the most disturbing features of the keyboard of the New York ‘Guarracino’ is that there are no aT’s combined with a bI as a split sharp as already discussed.  On the other hand it is singularly curious that one gI is provided in the middle of the compass.  The gI suggests modulations in the flat direction as far as DI major (bI minor) with 5 flats, whereas the lack of an aT means that it is possible to modulate only as far as E major (cT minor) with only 4 sharps, making it impossible to play in B major.  Christopher Stembridge[50] has pointed out that Neapolitan composers greatly favoured the sharp tonalities in favour to those in flats, and that one should really therefore expect aT’s.  Stembridge therefore makes the interesting suggestion that the keyboard begins with a normal C/E short octave as usual.  However, the main part of the keyboard is then to be considered a transposed keyboard at 12' pitch with the f notes playing a c and the g’s playing a c.  This then means that what looks like a dT/eI is playing an aT/bI, and the surprising fT/gI is playing cT/dI.  There is a number of difficulties with this scheme not the least of which is the lack of any evidence either among the surviving instruments and in the archival documents of instruments at 12' pitch transposing by a fifth[51].

         RCM 175 is therefore probably the only harpsichord by Guarracino so far brought to light that originally had a bass broken short octave.  Two other Guarracino harpsichords, one in the Boston Museum of Fine Arts (Cat. No. 1986.518) and another in the Beurmann Collection, Museum für Kunst und Gewerbe, Hamburg (Cat. No. 8)[52] both have standard chromatic C/E to c3 compasses with no split accidentals and are therefore like the harpsichord in private ownership in Rome in this respect.  Neither of the first two harpsichords is signed by Guarracino, but all of them can be attributed to him with a high degree of certainty.  By the establishment of this compass and keyboard that I propose here we avoid the highly-unusual C/E to d3,e3 compass suggested by John Barnes[53].  I therefore feel confident that it is possible to say that the original compass of this instrument was C/E to c3 with a broken short octave, 3 sets of split accidentals, and a total of 50 notes.

 

Pitches and Design Scalings of Instruments built by ONOFRIO Guarracino

         A thin straight line has been drawn on the graph of Figure 23 which gives Pythagorean scalings based on c2 = 12 Neapolitan once = 259.4mm.  The plotted points for the main bridge section (not including those for the strings and pins on the bass bridge mitre) all lie on, or very close to, this line.  It seems likely therefore that Guarracino based the design of this harpsichord on scalings of 12 Neapolitan once or simply on 1 Neapolitan palmo.  It turns out, therefore, that the design scalings of this harpsichord are strikingly simple when viewed from the point of view of the unit of measurement used by Guarracino to design and build this instrument.

         This design scaling is, however, different from that used in most of the other surviving harpsichords and virginals which are either signed by Guarracino or can be attributed to him.  Although most of the virginals use a design scaling of 13½ once, two of the surviving virginals use design scalings of 14 once (or an octave above this at 7 once) and one virginal is also designed around a c2 scaling of 12 once.

 

                                                                                                                 Design scaling

                        Instrument                                   Location Neapolitan once                              mm

                       1652 virginal                             Donati, Florence                              12                             259.4

                       1663 virginal                           Tagliavini, Bologna                           13½                             291.7

                       1667 virginal                              La Scala, Milan                               14                             302.7

                       1667 virginal               Musikinstrumentenmuseum, Berlin              12½                             270.4

                 1668 bentside spinet           Museo strumenti musicali, Rome                13½                             291.7

                       1677 virginal                 Museo strumenti musicali, Rome                13½                             291.7

                       1679 virginal                   Mirrey Collection, Edinburgh                   13½                             291.7

                       1692 virginal                 Museo strumenti musicali, Rome                13½                             291.7

                        n.d. virginal                   Harps. Clearing House, Boston                   14                             302.7

                  1694 octave virginal        National Music Museum, Vermillion                7                             151.4

 

                    1651 harpsichord                   Private collection, Rome                        14                             302.7

                    n.d. harpsichord               Gemeentemuseum, The Hague                   14                             302.7

                    n.d. harpsichord                     Fernanda Giulini, Milan                       14¼                             308.2

                    n.d. harpsichord            Metropolitan Museum of Art, N.Y.                14                             302.7

                    n.d. harpsichord                Museum of Fine Arts, Boston                    14                             302.7

                    n.d. harpsichord              Museo strumenti musicali, Rome                13½                             291.7

Table 8 - String scaling designs of Guarracino virginals, spinets and harpsichords

 

         Seen in the light of the other instruments made by Guarracino, RCM 175 is therefore unusual in its design scaling compared to the other surviving harpsichords, but is the same as the 1652 Guarracino virginal.  The intriguing question raised by the variety of the scalings used by Guarracino is both the absolute pitch of the instruments he built, and their pitch relationships to one another.  Normal brass scalings for a pitch of about a1 = 415.3Hz, an equal-tempered semitone below ‘standard’ concert pitch of a1 = 440 Hz, are about c2 = 285mm.  The scalings of the instruments designed around a scaling of 13½ once = 291.7mm are close to this and, with this scaling, were probably tuned to a pitch near a1 = 415.3Hz or perhaps a slight bit lower.  What, then, about the instruments with the other design scalings?  To what pitch were they tuned?

         The spectrum of design scalings used by Guarracino therefore ranges through the numbers 12, 12½, 13½, 14, and 14¼ once.  Notably missing from this range is a design based on c2 = 13 once = 281.1mm which might today have been considered the, almost ‘standard’, pitch scaling for a1 = 415.3Hz.  Among those scalings that do exist in the instruments of Guarracino one is immediately let to ask how these numbers relate to one another, and do any of them correspond to common musical intervals or pitch differences?  Many instrument makers, both north and south of the Alps, designed their instruments to be tuned to pitches a tone apart.  I have shown, for example, that the Ruckers family designed their instruments in two pitch families a whole tone apart with each pitch family consisting of instruments with pitches of a fourth, a fifth, and an octave between them[54].  I have discovered that in Venice as well, instruments were designed around f scalings that were either 9 Venetian once or 8 Venetian once.  The ratio  = 1.125 is the exact ratio of two notes a Pythagorean whole tone apart.  Clearly the different sizes of the oncia used in Venice and that used in Naples do not allow for such a tidy relationship between the scalings expressed in integral whole numbers of Neapolitan once.  However, Table 9 below shows that the same ratio is possible among the various design scalings used by Guarracino:

 

                Design scaling                               Pitch difference                           Musical

                     numbers                Ratio                 in cents                                  interval

                      13½/12                  1.125                   203.91                     Major whole tone

                      14/12½                  1.120                   196.20                     Minor whole tone

                    14¼/13½                1.055                    93.60            About a minor semitone

                        14/12                   1.167                   266.87 Between a major tone and a minor third

                    14¼/12½                1.140                   226.80        Greater than a major tone

 

Table 9 - Table of Guarracino’s design-scaling relationships

 

         Table 9 therefore shows that, like the Venetian instruments with design scalings of 9 and 8 Venetian once, those instruments with design c2 scalings of 13½ once and 12 once also have scalings which differ by exactly one Pythagorean whole tone from one another.  Instruments with these scaling differences therefore also differ in pitch from one another in exactly the same way that I found for the Flemish instruments of the Ruckers family referred to above and the Venetian instruments with f scalings based on 9 once and 8 once.  It seems likely therefore that RCM 175 and the 1652 Guarracino virginal were both designed to sound one whole tone above those relatively common instruments whose design scalings are 13½ once.  The 14/12½ design scalings also give a pitch difference with ratios close to the 1.125 ratio for the string lengths and frequencies of notes separated by a Pythagorean minor whole tone.  These numbers may represent a pragmatic solution using relatively simple numbers for the design scalings to give instruments at pitches a whole tone apart, with each designed for a slightly different pitch level.  Such a scheme could have been required during this period when the concept of a pitch ‘standard’ did not really exist., the 14¼ and 13½ scalings correspond to pitches about a minor semitone apart and may have been a way for Guarracino to design an instrument a semitone below those much more common instruments with scalings of 13½ once.

The bottom three scaling differences given in Table 9 do not seem to correspond to any of the usual pitch intervals in common use in the seventeenth century although, as just mentioned, they are all individually necessary to provide the first three pitch intervals.  The 14¼ and 13 design scalings gives a pitch relationship almost midway between a semitone and a tone, and therefore cannot have been thought of as bearing a pitch difference relative to one another corresponding to a musical interval in any of the Western scales.  The others ratios also do not yield any satisfactory musical intervals.  Unfortunately little seems to be known about pitch levels and pitch relationships for other instruments designed and built in Naples with which these numbers and pitches could be compared.

         If, for argument’s sake, it is assumed that those instruments with design scalings of 13½ once = 291.91mm could be tuned to a pitch of a1 = 415.3 Hz, the contrived and so-called ‘Baroque pitch’, then the pitches of the instruments with Guarracino’s other design scalings can be calculated.  These are shown in Table 10 below where the 13½ once  design scaling has been highlighted.  Here, the longer-scaled instruments are at a lower pitch and the shorter-scaled instruments are at a higher pitch.  Seen in this way the gamut of Guarracino instruments has pitches from about one semitone below a1 415.3Hz to about a major tone above this.  It is important to remember when looking at this table that most virginals have a design scaling of 13½ once, and that most harpsichords have a design scalings of 14 once.  This suggests that the pitch to which most harpsichords were tuned was lower than a1 = 415.3Hz, and that RCM 175 would originally have been tuned to a pitch about a major tone above a1 = 415.3 Hz.

 

                            Design scaling                                      Pitch level

                        once                mm                         in cents            in Hz

One whole tone  14¼               308.13                          -93.6                393.4

 

                  14                302.72                          -63.0                400.5

 

         13½               291.91                             0                   415.3

                         12½               270.29                          133.2               448.5

                          12                259.48                          203.9               467.2

Table 10 - Table of possible pitch levels for Guarracino’s different design-scalings

 

Here the difference between the pitch design scalings has been expressed in cents and also, assuming that the 13½ once design scaling corresponds roughly to a1 = 415.3 Hz, the pitch levels of the other design scalings has been calculated.  How these pitch levels might have fitted into the daily musical life of seventeenth-century musicians performing in Naples remains a matter of conjecture.

 

The lateral string spacing: 

         Like the lateral spacing of the jackslots, the lateral spacing of the strings was designed by the maker of any harpsichord in terms of his local unit of measurement.  The determination of this spacing therefore has the potential for a further determination of the unit of measurement used by the maker.  Normally the strings from the tenor part of the compass upwards are parallel to one another and usually to the spine side of the instrument.  In the bass, however, the strings are sometimes angled away from the spine.  The reason for this is clearly to make extra space between the bass end of the bridge and the spine side of the instrument in order to increase the flexibility of the soundboard at the bass end of the bridge and, in so doing, to improve the quality of the extreme bass notes.  Angling the strings away from the spine so that the bass end of the bridge is distanced from the spine and spine liner achieves the required area and flexibility in the bass part of the soundboard.  This therefore means that the lateral bridge-pin spacing is normally different in the bass from that in the rest of the compass.  However, if the strings are parallel to the spine throughout most of the compass, the lateral spacing of the strings should be the same as the lateral spacing of the jackslots from the tenor upwards.  This should come out of the analysis of the lateral string spacing and should be true even though the original nut is not present whereby the lateral spacing of the strings cannot be determined at the near end of the strings.

         In order to determine the lateral spacing of the strings, the distance of the bridge pins of the long strings of the present compass from the spine was measured.  The bridge pins for the present long strings are, indeed, the original bridge pins.  These measurements are given in Table 5 above.

         These measurements are plotted in Figure 27 below.  From this graph it is clear that the points form a single straight line for the whole of the compass.  This implies that the strings were not angled away from the spine in the bass as is normal practice.  The strings were therefore parallel to one another throughout the whole of the compass of the instrument. 

The measured distances of the bridge pins relative to the spine were fitted to the equation of a straight line in the same way as that used for the lateral spacing of the jackslots.  The slope of this line found in this way is 13.703mm/string pain which is slightly different from that found for the lateral spacing of the jackslots (which was 13.768mm/jackslot).  However the difference between these two is only 0.6%, and therefore the two agree with one another to within the experimental error[55].  Effectively the two slopes are the same and therefore, to within the experimental error, the spacing of the jackslots is the same as the spacing of the strings.  This is clearly what one would expect given that the new register, made by slicing up the original register, had the same spacing as in the original design.

 

Figure 27.wmf

Figure 27 – Graph of he lateral string spacings measured the inside of the spine to the bridge pins

 

 

Commentary:

         In his restoration report John Barnes gives 1580 as the approximate date of this instrument.  The catalogue gives the date c.1610[56].  However the fact that it was built by Guarracino means that it must date from the period between about 1651 to 1694, the date range of the of the surviving instruments of Guarracino.  On the basis of the similarity of the mouldings of this instrument to the mid-period dated virginals by Guarracino, I have tentatively dated the instrument to the middle period of Guarracino’s activity around 1675. However, the fact that the scalings of RCM 175 are based on 12 once = 1 palmo, the same as the 1652 virginal belonging to Pier Paulo Donati suggests that it might date from early in Guarracino’s career.  However, there is, so far as I can tell, no other indication on which to base a date for this instrument.  It is at any rate about 100 years later than originally thought.  Indeed, the gold designs painted onto the green ground which decorate the soundwell, nameboard, jackrail, and the front edge of the case are very freely and rhythmically painted in a style characteristic of the other instruments of Guarracino made during the second half of the seventeenth century.  Unlike late Renaissance decoration from around 1580 which is characteristically much more formal, geometric and restrained, the decoration on this instrument is very free and uninhibited in a manner totally atypical of the period around 1580 or even 1610.  Therefore, just on stylistic grounds the decoration (which might, of course not be original) of the instrument could certainly not date from the period around 1580.

         Although it is true that a number of single-strung Italian instruments is of sixteenth-century origin[57], there are also many seventeenth-century instruments which were originally single strung.  For example, the 1619 harpsichord with the signature of Giovanni Battista Boni harpsichord (Brussels Museum of Musical Instruments No. 1603), and an anonymous harpsichord dated 1630[58] in the Museo Nazionale degli Strumenti Musicali in Rome were both single-strung and date from the seventeenth century.  Indeed most, if not all, of the harpsichords which are either signed by or attributed to Onofrio Guarracino were originally single strung, and these all date from the second half of the seventeenth century.

         The alterations to this instrument to modify it to bring it into line with the musical requirements and taste fashionable at that time are very interesting.  The original compass of C/E to c3 is very modest for an instrument of this date even with a bass broken octave and an enharmonic treble compass.  The alteration to State 2 (see page 52 below) with a normal bass short octave and with a chromatic compass to f3, although it would have resulted in strongly non-Pythagorean scalings, extends the original compass considerably in the treble.  It seems likely that, in this state, the pitch was dropped to ‘normal’ pitch and that iron stringing would have been required in at least part of the treble compass (see Appendix 1).  The alteration to State 3 probably arose out of a need towards the end of the seventeenth century for a chromatic bass octave (especially for the notes EI and FT) and for the lack of necessity of transpositions involving top notes above c3.  There is no way of being sure about the scalings or pitch of the instrument in States 2 and 3 however, as the original balance rail (and the intermediate balance rails if there were any) are now missing, and there is no way of telling if the nut or the treble section of the bridge had been moved during the period of these alterations. 

         The final alteration to State 4 was much more ambitious than the previous ones and involved much more work.  The addition of the second choir of strings with the second register of jacks made the instrument thoroughly up to date.  Most of the instruments of Cristofori and Ferrini have a brass c2 scaling of about 285 to 289mm.  The deliberate shortening of the c2 scaling of this instrument to 261mm by moving both the treble section of the bridge and the nut closer to the gap, suggests that brass stringing was intended at a pitch a semitone to a tone higher than usual.  On the other hand, as the scalings are mostly non-Pythagorean, it is difficult to say with certainty to what pitch the instrument was intended to sound in this state.

         The blocks used to limit the movement of the registers in State 4 are also very interesting.  The fact that the rear block is immovable and that there is no easy way, such as with a register lever, of engaging or disengaging it suggests that the front register was turned off only during tuning when the jackrail could be removed or at least moved back.  Otherwise it must normally have been played with both registers engaged.  The 1.5mm movement of the front register to disengage it has always remained unaltered and provides an example of one aspect of original historical regulation.  The marking of the plucking direction of the jacks on the registers seems original to State 4 and provides an unambiguous indication that the rear register plucked to the left and the front register plucked to the right.

 

Conclusions:

         The attribution of this harpsichord to Onofrio Guarracino adds yet another layer to the story of harpsichord building in Naples.  As indicated here there are many other unsigned and undated harpsichords which can be attributed to Guarracino and these instruments, along with the virginals designed and made by him, display a large variety of sizes, compasses, string-scaling designs and pitch levels.  This study shows that there was an important school of harpsichord and virginal building that was extremely active in Naples, a city sometimes otherwise forgotten about in the history of the making of the harpsichord in Italy.  Naples during this period was, however, a wealthy, busy port and enjoyed a high level of culture and intellectual activity.  Work that I have done recently indicates that many other instruments were also built in Naples, but these unfortunately cannot be ascribed to any of the builders known to have been active in Naples because of the scarcity of any signed instruments with which the unsigned instruments can be compared.

         Although much attention has been paid to harpsichord building in Florence, Rome and especially in Venice, there was also an active and important school of instrument building with its own individual characteristics producing a large number of instruments in Naples.  The grouping of all Italian instruments into a single school or tradition is misleading:  the Italian peninsula was, within itself, as varied as the politically-separate states north of the Alps.  It must be remembered that, before unification - the so-called Risorgimento - in 1861, Italy was a disparate collection of church and city states along with minor principalities and independent regions.  It should therefore not be surprising that there is as much difference in the construction styles of Venetian and Neapolitan instruments as there is between instruments made in Paris and London which are geographically considerably closer to one another.  It should also therefore not be surprising that there is as much - or more - difference in the sounds of Venetian and Neapolitan instruments as there is a difference between  the sounds of English and French harpsichords which are well-known and recognisable to many who play and listen to the harpsichord.  More will be said about the sound qualities of Neapolitan harpsichords below. 

         The prevalence in Naples of harpsichords with enharmonic keyboards gives a significant insight both into the music performed there and into a musical climate interested in extending the musical possibilities of meantone tuning and enharmonic modulations which are possible with the additional split accidental notes.  It also suggests a lack of any rigid system for the tuning of the bass compass of some harpsichords which have an extended bass compass down to bottom F1 or G1 with what is, to a modern musician at least, a strange and unusual succession of bass notes.

         The determination of the original compass of RCM 175 simultaneously simplifies the previous interpretation of the compass and scalings of this instrument and also adds further to our knowledge both about Guarracino and about musical practice in Naples during the second half of the seventeenth century.

         The attribution of the instrument to Onofrio Guarracino, and the determination of the instrument’s original compass now, in 2008, some 34 years after the restoration and analysis of the instrument in 1974, to me also provides additional support to the cautionary philosophy which clearly needs to pervade restoration and conservation practices in modern times.  A great deal of research and effort was necessary to arrive at the conclusions reached in this report.  As it turns out, probably no evidence was destroyed or altered in the instrument during the 1974 restoration which would have then subsequently altered the conclusions reached here some 34 years later.  On the other hand, much more background work might have been done in 1974 if the maker and the original compass of the instrument had been known then.  It would certainly then have been inconceivable to provide the instrument with a stand clearly based on a Venetian model!!  Also a great deal of mis-information has been dispersed (including that in the RCM catalogue published as recently as 2000) in the intervening period.  We can, of course, not turn back the clock.  However, every restorer and every individual and institution commissioning restoration work is clearly interfering with the past and is often doing so in significant ignorance of the facts as happened with John Barnes and myself in 1974.  If anything, therefore, these final remarks are directed towards a continuation of the cautionary philosophy generally prevalent in the field of the restoration of musical instruments.  The future will reveal new facts and new truths, and no restorer wants to be guilty of destroying evidence of original working practices and methods that will yield important evidence not only about historical instrument building techniques, but also the way in which these affect our interpretation of the music performed on these instruments. 

 

The Sound of Neapolitan Harpsichords:

         My attention was first drawn to the remarkable sound of a harpsichord that I restored for the Museo Nazionale degli Strumenti Musicali in Rome in 1980[59].  I have been able to show that, although not made by Guarracino, this instrument uses the Neapolitan unit of measurement and, along with numerous other characteristics of the Neapolitan style, must have been made in Naples.  The sound of this instrument has remained with me to this day.  During the course of my career I have restored quite a large number of instruments of all national and regional schools.  However, this instrument (which I will refer to here as Rome 779) had what seemed to me to be the perfect sound for a harpsichord - it was a simple, clear sound that didn’t impose itself on the listener but was, on the other hand, the perfect servant of the music.  It was a sound that was at once extremely beautiful but at the same time did not in any way intervene between the musician or listener and the music being played on it.  Despite the clarity and even-ness of the tone, one was aware first and foremost of the music and not of the sound of the instrument.  In this respect its sound was quite unlike the seductive sound of the fashionable French and Flemish instruments which are much more common then and today.  Quite frankly I was dumbfounded by the beauty, clarity and even-ness of the tone of such a simple, plain instrument. 

         At the time that I carried out the restoration of this little Rome harpsichord I myself had no idea that it was made in Naples.  At that time I had not yet developed my ‘unit of measurement’ procedure for determining where an instrument was originally made[60].  Once I was able to determine where instruments were built, two startling realisations were made:  firstly, there was a vast number of unsigned and usually undated harpsichords made in Naples in many of the museums and collections throughout the world and un-recognised as such.  Secondly, there was a number of seventeenth-century Neapolitan harpsichords which had been re-worked, up-dated and up-graded by Bartolomeo Cristofori and Giovanni Ferrini working in Florence at the beginning of the eighteenth century.

         Why, I asked myself, would a builder of the stature of Cristofori, the inventor of the piano escapement mechanism and one of the most inventive and skilful builders I know, bother to re-work instruments from as far away as Naples to Florence, which was land-locked and a long way from access to the sea and cheap transport?  Why didn’t Cristofori and Ferrini simply up-date the seventeenth century harpsichords made locally in and around Florence by such makers as Vincentius Pratensis, Giovanni da Pertici and Stefano Bolcioni?  There must have been some significant reason for choosing to import the Neapolitan instruments into Tuscany and Florence[61].  Could it be that I already knew the answer and that it lay in the sound of Rome 779 that I had restored as long ago as 1980?

         Since the early days of my realisation that there was something special about the sound of Neapolitan harpsichords, I have only been re-assured in my conviction that Neapolitan harpsichords are indeed significantly different from the harpsichords made elsewhere in the Italian peninsula.  More than ten years of experience now in the building and restoration of Neapolitan harpsichords has confirmed my belief that Neapolitan instruments generally have many of the qualities of Rome 779 made in 1630.  I once, in a desperate search to describe the indescribable properties of these instruments, said that their sound was ‘Like the sound of pearls dropping onto sheets of diamond!’.  Meaningless as this is as an adequate description of their true sound, it at least goes some way towards understanding the sensation one has when listening to a correctly-strung and well-restored Neapolitan harpsichord.

         In Appendix 3 we read that even mother and child virginals were made in Naples, possibly even by Guarracino.  The elusive gut-strung tiorbino was also built in Naples.  These are both unusual and very interesting types of keyboard instrument.  The realisation that such unusual instruments were made in Naples is then immediately followed by the question:  “How were these instruments used in the day-to-day musical life of this rich and cultured city?”  Additionally it is clear from what has been explained here that  instruments with elaborate and complicated enharmonic keyboards seem to have been commonplace in Naples, but have been very little researched[62].  I hope that this paper generates further interest in the Neapolitan school of harpsichord building, in the instruments of Onofrio Guarracino, in the music composed for these instruments, and particularly in the universality of the sound of Neapolitan instruments in the interpretation of the music of other periods and locations when performed on them.  Those at the Medici Court and Bartolomeo Cristofori and Giovanni Ferrini, at least, must have been convinced of this and convinced that Neapolitan harpsichords were really something rather special!

 

The Attribution of RCM 175 to Onofrio Guarracino:

            The use in this instrument of Guarracino’s unit of measurement for its design and construction is, on its own, probably sufficient proof that Onofrio Guarracino is its maker.  But as John Barnes was always careful to point out there is now never any ‘proof’ that an instrument is by a given maker.  Only someone who has actually seen him at work has ‘proof’ that he was the maker.  As John Barnes also pointed out the process of attribution involves the consideration of many different features which eventually all do or do not point in one direction namely, to the author of the work under consideration. 

            The harpsichords, virginals and spinets discussed here display a wide variety of appearances: some are virtually unaltered from their original state, and others have been altered and now bear little resemblance to their initial state.  When confronted by a new instrument how is one to decide definitely whether or not it is a genuine product of the Guarracino workshop tradition?

            Partly the ability to establish the authenticity of a Guarracino instrument is based upon experience, and recognising which characteristics can be used to authenticate an instrument becomes easier as experience is gained.  As more and more instruments are examined, recurrent features become apparent which can be used to decide if a new instrument is genuine.  After one has seen a number of undoubted examples, all of which have similar features, it is possible to compare these with new instruments which have been altered or which may have lost certain of their identifying features such as their rosette and keyboard.

            In theory a genuine Guarracino instrument must exhibit all of the features outlined in below.  But in practice it is usually necessary to look for only a few features of the construction and marking out.  Fortunately it is only rarely really difficult to decide about the authenticity of an instrument.  Usually if even a few of the characteristic features are missing or are atypical or wrong, then one soon finds that there is nothing that, in detail, is typical of Guarracino’s usual practice.  On the other hand if only some of the characteristics of Guarracino’s usual practice are present, then as the examination continues, more and more features are discovered which are typical of Guarracino’s normal workshop methods, until the evidence that the instrument is genuine becomes overwhelming.  It is perhaps most notable in the discussion below that there is no contra-indication among the instruments that any of them is not by Guarracino

            In my opinion too much weight has been placed in the recent past on the use of moulding shapes to identify and authenticate instruments.  We have no more ‘proof’ that mouldings were made in house in the maker’s workshop than we do that the soundboard rosettes were made in house.  It is only an assumption that the mouldings were made in house[1].  There is, for example, a number of Venetian instruments by at least 4 different makers all of which use soundboard rosettes made using the same pattern.  If soundboard rosettes weren’t made in house but were bought from a common supplier, why should we assume that mouldings were?  In my work on the instruments of the Ruckers family I found evidence that the sharp bocks and the jacks (and therefore probably the registers) were provided to the workshops of Ioannes and Andreas Ruckers by a common supplier[2]

            On the other hand there is a large body of evidence both from archival sources and from the instruments themselves that scratch moulding cutters were used in harpsichord makers’ workshops and that they were used in house to make mouldings.  In an inventory of 1702 of the tools left in the workshop of Giuseppe de Bonis Cortona, the son of the more well-known Gian Battista Boni of Cortona[3], there were “Diece ferri da scorniciare” or 10 irons for making mouldings.  But, for example, the Medici archives indicate that no less a maker than Bartolomeo Cristofori working in Florence for the Medici Court in 1700[4] farmed out the making of the cases to an ebanista (furniture maker and finisher).  Did the ebanista work in Cristofori’s workshop or in his own?  Did he use Cristofori’s scratch moulding cutters or did he use his own?  The fact that at least one maker is known to have had the cases of his instrument made by someone else means that we cannot be absolutely sure that the attribution of an instrument to a maker can be made with complete certainty on the basis of the mouldings alone.  Nonetheless, the use of moulding profiles in the attribution and dating of some instruments has been extremely fruitful in the attribution of some instrument to specific makers[5].  In my opinion, however, a moulding comparison is but only one part of an attribution since it is based on the assumption that mouldings were always made in the workshops of the makers themselves.

            In using the unit of measurement as a tool for the identification of the maker of harpsichords, virginals, spinets and clavichords I am also making an assumption.  As John Koster has pointed out there is always the possibility that the tools, including his rulers, were handed on from master to apprentice or from father to son.  This means that Guarracino may have inherited his rulers from his master, and that some of the instruments attributed here to Guarracino on the basis of the unit of measurement used in their design and construction were designed and built by one (or more!) of Guarracino’s antecedents.  Thus an instrument found to be based on the same unit of measurement as used by Guarracino may have actually been built by one of Guarracino’s masters.

            Another assumption made when using the unit of measurement to locate an instrument’s maker is that the maker did not bring his tools and measuring sticks with him from somewhere else.  This clearly happened with the maker Giovanni Natale Boccalari who worked in Naples and whose activity there is documented in the Neapolitan archives.  Four of his instruments survive:  in the collection of Fernanda Giulini, Briosco, Italy[6], in the collection of Jean-Pierre Decavalé Toulouse, France, in the collection of J. C. Zehnder, Basel, and in the Musikmuseum of the Historisches Museum Basel (Nº 1956.639).  The Giulini harpsichord is signed “Io Giouanni Natale Bocalari de Vfida fece questo Cibolo[sic] in Napoli alli 20 de Luglio de lanno 1679  Che da lasame / a(n)ni fa (?)” [I Giovanni Natale Boccalari from Ufida made this harpsichord in Naples on the 20th of July in the year 1679.  {I} took the exam ?? years ago] in ink on the baseboard in the keywell just in front of the line scribed on the baseboard marking the wrestplank position.  The very last part of the inscription giving the number of years before the construction of the instrument that Boccalari took the exam is illegible.  ‘Ufida’ is an antiquated dialect form of Offida, a small town in the Province of Ascoli Piceno in Le Marche which lies about 15 km inland from the Adriatic and about half way between Ancona and Pesaro.  The signature implies that Boccalari, although he was from Offida, took the necessary exam to qualify as a harpsichord builder within the Neapolitan guild system.  Le Marche was part of the Papal States when the instrument was built, and Offida, like Rome and most of the other centres in the Papal States used a unit which was a multiple of the oncia = 18.62mm[7].  The instruments by Boccalari in the Giulini Collection and the two in Basel have all been measured by the author and have been found to use a unit of measurement which averages 18.65mm – ie the unit used in Offida, Rome and throughout the Papal States – and not that used in Naples.  Therefore the use of the unit of measurement in assigning the location of construction of these instruments would, used on its own, be totally misleading.

            Therefore, in making the attribution of this instrument and the others discussed here to Guarracino, I have tried to use many of its features and style of construction to compare these attributed instruments with those signed by Guarracino without relying on any single feature on its own.  I admit that not all of the instruments have all of the characteristics listed below, but it is also true that the instruments of Guarracino cover a period of almost a half a century during which time his working methods doubtless changed and evolved.  A comparison of their mouldings does form a part of the attributions (see below).  However, the discovery that a large group of otherwise anonymous harpsichords uses the same unit of measurement and that they have many other features in common lends at least a certain weight to the likelihood that the attributions are sound.  Indeed, no other maker is known who used his highly idiosyncratic and individual size of unit slightly different from the ‘text-book’ Neapolitan unit of measurement which, as illustrated above, would have been quite different from that found here.  In addition this instrument has many Neapolitan characteristics[8] and characteristics of the distinctive working methods of Guarracino.  These features are listed on my website together with additional details to those given below at:  http://www.claviantica.com/Characteristics.htm. 

            Among these characteristics are the following:

1.      The case has a sharply-pointed tail, a feature of many, but not all Neapolitan harpsichords.  However, all of the harpsichords signed by or attributed to Guarracino have sharply-pointed tails.

2.      The case sides project below the baseboard and usually consist of a lower part of chestnut (see Figure 8 and Figure 10).  The chestnut strip is a feature of some other Neapolitan builders, but is not necessarily found on all of the harpsichords attributed below to Guarracino.

3.      The case sides of this instrument are made of sycamore and not of cypress.  Although most of the instruments of Guarracino have case sides of cypress, at least two of the other harpsichords that I have attributed to Guarracino also have case sides of sycamore.  On the other hand other Neapolitan instruments are sometimes constructed using sycamore, a material which seems to have been used only in Naples.

4.      The plan of the harpsichords by Guarracino are usually scribed onto the baseboard (part of this plan can be seen in Figure 12).  This plan was scribed onto the baseboard as part of the initial design of the instrument before any of the framing, keyblocks or wrestplank were glued in place.  The plan scribed onto the baseboard of this harpsichord is similar in plan and execution to that found on the baseboards of many of the other harpsichords by or attributed here to Guarracino.

5.      The rosette is in the style usual in instruments of Guarracino (see Figure 2) and of other Neapolitan makers.  The other instruments of Guarracino whether virginals, spinets or harpsichords all have rosettes in two parts with one flat layered part glued to the top of the soundboard and an inverted ‘wedding cake’ type of rosette in several levels each consisting of several individual layers of parchment and wood positioned below the soundboard.  The style and grammar of the elements used in the rosette designs are all very similar, with the same type of individual elements used together in different ways to make up the whole rosette.  The parchment layers are, like those of the rosette in this instrument, often gilt or coloured, and often at least one layer in each level is of a thin veneer of wood with a thin parchment backing.  In some cases, especially in the virginal rosettes, all of the layers making up the different levels appear to be of wood with a parchment backing.

6.      The jackrail and jackrail supports in RCM 175 are in the usual Neapolitan style (see Figure 18 and Figure 19).  All of the instruments of Guarracino whether spinets, virginals or harpsichords use this style of jackrail and jackrail support.

7.      The keyboard slides in and out like a drawer, a feature found on all of the other instruments by Guarracino, but also on some non-Guarracino Neapolitan harpsichords.  The side projections of the keyboard can be seen indistinctly in the photograph shown in Figure 15.

8.      The keyboard rests on a heavy support glued to the baseboard.  This feature, although found on other non-Neapolitan instruments, is a feature of all Guarracino instruments.

9.      The outside edges of the top cap mouldings are flush with the mouldings below them and do not project in the way characteristic of most north-Italian instruments.  This can be seen clearly at the top of Figure 8 and of Figure 14.  This is a characteristic of all of the instrument by, or attributed here, to Guarracino, but not of all Neapolitan instruments.

10.  There are carved keywell scrolls typical of the Neapolitan tradition (see Figure 5 in the colour section on page xxx).  Carved scrolls are, however, also not universal on Neapolitan instruments.

11.  The nameboard is ‘panelled’ in the usual Neapolitan style.  Normally North-Italian harpsichords and virginals have flat nameboards, whereas Neapolitan have a central recessed rectangular panel usually scratched out of the wood of the nameboard itself with a moulded scratch cutter.

12.  The nut, as indicated by the plan of the instrument scribed onto the baseboard (see Figure 12), was originally in two straight sections with a 'nick' between them at the note middle c1. The nick in the nut which is otherwise in two straight sections is a characteristic of all of the other Guarracino harpsichords with the plan scribed on the baseboard[9], and the ‘nick’ always occurs at the original note c1.

13.  The mouldings on the harpsichords signed or attributed to Guarracino show quite a range of sizes and shapes throughout the period of almost half a century during which Guarracino was active.  For example, the upper and lower outside case mouldings in the ‘Guarracino’ harpsichord in the Hague are notably different from the mouldings of the harpsichord described here as can be seen in Figure 14 below.  This moulding seems to be unique to this instrument and is, as far as I know not found on any other instrument whether by Guarracino or not.

Figure 14.wmf

Figure 14 – A schematic section of the spine side of the ‘Guarracino’ harpsichord in the Gemeentemuseum, The Hague, Cat. No. 1933.0543.  Note the shape of the upper and lower outside moulding.  The lower part of the lower moulding does occur on a number of instruments, but the top outer moulding and the top part of the lower moulding seem to be unique to that instrument. 

 

However, the mouldings on the harpsichord that is the subject of this report are closely similar to those found on the signed virginals by Guarracino from the period 1668 to 1679.

14.   Lastly, the most compelling reason for attributing this instrument to Guarracino is the use in its design and construction of the unique workshop unit of measurement used by Guarracino in this and in other instruments either signed or attributed to Guarracino.  This unit of measurement (see Table 4 above) seems to be characteristic of all of the signed instruments by Guarracino, or which I have attributed to him.  The only exception to this is the little rectangular octave virginal signed by Guarracino and dated 1694[29] (and therefore made right at the end of his career) which uses the ‘normal’ Neapolitan unit close to 21.83mm. 

 

         All of the unsigned instruments attributed by me to Guarracino, in addition to using this unit of measurement in their design and construction, also have many of the other characteristics of Guarracino listed above and can therefore be confidently attributed to him.  There is, at any rate, not a single contra-indication that this instrument nor any of the instruments under discussion here is not by Guarracino.  In addition I know of no other Neapolitan maker who used the characteristic and individual unit of measurement found here in the design of his instruments.  In the light of all of this evidence I therefore feel that there can be little doubt that this instrument, which would have lost its original signature on the top or bottom keylever when it lost its original keyboard and keylevers[30], is by Onofrio Guarracino and dates from sometime between 1651 and 1694.  Lacking any evidence to the contrary, I have tentatively dated it to the period about 1675.

 

 

FOOTNOTES:


[1] Edwin M. Ripin, ‘The surviving oeuvre of Girolamo Zenti’, Metropolitan Museum of Art Journal, 7 (1973) pp. 71-87 was the first to suggest the use of Italian case moulding profiles in the identification of Italian stringed keyboard instruments.  Ripin, however, like most later authors, never explicitly notes that it is an assumption that the mouldings on their instruments were made in their own workshops and not bought in from a common supplier.

[2] See my book, Ruckers.  A harpsichord and virginal building tradition (Cambridge: Cambridge University Press, 1990; digital reprint, Cambridge: Cambridge University press, 2008) p. 120 & 126.

[3] See Patrizio Barbieri, ‘Cembalaro, organaro, chitarraro e fabbricatore di corde armoniche nella “Polyanthea technica” di Pinaroli (1718-32).  Con notizie sui luitai e cembalari operanti a Roma’, Recercare, 1 (Lucca: Libreria Musicale Italiana Editrice, 1989), p. 135.

[4] See Giuliana Montenari, ‘Le spinette ovale e la collezione di strumenti a penna del Granprincipe Ferdinando de’ Medici/The oval spinets and Grand Prince Fercinando de’ Medici’s collection of quilled instruments’, Bartolomeo Cristofori.  La spinetta del 1690/The 1690 oval spinet, edited by Gabriele Rossi-Rognoni, (Florence: Sillabe for the Galleria dell’Accademia, 2002) p. 3.

[5] See Denzil Wraight, ‘The identification and authentication of Italian string keyboard instruments’, The Historical Harpsichord, Volume Three, Howard Schott editor, (Stuyvesant, N.Y.: Pendragon Press, 1992) pp. 59-161 and The stringing of Italian keyboard instruments c.1500 - 1650, Ph.D. dissertation, Queen’s University of Belfast, Part 1 and Part 2 (Ann Arbor, MI: UMI, UMI number 9735109, 1997).

[6] See John Henry van der Meer, , Alla ricerca dei suoni perduti - In the search for lost sounds  Arte e musica negli strumenti della collezione di Fernanda Giulini - Art and music in the instrument collection of Fernanda Giulini, edited by Fernanda Giulini (Briosco: Villa Medici Giulini, 2006) Cat. No. 1, pp 98-103.

[7] The piede was usually 18 once, the braccio mercantile was usually 36 once, the palmo or palmo architettonica was 12 once, the canna was 120 once, etc.  See the various centres in the Papal States listed in Giovanni Croci, Dizionario universale dei pesi e delle misure in uso presso gli antichi e moderni con ragguaglio ai pesi e misure del sistema metrico, (Milan: The Author, 1860), Ludovico Eusebio, Compendio di Metrologia Universale e Vocabolario Metrologico, (Turin:  Unione Tipografico Editrice Torinese, 1899; reprint by Bologna: Forni Editore, 1967), Luigi Pancaldi, Raccolta ridotta a dizionario di varie misure antiche e moderne coi loro rapporti alle misure metriche…, (Bologna: Sassi, 1847) and numerous other sources on Italian metrology including Martini and Doursther mentioned elesewhere.

[8] Attention was first drawn to many of these features by John Koster in an unpublished paper he delivered at the meeting of the American Musical Instrument Society in San Antonio, Texas in 1992.  This was then published by him in ‘1. Harpsichord.  Maker unknown, Italy (probably Naples), about 1550’, Keyboard Musical Instruments in the Boston Museum of Fine Arts, (Boston: Museum of Fine Arts, 1994) pp. 3-8.  He also reiterated many of these in:  ‘Keyboard Instruments traced back to 16th-century Naples’, The Shrine to Music Museum Newsletter, 23, No 1 (1995) pp. 1-3.

[9] Even if the nut is not original on these harpsichords, the drawing of the nut is usually scribed on the baseboard and shows this feature.

 

 

 

 

References: 

A.A. Stanley, Catalogue of the Stearns Collection of Musical Instruments, (Ann Arbor: University of Michigan, 1918).

F. J. Hirt, Meisterwerke des Klavierbaus. Geschichte der Saitenklaviere von 1440 bis 1880, (Olten: IM Urs Graf-Verlag, 1955) pp. 218-221.  This reference shows a photo of the Leipzig ‘Vincentius’ which may also be an instrument by Onofrio Guarracino.

Edwin M, ‘The single-strung Italian harpsichord’, Keyboard Instruments (Edinburgh: Edinburgh University Press, 1971, reprint New York: Dover Publications, 1977) footnote 1, p. 36.

John Barnes, Anonymous Italianate Harpsichord No. RCM 175. Unpublished report written for the Museum of Musical Instruments, Royal College of Music, London, (Edinburgh: Museum of Musical Instruments, Royal College of Music, London, 1974).

Grant O’Brien, Anonymous Single-manual Italian Harpsichord, RCM 175. Technical drawing with additional notes (Edinburgh: Museum of Musical Instruments, Royal College of Music, London, 1974).

Michael Thomas, ‘String gauges of old Italian harpsichords’, The Galpin Society Journal, 24 (1971) p. 69.

Friedemann Hellwig, ‘The single-strung Italian harpsichord’, Keyboard Instruments.  Studies in Keyboard Organology, 1500-1800, Edwin M. Ripin editor, (Edinburgh: Edinburgh University Press, 1971; reprint, New York: Dover Publications, 1977) p. 36, footnote 1.

Elizabeth Wells, The Royal College of Music Museum of Instruments: Guide to the collection (London: The Royal College of Music, 1984) p. 5.

Grant O’Brien, Ruckers.  A harpsichord and virginal building tradition (Cambridge: Cambridge University Press, 1990; digital reprint, Cambridge: Cambridge University press, 2008).

BBC, Early Music Special Issue, (London: BBC, 1994), p.18 & ill.

Edward Kottick & George Lucktenberg, Early Keyboard Instruments in European Museums (Bloomington and Indianapolis: Indiana University Press, 1997), pp. 235–6

Denzil Wraight, The Stringing of Italian Keyboard Instruments c. 1500–c. 1650, PhD dissertation, The Queen’s University of Belfast, 1997 (Ann Arbor, Michigan: UMI Dissertation Services, No. 9735109), v.1 p. 169; v.2 p. 506 & p. 526.

Grant O’Brien,  ‘Bartolomeo Cristofori/Giovanni Ferrini as restorers and re-builders.  A ‘Neapolitan’ connection in two Italian harpsichords in Britain’, paper presented at the Galpin Society Conference, Edinburgh, July, 1997, and to a conference in Bologna.  This paper has never been published but is given on my website:  http://www.claviantica.com/Publications_files/Cristofori_Naples.htm.

Francesco Nocerino, ‘Arte cembalaria a Napoli.  Documenti e notizie su costruttori e strumenti napoletani’, in Ricerche sul ‘600 Napoletano, Saggi e documenti 1996-1997, (Naples: Electa Napoli, 1998), pp. 85-109

Grant O’Brien, ‘The use of simple geometry and the local unit of measurement in the design of Italian stringed keyboard instruments: an aid to attribution and to organological analysis’, The Galpin Society Journal, 52 (1999), pp. 108-171.

John Budgen, William Debenham, Jenny Nex, Christopher Nobbs and Lance Whitehead, Royal College of Music Museum of Instruments, Catalogue Part II, Keyboard Instruments, edited by Elizabeth Wells, (London: Royal College of Music, 2000), Cat. No. 175, pp. 36-39.

Francesco Nocerino, ‘Napoli centro di produzione cembalaria alla luce delle recenti ricerche archivistiche’, Fonti d’archivio per la Storia della musica e dello spettacolo a Napoli tra XVI e XVIII secolo, (Naples: Editoriale Scientifico, 2001) pp. 205 - 225.


 

 

Appendix 1 - Measurement of the Original Scalings from the Drawing of RCM 175

 

         The original scalings of RCM 175 could be measured from the drawing made of the instrument by the author for the Royal College of Music in 1974 (see footnote 44).  This was done by transferring the position, marked on the baseboard, of the original nut onto the drawing.  The string lengths were then measured from the drawing and not from the instrument itself using the original positions of the nut and its pinning to the bridge pins.  In the treble the string lengths were corrected using the original bridge and bridge pin positions indicated by the clearly-visible marks on the soundboard indicating where the bridge was located before it was moved.  The string lengths of each note in the C/E to c3 enharmonic compass with 50 notes resulting from these measurements are given below along with the corresponding notes for the intermediate states 2 and 3.  The string lengths of the present compass with an altered nut position are given in Table 6.  All of these states had a 50-note compass.

 

Note Length   Note Length
 c3 134    d1 474
 b2 143    cT1 497.5
 bI2 150.5    c1 525
 a2 159    b 550.5
 gT2 167    bI 576.5
 g2 176.5    a 607
 fT2 185    gT 639.5
 f2 195    g 676
 e2 205.5    fT 715
 eI2 215    f 760
dT2 227    e 805
 d2 239    eI 857
 cT2 250.5    d 912
 c2 264    cT 973
 b1 277.5    c 1038
 bI1 292.5   B 1103
 a1 306.5   BI 1179
 aI1 321.5   A 1253
 gT1 337.5   GT 1285
 g1 354.5   E 1304
 fT1 371.5   G 1323
 f1 391   FT 1343
 e1 410   D 1362
 eI1 429.5   F 1380
 dT1 452.5   C 1398

 

Table 12 - String-length measurements made from the 1974 RCM drawing used to reconstruct the original state and each of the subsequent states discussed in this report.  The split enharmonic notes and the split notes of the bass broken octave (5 extra notes) of the original state are indicated in bold.

         These string lengths are plotted for each c and f note as well as for GT and A on either side of the bass bridge mitre in the graph shown in Figure 23.

 

 

Appendix 2 - An Outline of the Historical Musical States of RCM 175

         The analysis used to arrive at the history of this instrument is based upon an examination of the keyframe, keyplates, jacks, registers, the lettering of the key and jack numbers, the pinning of the bridge, nut and hitchpins, alterations to the bridge position, wrestplank and upper bellyrail, and construction marks on various parts of the instrument.  The outline below does not necessarily present all of the evidence for each of these states.

 

State 1:  As determined here, the original enharmonic compass of this harpsichord was 4 octaves C/E to c3 50 notes with 27 naturals and with a broken short octave in the bass.  It had enharmonic split notes dT1/eI1 and gT1/aI1 in the third octave and only dT2/eI2 in the top octave.  There was only one choir of strings and one set of jacks and, in order for the keyboard to be centred in the keywell with keyblocks of equal width, the jacks must have plucked towards the left as is normal Italian and Neapolitan practice in single-register instruments and universal among the other instruments of Guarracino with only one register.

         The position of the nut was scribed on the baseboard below the keyboard and this and the punched mark and the ‘nick’ in the nut indicating the position of the note middle c1 has enabled the original scalings to be determined (see Appendix 1).  The original scalings for the top strings were estimated in this table using the position of the original nut scribed on the baseboard in conjunction with the marks on the soundboard which indicate the original position of the treble section of the bridge.  The estimated c2 scaling of the original state is 264mm, and these scalings, plotted in the graph of Figure 23, are Pythagorean and halve uniformly with each octave rise in pitch.  These scalings, in terms of the unit of measurement being used by Guarracino, seem clearly to be based on 12 once = 1 palmo = 259.4mm.

 

State 2:  Here the instrument was left still with one choir of strings and one register, but the compass was increased to C/E to f3 with 30 naturals and still with 50 notes.  It seems likely that this would have been a normal bass short-octave chromatic compass. In carrying out this procedure the keyboard must have been re-made and was moved one note to the left by a distance of 16mm.  The keyblocks were narrowed and new keyplates were made for c3, d3 and e3, and the original c3 keyplate became the new f3.  There is no way of telling in which direction the jacks plucked at this stage although the simplest and most logical way would have been for them still to pluck to the left as in the original state.  If the direction in which the jacks plucked remained the same, then the length of c2 would have been 321½ mm.  However, because the original compass was enharmonic and this new compass was chromatic the new scalings would have been strongly non-Pythagorean (to the extent that they are also non-Pythagorean with the present compass) so that the length of c2 does not give a clear idea of the pitch of the instrument in this state.  Assuming that the treble section of the bridge had not been moved at this stage, the new top note f3 would have a string that was critically stressed and would have had a c3-equivalent scaling of  = 357.3mm.  If used at a pitch of about a1 = 415.3 Hz this would have required iron stringing at least for a part of the treble compass. 

         In this state the boxslide register would not have required any alteration because this state with an increased chromatic compass had the same number of notes as the original 4-octave enharmonic compass with a broken bass octave.

 

State 3:  In this state the major alteration was to change the compass to C,D to d3, the same as at present, but still leaving only one register of jacks.  The 16mm wide strip added to the keyframe in state 2 was removed and a further 40mm cut from the keyframe.  A piece of chestnut was added to the treble as a substitute for the removed portion of frame, and this still remains with a rebate for the balance rail in the forward position predating the present position of the balance rail slightly further back.  The total sideways movement of the keyboard of 56mm gives room for two extra naturals in the bass and two fewer in the treble.  Here the bottom full-width E keyplate was moved down to become the bottom full-width C keyplate.  The top f3 of State 2 (the original e3) was moved down to become the D keyplate with a cut-out in its rear right edge for the EI keylever.  The top e3 keyplate became E and the eI3 accidental became EI.  The space left in the top d3 where the eI3 used to be had to be filled in, and this is still visible in the present keyboard.

         It is assumed that the plucking direction of the jacks in this state was still towards the left.  The c2-equivalent scaling of the top note with the bridge in its original position would have been 134mm, the same length as in the previous and original states (see Appendix 1).  The top d3 string would then have had a c2-equivalent scaling of 301½mm.   The scalings in this state, like those in State 2, are however strongly non-Pythagorean.  The length of c2 itself is only 277½mm, a length suitable for brass stringing at a pitch near modern concert pitch.  However, because of the non-Pythagorean nature of the scalings, the strings of the top notes would certainly break at this pitch.  The problem could be avoided by 1. tuning the whole instrument to a lower pitch;  2. stringing the top notes with iron wire until a point is reached where brass will still hold pitch without breaking; or 3. moving the top treble section of the bridge.  The later solution was certainly a feature of State 4 below, and may well have been introduced in this state although there now seems no proof to indicate at which point in the instrument’s history the bridge was moved.

 

State 4:  The compass and arrangement of the keyplates relative to State 3 was unaltered, but an additional register and choir of strings was added.  To do this an extra set of hitch and bridge pins was added, the gap was widened by cutting 12mm from the soundboard and upper belly rail, and by removing the wrestplank and replacing it with one 6 mm narrower.  A new nut was made (see Figure 28 below) and positioned in a new location somewhat nearer to the gap than the original nut.  The treble section of the bridge was also moved nearer the gap (this may of course have been done in either States 2 or 3).  Two rows of tuning pin holes were drilled in the new wrestplank, a new row of jacks must have been made, and the original boxslide was sliced into sections from which the present upper and lower guides are made.  It is likely that the two identical rows of tuning pins date from this alteration although it is possible that the stock from which the second row of tuning pins was made was of the same diameter as that of the original pins and that the style of the old pins was copied when make the new pins.  The diapason rack in the keyframe was replaced with the new system of guide pegs between the keylevers since there was now no space at the rear of the keyboard for a diapason rack.  The old keylevers were replaced with new ones of chestnut but the old keyplates were transferred to the new keylevers and a new set of arcades was made.  The balance rail was moved further to the rear of the keyframe so that the balance of the keys with two registers of jacks instead of one would give a smaller lever ratio making the keys easier to play[63].

         The replacement of the nut in a new position closer to the gap, and the movement of the treble section of the bridge, both contributed to the shortening of the scale in this state.  The long string for c2 has a scaling of 255mm, and the long g2 string (174mm long), which is now the most critically stressed string, has a pitch scaling of 261mm.  These scalings would be suitable for brass stringing at a pitch close to modern pitch a1 = 440 Hz.

         There is a number of features of the re-building of this instrument that suggests that the work was done by Bartolomeo Cristofori or by his pupil Giovanni Ferrini[64].  The system of guiding the keys at the rear with wooden pegs is the same as that of the harpsichord-piano combination by Giovanni Ferrini in the Collection of Luigi Ferdinando Tagliavini, Bologna[65].  Cristofori and Ferrini and some of their Florentine contemporaries are among the few harpsichord builders working in the Italian tradition who used a system of upper and lower guides instead of boxslides.  Hence the existence of upper and lower guides in this instrument is at least a signal that the instrument was re-worked in Florence.  The moulding on the edges of the upper guides is the same as that on the soundboard mouldings of instruments by Cristofori in Leipzig (Nos. 84, 85 and 86[66]), and the Stearns Collection harpsichord[67] and a cembalo traverso belonging to Dom Pineschi, Pistoia, both by Ferrini.  The moulding on the top of the nut, and the bevelled cutting of the ends of the nut and the moved treble section of bridge are the same as that on Leipzig Nos. 84 and 85 by Cristofori, and as the Stearns Collection harpsichord and Pistoia cembalo traverso by Ferrini.  Also the moulding of the key arcades and the construction and placing of the positioning blocks on the ends of the registers are the same as on most of the above-mentioned instruments by Cristofori or Ferrini.  It seems likely therefore that the alteration of State 4 is by either Cristofori or Ferrini, whose work I have not yet been able to distinguish, and that it dates from the first half of the eighteenth century.

 

Figure 28 - The new nut (left) and the original bridge (right).  Scale 1:1.

Appendix 3 - Biography of Onofrio Guarracino (b., Naples, January, 1628 – d, Naples, post 1698).

         Onofrio Guarracino is today reasonably well known as a maker of rectangular virginals but he also made at least two bentside spinets[68] and a number of harpsichords[69].  The only harpsichord which still retains his signature is the instrument dated 1651 which is in a private collection in Rome.  However, on the basis of many of the features that they have in common with the other signed instruments by Guarracino, a number of unsigned harpsichords can also be attributed to him (see Table 4).  Among these are the harpsichords in the Gemeentemuseum, The Hague (Cat. No. 1933.0543), in the Giulini Collection in Milan (Cat. No. 3[70]), in the Metropolitan Museum of Art, New York (Inv. No. 45.41), in the Boston Museum of Fine Arts (Cat. No. 1986.518), in the Museum für Kunst und Gewerbe, Hamburg (Beurmann Collection, Cat. No. 8) and in the Musikinstrumentenmuseum, Berlin (Cat. No. 4650) .  Other harpsichords which may also be by Guarracino are in private ownership in Buenos Aires, in the Palazzo Gritti, Venice and in the Händelhaus, Halle in Germany (Inv. No. MS-69) although I have not had the opportunity to make a detailed examination of the latter instruments.  Although several of his virginals are in the Museo Nazionale degli Strumenti Musicali in Rome, it would appear that none of the harpsichords in the catalogue of this museum[71] is by Guarracino even though many of them are certainly Neapolitan in origin.  However, a harpsichord which is in the reserve collection and not on public view, has recently been discovered[72] that is certainly by Guarracino.

         There are many documentary references to Guarracino and much of the archival information about him is known thanks to the hard work and diligent research of Francesco Nocerino[73].  According to Nocerino, even in his own time Guarracino was the most important and respected Neapolitan harpsichord builder in the city[74].  Probably as a result of this high esteem by his contemporaries he was often called upon to give valuations of instruments and estimates of work, and many of the archival references deal with this aspect of his activity.  The archives show, for example, that he made a valuation of a mother and child virginal[75] made by one Giovanni Rispolo.  The document relating to the evaluation shows for the first time that such instruments were made in Naples, and not just in Antwerp[76] and England[77], as part of a harpsichord builder’s normal activity.  This further suggests that, since he knew of their existence, Guarracino may well also have made them himself, although there is no known extant example of this aspect of his work.  It is also known that he was involved in the construction and sale of the elusive tiorbino, a gut-strung double-strung keyboard instrument[78].

         Francesco Nocerino has been able to find the certificate of baptism of Guarracino[79] in the Neapolitan archives.  According to this, Honofrio Antonio Guarracino, son of Fabritio Guarracino and Anna d’Accetto, was baptised in the Church of San Gioseppo Maggiore in Naples on the 4th of January, 1628.  He was therefore probably born only a few days before this.  This means that he was probably born after 1 January, 1628, although his birth may have take place in the dying days of 1627.  His earliest known instrument is the only surviving signed harpsichord which is dated 1651, and was therefore made when he was 24 years old.  Guarracino was married in 1654, aged 27, to Ursula Perrone, daughter of Giuseppe Perrone.  On the marriage certificate he states his occupation to be cimbarairo[sic] = cembalaro or harpsichord builder.  The marriage certificate also makes clear that Guarracino’s brothers-in-law Aniello and Michele Perrone were both wood carvers.  One of the features of Guarracino’s harpsichords is their elaborately-carved keyscrolls.  Sometimes these take the form of elaborate gilt vine scrollwork, sometimes they are carved in plain wood and are of human or mythological figures, or of cornucopiæ.  The family relationship between Guarracino and Aniello and Michele Perrone strongly suggests that one or both of them may have been responsible for carving the keywell scrolls on his instruments (see Figure 5 in the colour section on page xxx).   Although the sons of many of Guarracino’s harpsichord-building contemporaries went on also to become harpsichord makers, none of Guarracino’s 8 children continued the tradition built up by their father.  One of Guarracino’s sons Francesco, however, is known to have gone on to become a Prorationale della Regia Camera (Administrative Officer of the Royal Chamber).  Although the last signed and dated instrument by Guarracino was made in 1694 he is known from the archives through the dealings that he made to have survived at least until 10 January, 1698, by which time he would have been 70 years old.   The date of his death is not known, but by 1711 he was declared already to have died by his son Francesco Guarracino.  His surviving instruments cover dates during the period between 1651 to 1694.

         According to the research of Francesco Nocerino, Guarracino worked in Naples in the Strada del Spirito Santo, near to the Banco del Spirito Santo in the archives of which much of Nocerino’s work has concentrated.  At first Guarracino rented a workshop there which was physically connected to the house where he lived.  Then, with the passage of time accompanied by his increasing wealth and activity, he was able to rent three apartments in a palazzo with various annexes, a courtyard and a number of workshops.  He held accounts in numerous banks throughout Naples and there were regular and consistent movements of funds in and out of these accounts.  These included payments for the rent of the buildings he used, and payments for paintings, furniture, glass, instruments, and in respect of the marriage of his daughters.  He made numerous loans to persons of all social and economic levels including his own relatives and family.  He was also often involved with dealings in used instruments and their repair and resale.  Guarracino was obviously a natural survivor:  as Francesco Nocerino has pointed out[80] he lived through numerous plagues in Naples, an eruption of Vesuvius, a famine and a major earthquake.  He lived through a period of enormous political and economic upheaval and of considerable human adversity.  He survived and flourished despite competition and the lower prices at which others offered their instruments.  He seems to have enjoyed a high standard of living and a high level of respect from his friends, family and professional contemporaries throughout his working career.

         Like makers working throughout Europe during the whole of the historical period, Guarracino based the design and construction of his instruments on his local unit of measurement.  Each city and town had its own unit of measurement and the local standard was usual displayed for all to see and use either inside or on an outside wall of the town hall.  Carpenters, joiners, land surveyors, cloth merchants, etc would have made up their own rulers from this standard by copying its length onto a piece of wood or metal and dividing up its length into any sub-units of the standard using some geometrical method.  Thus each harpsichord maker working in Naples probably also made up their own rulers, possibly as a part of their apprenticeship training, from the town standard and, because of the slight error in transferring the measurement of the standard to their own workshop rulers, there is a kind of individual workshop unit for each maker which is very close, but not exactly equal to the standard held by the town officials.  In the case of Guarracino this unit of measurement was 21.622mm rather than the ‘textbook’ unit found in books on metrology of 21.834mm (see footnote 16).  The difference between these two is easily distinguishable in the instruments of Guarracino when calculating the unit of measurement he used using either the case measurements of an instrument or its lateral string/bridgepin/nut pin spacing.  The use in an instruments of this unique and characteristic unit of measurement is one of the most compelling arguments for attributing such an instrument to him.  I know of no other maker that uses this unit of measurement in the design and construction of their instruments.  It seems likely that, at the end of his career around 1694, Guarracino started to use a unit of measurement closer to the textbook standard value when, for example, he built the little octave virginal in the National Music Museum in Vermillion, South Dakota (see footnote 29).  Obviously instruments built from this date or later will probably not use the characteristic 21.622mm unit.

         Guarracino did not seem to sign his name on the nameboard of any of his instruments as is more common with some other harpsichord and virginal makers.  Instead, if at all, he seems to have signed only the top surface of either the top or the bottom keylevers.  Virginals, by virtue of their complicated geometry, are difficult to alter by increasing their compass and number of notes.  They therefore normally retain their original keyboards and, in the case of Guarracino, therefore also retain the signature on the keylevers.  However, harpsichords are frequently subjected to alterations, the most common being the alteration of a C/E to c3 short-octave compass to a C to c3 chromatic compass, usually with jacks and strings added at both ends of the keyboard to squeeze in the extra notes.  Also, as seen here, Guarracino seems often to have made instruments with enharmonic keyboards, often with unusual overall compasses.  After the general introduction of sixth-comma temperaments, such enharmonic compasses became redundant.  Hence these instruments were also often given new keyboards to bring them up to date with the modern literature and with modern tuning systems.  Therefore in many of the instruments that I have ascribed to Guarracino the keyboards and signed keylevers have disappeared in the process of up-dating the instruments in conformity with later musical needs and practices.

         Thus in all of the harpsichords that have been ascribed to him here in which the keylevers have been replaced, the signature has disappeared in the process of making the new keyboards.  Before 1677[81] Guarracino signs himself on his instruments using the normal Italian form “Onofrio Guarracino”, but after 1677 he uses the more elegant and sophisticated form “Honofrio” (Honofrius in Latin) for his first name.  It would seem that by 1677 Guarracino’s social and economic position in Neapolitan society was such that he wanted to reflect this wealth and social situation in the form in which he signed his instruments.  In the archival sources the form in which his name would be written would be outwith the control of Guarracino himself except when he actually signs his name himself.  It is not known if Guarracino’s signature in the archival sources follows the same tendency. 

 

                                                                            - © Dr Grant O’Brien, Edinburgh, 2008

      

FOOTNOTES:

[1] See John Budgen, William Debenham, Jenny Nex, Christopher Nobbs and Lance Whitehead, Royal College of Music Museum of Instruments, Catalogue Part II, Keyboard Instruments, edited by Elizabeth Wells, (London: Royal College of Music, 2000), Cat. No. 175, pp. 36-39.

[2] This stand was made from parts supplied by the author of this report.  This stand is in a typically Venetian style, and is not typical either of the simple nor of the elaborate Neapolitan stands which survive under numerous Neapolitan harpsichords and virginals.  Unfortunately at the time that this stand was made for the instrument, it was not known that the harpsichord was Neapolitan in origin.

[3] Details of the attribution of the final state of the instrument were made in a paper presented by me at the Galpin Society Conference, Edinburgh, July, 1997:  ‘Bartolomeo Cristofori/Giovanni Ferrini as restorers and re-builders.  A ‘Neapolitan’ connection in two Italian harpsichords in Britain’, This paper has never been published but is given on my website:  http://www.claviantica.com/Publications_files/Cristofori_Naples.htm.

[4] All photographs and drawings are reproduced here by kind permission of the curatorial staff of the Royal College of Music, London.

[5] I am grateful to Francesco Nocerino for making this suggestion.

[6] See Hubert Henkel, Kielinstrumente.  Katalog des Musikinstrumentenmuseums der Karl-Marx Universität Leipzig, Vol. 2 (Leipzig: VEB Deutscher Verlag für Musik, 1979) Cat. No. 69, pp. 59-61, Plates 24 and 25 near the end of the volume.

[7] The final confirmation of the authorship of Leipzig 69 will be a determination of the unit of measurement used in its design and construction to see if it is the same as that used in the other instruments by Guarracino.  The author plans to carry out the analysis of the unit of measurement used in Leipzig 69 in the near future.

[8] Practical experience gained by the author and his colleague Graziano Bandini working in Castel San Pietro Terme near Bologna has shown that both chestnut and cypress are easily bent after being subject to the application of heat.

[9] The author has used this system of case construction himself in a number of Guarracino-style harpsichords (see Figure 9 above).  Further details of the construction of these instruments can be seen on his website:  http://www.claviantica.com/Design_files/Construction_particulars_files/Case_side_construction.htm.

[10]  The top moulding adds an additional 3½mm.

[11] The addition of this small piece of bridge was necessary when the disposition was changed from 1x8' to 2x8'.  Because the second string was added on the right it was necessary to extend the bridge in the treble but not in the bass.

[12] The pitch of this note is incorrectly given on the drawing of the instrument made by me in 1974 as a cT1 and not c1, and is the first indication that the interpretation of the original compass of the instrument of C/E to d3,e3 without a top eI3 by John Barnes in his restoration report is incorrect.  See footnote 34.

[13] See my paper, ‘The use of simple geometry and the local unit of measurement in the design of Italian stringed keyboard instruments:  an aid to attribution and to organological analysis’, The Galpin Society Journal, 52 (1999) pp. 108-171.

[14] It seems clear that the baseboard has shrunk slightly near the tail where it is not prevented from doing so by any transverse bracing.  It is estimated therefore that the tail of the baseboard matched the wood of the tail itself and was about 2mm longer than it is at present.

[15] Here it is assumed that the value of the oncia lies within the range of about 18mm to 60mm, the range covered by the units used historically in the Italian peninsula.  If values of the oncia smaller than 18mm are allowed, then numerous other lengths of the oncia are possible.

[16] See: Horace Doursther, Dictionnaire universel des poids et mesures anciens et modernes, (Brussels: M Hayer, 1840; reprint  Amsterdam: Meridian Publishing Co, 1965) p. 375 and Hercule Cavalli, Tableaux comparatifs des mesures, poids et monnaies modernes et anciens…, (Paris: Paul Dupont, 2/1874).  Both of these authors give the same value for the Neapolitan oncia

[17] Angelo Martini, Manuale di metrologia ossia misure, pesi e monete, (Turin Rome and Florence: Ermanno Loescher, 1883; reprint Rome: Editrice Edizioni Romane d’Arte, , 1976) p. 394.

[18] The value of the error given here is the value of the normal standard deviation error as determined by the regression analysis.

[19] Andreas Beurmann, Historische Tasteninstrumente.  Cembali Spinette Virginale Clavichorde, (Munich, London and New York: Prestel Verlag, 2000) Cat. No. 8, pp. 40-43.  The ascription to Celestini (whose name appears on the nameboard) is clearly a forgery.

[20] Edwin M. Ripin, ‘The surviving oeuvre of Girolamo Zenti’, Metropolitan Museum of Art Journal, 7 (1973) pp. 71-87 was the first to suggest the use of Italian case moulding profiles in the identification of Italian stringed keyboard instruments.  Ripin, however, like most later authors, never explicitly notes that it is an assumption that the mouldings on their instruments were made in their own workshops and not bought in from a common supplier.

[21] See my book, Ruckers.  A harpsichord and virginal building tradition (Cambridge: Cambridge University Press, 1990; digital reprint, Cambridge: Cambridge University press, 2008) p. 120 & 126.

[22] See Patrizio Barbieri, ‘Cembalaro, organaro, chitarraro e fabbricatore di corde armoniche nella “Polyanthea technica” di Pinaroli (1718-32).  Con notizie sui luitai e cembalari operanti a Roma’, Recercare, 1 (Lucca: Libreria Musicale Italiana Editrice, 1989), p. 135.

[23] See Giuliana Montenari, ‘Le spinette ovale e la collezione di strumenti a penna del Granprincipe Ferdinando de’ Medici/The oval spinets and Grand Prince Fercinando de’ Medici’s collection of quilled instruments’, Bartolomeo Cristofori.  La spinetta del 1690/The 1690 oval spinet, edited by Gabriele Rossi-Rognoni, (Florence: Sillabe for the Galleria dell’Accademia, 2002) p. 3.

[24] See Denzil Wraight, ‘The identification and authentication of Italian string keyboard instruments’, The Historical Harpsichord, Volume Three, Howard Schott editor, (Stuyvesant, N.Y.: Pendragon Press, 1992) pp. 59-161 and The stringing of Italian keyboard instruments c.1500 - 1650, Ph.D. dissertation, Queen’s University of Belfast, Part 1 and Part 2 (Ann Arbor, MI: UMI, UMI number 9735109, 1997).

[25] See John Henry van der Meer, , Alla ricerca dei suoni perduti - In the search for lost sounds  Arte e musica negli strumenti della collezione di Fernanda Giulini - Art and music in the instrument collection of Fernanda Giulini, edited by Fernanda Giulini (Briosco: Villa Medici Giulini, 2006) Cat. No. 1, pp 98-103.

[26] The piede was usually 18 once, the braccio mercantile was usually 36 once, the palmo or palmo architettonica was 12 once, the canna was 120 once, etc.  See the various centres in the Papal States listed in Giovanni Croci, Dizionario universale dei pesi e delle misure in uso presso gli antichi e moderni con ragguaglio ai pesi e misure del sistema metrico, (Milan: The Author, 1860), Ludovico Eusebio, Compendio di Metrologia Universale e Vocabolario Metrologico, (Turin:  Unione Tipografico Editrice Torinese, 1899; reprint by Bologna: Forni Editore, 1967), Luigi Pancaldi, Raccolta ridotta a dizionario di varie misure antiche e moderne coi loro rapporti alle misure metriche…, (Bologna: Sassi, 1847) and numerous other sources on Italian metrology including Martini and Doursther mentioned elesewhere.

[27] Attention was first drawn to many of these features by John Koster in an unpublished paper he delivered at the meeting of the American Musical Instrument Society in San Antonio, Texas in 1992.  This was then published by him in ‘1. Harpsichord.  Maker unknown, Italy (probably Naples), about 1550’, Keyboard Musical Instruments in the Boston Museum of Fine Arts, (Boston: Museum of Fine Arts, 1994) pp. 3-8.  He also reiterated many of these in:  ‘Keyboard Instruments traced back to 16th-century Naples’, The Shrine to Music Museum Newsletter, 23, No 1 (1995) pp. 1-3.

[28] Even if the nut is not original on these harpsichords, the drawing of the nut is usually scribed on the baseboard and shows this feature.

[29] This instrument is in the National Music Museum, Vermillion, South Dakota, Cat. No. NMM 6041

[30] As explained near the end of the biography of Onofrio Guarracino at the end of this paper, Guarracino’s normal practice was to sign the top or bottom keylever and not to place his signature on the nameboard as is somewhat more common.  The fact that the keylevers are not original means that, like a number of other Guarracino instruments where this has also happened, the original signature was lost when the keylevers were replaced.

[31] By ‘natural balance’ I mean that if the natural keylever is balanced on a rounded fulcrum, this is the point where the keylever balances front to back relative to the balance-pin holes.  Normally, even though the balance pins of the sharps is placed well behind the balance pins of the naturals, the natural balance of the sharps is much further behind the balance pin hole than it is for the naturals because of the light weight of the front part of the sharp keylevers compared to the naturals.  So normally only the natural balance of the natural keys is relevant.  The natural balance is clearly affected by any carving away of wood on the keylever in order to lighten (or otherwise) the touch.

[32] The complete description of the intermediate states is to be found in the detailed restoration report by John Barnes held by the Royal College of Music.

[33] It should also be noted that generally the balance ratio of the keylevers of early instruments is much more ‘unfavourable’ than in later instruments.  This is probably the result of a general change in keyboard technique which seems to have occurred throughout most of Europe.  The movement of the balance rail away from the player was therefore probably partially intended to compensate for the addition of an extra row of jacks at the ends of the keylevers, but can also be seen as a general trend towards the lightening of the touch in the eighteenth century.

[34] John Barnes, Anonymous Italian Harpsichord No. RCM 175 (Edinburgh: Unpublished restoration report to the Royal College of Music, March, 1974).

[35] See footnote 3.  The instrument is Russell Collection Cat. No. HS1-A1620.2.

[36] The position both of the bridge pins in Table 11 and those in Table 5 below were measured using an ordinary tape measure.  The estimated error in each measurement was estimated to be ±0.1mm, and the measurements were reproducible to within this accuracy.

[37] This error includes, in part, the error made by the author in the actual measurements of the positions of the jackslots.  Hence the error given here is, effectively, the total of the author’s measurement error plus Guarracino’s error in positioning and cutting the jackslots.  Hence Guarracino’s error is even less than this number.

[38] In terms of normal experimental error, this value is extremely small.

[39] John Henry van der Meer, Alla ricerca dei suoni perduti - In the search for lost sounds  Arte e musica negli strumenti della collezione di Fernanda Giulini - Art and music in the instrument collection of Fernanda Giulini, edited by Fernanda Giulini (Briosco: Villa Medici Giulini, 2006) Cat. No. 3, pp 110-117.

[40] There is a fairly substantial number of other instruments with a disposition of 2x8' made both in Naples and elsewhere in Italy in which it is clear that the ‘P’ = Principale register was placed at the rear of the gap and that its jacks plucked the long strings on the left with the ‘S’ = Secondo register plucking the short strings on the right.  However, the process of determining this is usually possible only when a large portion of the remaining jacks is original.  The jacks and registers are, of course, hand made and normally each jack is made individually to fit its particular register slot.  In order to determine how the jacks fit the registers and their plucking direction, it is necessary to try each of the jacks in its correct position in both of the registers with both plucking directions.  Usually the process of trying the jacks in their slots results in there being a unique row and plucking direction for the jacks since otherwise they wouldn’t fit in their correct slot.  As mentioned here, because of the mouldings on the registers and because the plucking directions are actually marked on the registers of RCM 175, there is no ambiguity in the determination of either the order or the plucking direction of the jacks.  The result is to confirm the ‘classic’ disposition of Italian instruments with two registers as given above.

[41] I would like to express my thanks to the owner for his kindness and generosity in allowing me to examine, measure and photograph his harpsichord.

[42] See footnote 34.

[43] Because the alteration was carried out in Florence it is appropriate to use the Florentine unit of measurement which would  have been that used in the Cristofori/Ferrini workshops.  The Florentine soldo was divided into 12 denari.  Four denari, or one third of a soldo was called a quattrino and this division into thirds is the most common sub-division of the soldo.  To us this division seems a bit strange, but it was completely normal for a Florentine working in the seventeenth and eighteenth centuries.

[44] Grant O’Brien, Anonymous Single-manual Italian Harpsichord, RCM 175. Technical drawing with additional notes (Edinburgh: Museum of Musical Instruments, Royal College of Music, London, 1974).

[45] The 1651 Guarracino harpsichord in private ownership in Rome, the only signed and dated harpsichord by Guarracino, leaves 35½mm between the bottom bridge pin and the spine and 44mm (= close to 2 once) between the top bridge pin and the cheek.

[46] John Barnes noted that this pin was not original, but then assumed, reasonably perhaps, that there was never an original pin in the same hole and at the same position.  Why the original pin disappeared from the bridge is not now clear, but if there wasn’t originally a pin at this point, then there would have been an unused portion of the tail mitre of the bridge that was 43mm long.  This is atypically long for Guarracino and for Italian harpsichords in general.  If the original pin was in the same position as the present long C pin, then it is only 18mm from the end - a distance that is much more reasonable and in keeping with Guarracino’s normal practice.

[47] Numerous other harpsichords and virginals have enharmonic compasses which are notably either lacking any enharmonic split notes in the top octave or they have fewer enharmonic notes than the octave immediately below the top octave.

[48] See footnote 39.

[49] From the style of the registers and sharps it seems likely that the alteration to the instrument was carried out by Ignazio Mucciardi in Naples around the end of the eighteenth century.  See Maria Luisa Cervelli, La Galleria Armonica, (Rome: Istituto Poligrafico e Zecca dello Stato, 1994), No 760, p. 279, Plate p. 293.

[50] Private communication.

[51] Private communication from Francesco Nocerino.

[52] See footnote 19.

[53] Even John Barnes, in his 1974 restoration report, admits that this is a highly unusual compass.  See footnote 34.

[54] See my book Ruckers.  A harpsichord and virginal building tradition (Cambridge: Cambridge University Press, 1990; digital reprint, Cambridge: Cambridge University press, 2008) pp. 223-4.

[55] The error in the measurement and positioning of the strings is much greater than that in the measurement and positioning of the jackslots.  This manifests itself as a noticeable scatter in the points in Figure 27 compared to that in Figure 17.  Also the standard error as given by the regression analysis is also greater for the string lateral spacing than the jackslot lateral spacing as seen in Table 4.  The total combined error in the determinations of the unit of measurement calculated using the regression analysis for these two spacings is 0.03mm whereas they, in fact, differ by only 0.007mm.  They are therefore said to agree with one another to within the calculated experimental error.

[56] See footnote 1.

[57] Friedemann Hellwig, ‘The single-strung Italian harpsichord’, Keyboard Instruments, Edwin M. Ripin editor, (Edinburgh: Edinburgh University Press, 1971; reprint, Dover Publications, New York, 1977).  Of the 8 instruments listed by Hellwig, only three can be definitively assigned to the sixteenth century.  The others may all be Neapolitan instruments, most of which are of the seventeenth-century.  Hellwig mentions RCM 175 as a single-strung harpsichord in footnote 1.

[58] Maria Luisa Cervelli, La Galleria Armonica, (Rome: Istituto Poligrafico e Zecca dello Stato, 1994) No 779, p. 305, Plate page 317.  Work by me on the design of this harpsichord and the unit of measurement used in its construction indicates that it is also of Neapolitan origin, although it is not by Guarracino.   The date 1630 on the top surface of the bottom keylever along with the initials “GA” does, however, appear to be original and authentic.

[59] This instrument is the single-manual harpsichord Cat. No. 779 with an enharmonic compass of C/E to c3 with split dT/eI, aT/bI, dT1/eI1, gT1/aI1, aT1/bI1, dT2/eI2, 51 notes.  The instrument is signed only “G. A.” and is dated 1630.  See Maria Louisa Cervelli, La Galleria Armonica, (Rome: Istituto Poligrafico e Zecca dello Stato, 1994) Cat No 779, pp. 305 and 317.

[60] For an explanation of this method see my article in The Galpin Society Journal, 52 (1999) pp. 108-171 or my website:  http://www.claviantica.com/Geometry_ital.htm.

[61] John Koster has already noted that Neapolitan harpsichords were the preferred instruments in the Medici court for much of the seventeenth century.  See his article ‘The early Neapolitan school of harpsichord building’, forthcoming in Luisa Morales, ed., Domenico Scarlatti en Espana / Domenico Scarlatti in Spain (Proceedings of the FIMTE Symposia 2006-2007).  See also Hammond, “Musical Instruments at the Medici Court in the Mid-Seventeenth Century,” in Friedrich Lipmann et al., eds. Studien zur italienisch-deutschen Musikgeschichte 10 – Analecta Musicologica 15 (Cologne: Arno Volk Verlag Hans Gerig, 1975), pp. 202-219.

[62] However, see Christopher Stembridge, ‘Music for Cimbalo cromatico and other split-keyed instruments in seventeenth-century Italy’, Performance Practice Review, 5 (1992) pp. 5-43, and ibid, ‘The Cimbalo cromatico and other Italian keyboard instruments with nineteen or more divisions to the octave (Surviving specimens and documentary evidence)’, Performance Practice Review, 6, Nº 1 (Spring, 1993), pp. 33-59, and Denzil Wraight and Christopher Stembridge, ‘Italian Split-Keyed Instruments with Fewer than Nineteen Divisions to the Octave’, Performance Practice Review, 7, no. 2 (1994), pp. 150-181. 

[63] See footnote 33.

[64] See footnote 3.

[65] See Luigi Ferdinando Tagliavini and John Henry van der Meer, Clavicembali e spinette dal XVI al XIX secolo - Collezione L.F. Tagliavini, (Bologna: Cassa di Risparmio in Bologna, 1986; 2/1987) Cat. No. 16, pp. 186-200.

[66] See Hubert Henkel, Kielinstrumente.  Katalog des Musikinstrumentenmuseums der Karl-Marx Universität Leipzig, Vol. 2 (Leipzig: VEB Deutscher Verlag für Musik, , 1979).

[67] This harpsichord is Catalogue No. 1332 in the Stearns Collection, University of Michigan, Ann Arbor, Michigan.

[68] A spinet in the Museo Nazionale with the catalogue number 898 and thought to be Guarracino is definitely not by him.  The two genuine bentside spinets known definitely to be by Guarracino are the instruments dated 1688 in the Museo Nazionale degli Strumenti Musicali, Rome, Nº 888 and another in the Convento delle Suore di Santa Francesca in Naples.

[69] See Donald H Boalch, Makers of the Harpsichord and Clavichord, 1440-1840, 3rd edition by Charles Mould, (Oxford: Clarendon Press, 1995) pp. 75 and 343-46.  A further list of harpsichords which have been attributed by me to Guarracino are given in Table 4 and Table 8.

[70] See footnote 39.

[71] See, for example, the harpsichord mentioned in footnote 58. 

[72] Personal communication to me by Andrea di Maio.  The instrument, which has the old inventory number 1827, is now missing its keyboards.  But it originally had a 45-note compass which almost certainly was the standard 45-note C/E to c3 compass.  It is, typically, unsigned and undated on the nameboard and, because it has lost its keyboards, has therefore lost the only trace of Guarracino’s signature which would have been on the top or bottom keylever.  Further details of this instrument with an analysis of the unit of construction used to design and build it will be published by me in a forthcoming study just of the harpsichords by Guarracino.

[73] See Francesco Nocerino, ‘Arte cembalaria a Napoli.  Documenti e notizie su costruttori e strumenti napoletani’, in Ricerche sul ‘600 Napoletano, Saggi e documenti 1996-1997, (Naples: Electa Napoli, 1998), pp. 85-109 and ‘Napoli centro di produzione cembalaria alla luce delle recenti ricerche archivistiche’, Fonti d’archivio per la Storia della musica e dello spettacolo a Napoli tra XVI e XVIII secolo, (Naples: Editoriale Scientifico, 2001) pp. 205 - 225.

[74] See footnote 73 above, ‘Napoli centro - - - ’, p. 210.

[75] See Francesco Nocerino, ‘Evidence for Italian mother-and-child virginals:  an important document signed by Onofrio Guarracino’, Galpin Society Journal, 53 (2000) pp. 317 - 321.

[76] See my book, Ruckers.  A harpsichord and virginal building tradition (Cambridge University Press, Cambridge, 1990; digital reprint, 2008) pp. 35 and 39.

[77] The existence of an ‘orphaned’ child virginal made in England shows that mother and child virginals were also made in England.  This instrument dated 1638 is by Thomas White, a seventeenth-century English harpsichord and virginal builder, and is in private ownership in London, England.  This small  instrument, as well as clearly being at octave pitch, has a slot in the baseboard through which the jacks of the ‘mother’ instrument must have passed when it was placed on top of the mother instrument so that both instruments could be played at the same time from the keyboard of the mother virginal.  See John Barnes, ‘Keyboard Instruments’, Made for Music.  An Exhibition to mark the 40th Anniversary of the Galpin Society for the Study of Musical Instruments at Sotheby’s, (London: The Galpin Society, 1986) No 177.

[78] The existence of these instruments and their identification as a type of keyboard instrument was revealed in a paper by Francesco Nocerino entitled ‘The “Tiorbino”:  an instrument built by harpsichord makers’, given at the combined meeting of the Galpin Society and the American Musical Instrument Society in Edinburgh on August 8, 2003 (see the end of this footnote).  In this paper Nocerino mentions a tiorbino made by Onofrio Guarracino “con tastiatura d’avolio” (with an ivory keyboard).  I have been able to show that the ‘spinetta all’ottava’ in the Museo Teatrale alla Scala, Cat. No. MTS TP/04 uses the Neapolitan oncia in its design and construction and that it was therefore made in Naples.  It also has many other Neapolitan features such as its rosette, the keyboard which slides in and out like a drawer, etc.  Its string scalings do not fit into the normal pattern of metal-strung instruments implying that it should be strung with strings of gut.  In my opinion it should therefore be classified as a tiorbino.  It has none of the usual features of Guarracino’s instruments, however, and is dated 1707 by which time, if he was even still alive, Guarracino would have been 80 years old.  It is therefore probably too late to have been made by him. See also:  Grant O’Brien and Francesco Nocerino, ‘The Tiorbino: an unrecognised instrument type built by harpsichord makers, with possible evidence for a surviving instrument’, Galpin Society Journal, 57 (2005) pp. 184 – 208 with colour plates pp 232-5.

[79] See footnote 73 above, ‘Arte cembalaria - - -’, p. 95 where there is an image of the baptismal certificate.

[80] Francesco Nocerino, Harpsichord Makers in Naples during the Spanish viceroys (1503-1707).  Recent news and unpublished documents, a paper presented at the combined meeting of AMIS and The Galpin Society, Vermillion, South Dakota, 19-23 May, 2006.

[81] The first instrument by Guarracino to use the form “Honofrio” or “Honofrius” in the signature is the 1677 rectangular virginal in the Museo Nazionale degli Strumenti Musicali, Rome, Nº 901.  See Maria Luisa Cervelli, La Galleria Armonica, (Rome: Istituto Poligrafico e Zecca dello Stato, , 1994), Cat No 901, pp. 348-9, illus. p. 353.

This paper was published by the author:  ‘The single-manual Italian harpsichord in the Royal College of Music, London, Cat. No. 175:  an organological analysis’, Galpin Society Journal, LXII (2009) pp. 55 – 99 with colour plates pp 194-5.