Wood and Wire — A Lecture by Peter Forrester. (2005)

Originally published in Lute News: The Lute Society Magazine. Reprinted by permission of Peter Forrester and The Lute Society. On-line edition corrected and amended by Andrew Hartig per Peter Forrester, February, 2006.

An expanded version of the same, with photographs, appears as "Wood, Wire and Geometry" in Michaelsteiner Konferenzberichte 66: Gittare und Zister — Bauweise, Spieltechnik und Geschichte bis 1800. (2005). ISBN 3-89512-125-8. (Available directly from Kloster Michaelstein, http://www.kloster-michaelstein.de.)

The existence of citterns and other similar instruments depends upon the availability of wire produced from metallic ores by experts using difficult, expensive and time-consuming processes. Around 1600 we are mostly concerned with iron and brass, and in particular with the tensile strength of the wire available. This relates directly to the possible open string range, and it was not until around 1590 that the orpharion was able to emulate the lute. Metal working in Germany was far superior to that in the rest of Europe, and the history of English iron and brass is full of the importation of workers and knowledge directly or via France. All knowledge at this period was empirical. The chemical analysis of iron and steel did not start until well into the 19th century, and brass was not made directly from copper and zinc until the 18th.

Some thirty years ago, at Lute Society summer schools, there was general agreement that brass and iron strings sounded better than steel. This continues, perhaps because of the use of modern alloy steels. Around 1600 steel was iron with a carbon content of between 0.5 and 2.0%. Wrought iron had less than 0.2%, cast iron has 3.0 to 4.5%. Cast iron melts at a temperature of 1130°C., achievable by early blast furnaces which reached England by 1500, but, as the carbon content decreases, the melting temperature increases until that of wrought iron is over 1500°C.

The earliest method of producing wrought iron, used in the Weald from Roman times, was by means of a bloomery furnace. Here the 'bloom', of iron mixed with slag, never became completely liquid, and was reheated in a second hearth to burn out carbon, and hammered to expel the liquid slag. With additional heat and charcoal 'natural steel' could be produced by leaving a certain amount of carbon present. After the introduction of the blast furnace, cast iron could be treated similarly, but with greater control. With care and experience the process could be stopped at a suitable point. Wrought iron could also be given a steel surface, 'case hardening', especially for armour, by heating in close contact with charcoal for some hours. In 'The Faerie Queene' we read of good Sir Satyrane that

He was all armd in rugged steele unfilde,

As in the smoky forge it was compilde. [note 1]

— good working armour. Filing and polishing would reduce the thickness of the steel surface. In summary, iron for wire was available, occasionally 'steely' of a higher tensile strength. One Wealden iron works was opened in 1565 at Robertsbridge by Sir Henry Sidney, using 'aliens', for the production of steel and iron wire.

Brass also was produced from Roman times using a method described by Theophilus in the 12th century. Metallic zinc was known and imported during the 16th century, but the biggest problem with brass making is that zinc boils at a temperature lower than the melting point of copper. In practice, refined copper ingots or plates were heated in contact with a zinc ore, calamine, until their weight had increased by the desired amount. As with iron the quality of the metal could vary considerably according to the content of the ore and the capability of the workman.

The orpharion requires a higher tensile strength for its top course than other instruments, and its short life seems to coincide with the working life of one particular wire drawer in Nuremberg. In 1621 Heinrich Schutz wrote to the secretary of the Elector of Saxony about a purchase of 'staline', perhaps steel, instrument strings, unobtainable elsewhere; and asking that the Nuremberg Council be directed to give Jobst Meuler permission to make them. In 1592 one Friedrich Held had obtained a privilege, from the Town Council, to control production of certain types of wire. In 1608 he obtained a second, imperial, privilege from the authorities in Vienna. In 1609 and 1610 he attempted unsuccessfully to close down Meuler's production, and in 1613 had Meuler's tools confiscated by pretending that Meuler owed him money. The Town Council reacted by putting Held under arrest, although not for long, because he had the imperial privilege. In 1621 he obtained another privilege from Vienna, after which Meuler's wire seems to have become unobtainable, and the orpharion to disappear. Octave strings on citterns in Dutch paintings also start to vanish from this date. Meuler died in 1632. [note 2]

William Cecil, Lord Burghley, was anxious to make England independent of imports, especially for the production of brass ordnance and wire. Iron wire was required for carding wool, and brass for pins, etc. In 1565 William Humfrey, assay master of the Mint, together with a German, Christopher Schutz from Saxony, applied for a patent for brass and wire manufacture from Cecil. This resulted in the establishment of works at Tintern from 1566, using a high quality iron ore from Glamorgan, and calamine from Somerset. Copper was produced by another German organisation, later called the Mines Royal, from Keswick in the Lake District. The production of iron wire was an immediate success producing 25 cwt. of wire, described as of excellent quality, each week. The Mines Royal, and the Company of Mineral and Battery Works were officially incorporated in 1568, with Cecil as a major shareholder in both. Brass production proved more difficult. The brass works were not completed in 1577, however a second works was established at Isleworth in 1582, and by 1600 brass production was the most profitable part of the Company.

After an auspicious start, wire manufacture entered a series of ups and downs. In 1571 Tintern was farmed out giving the shareholders a respite from the hitherto steady expense, but producing complaints about the quality of the wire. In 1581 the wire was described as "better than foreign and being exported". In 1594 and 1596 there were more complaints. Meanwhile there were various attempts to discourage the import of foreign brass and wire by prohibition or high duties. Sir Henry Sidney and Edmund Roberts, both shareholders of the Company, were purchasing iron from the Company's Glamorgan furnace for their own wireworks at Robertsbridge in 1568. In 1580 Sir John Zouch established a wire works in Derbyshire, closed the following year after a letter of prohibition for infringing the Company's patent, but later reopening. By the beginning of the 17th century, instead of Burghley's use of monopolies to improve national resources, they were increasingly used as rewards for loyal courtiers.

Derryck's letter of 1594, where he asks for an "orphytre or bandora, of the new fashion which hath both the bridge and the stops slope, with the trebles of wire. as the other strings.", must represent the situation in England, at least, with a very intermittent supply of good quality wire, perhaps only occasionally of iron with sufficient carbon content to produce a high tensile strength. His "new fashion" of sloping bridge and frets can be dated some twelve years earlier by a 1582 carving from Lord Burghley's home at Theobalds where a 7 course bandora is shown. [note 3]

How strong was good wire when available? Meuler's wire would seem to have been about a semitone less strong than gut, based on the string length of Palmer's orpharion. Wire on citterns could be around 2 semitones less strong than Meuler's when using octaves. A painting by Floris de Vriendt [note 4] shows nine pegs, indicating octaves on the lowest (g) course only. Later citterns using the 'French' tuning had ten pegs with octaves on the fourth (a) course also. The 'Italian' tuning used in England seems never to have achieved octaves on its fourth (b) course. Octaves can be clearly seen in some tablature illustrations — Le Roy, 1565; Vreedman, 1568; Kargel, 1578. The triple third course is shown on the full-size marquetry cittern on the 'Eglantine' table. [note 5] Outside Italy, strings were arranged: top course, iron; 2nd course, brass (a fragment of brass on the fourth peg of one of the cittern remains from the Lelystad shipwreck, circa 1620, confirms this); 3rd course, one twisted brass string plus an octave pair of iron; 4th course ditto, for the 'French' tuning, or a plain brass pair in England. This variety seems to retain intonation accuracy on the upper frets. Each pair of a course was a single length of wire from peg to peg, via a hitch pin on constructed citterns, or through a slot in the comb, around a horizontal rod, and back through the next slot, on Italian carved citterns. [See illustration 1] Only one rod, probably not original, survives, on Campi's ceterone, but their use can be deduced from the lack of wear on some of the combs of other citterns, such as the Salvatori. On all citterns,where double courses are shown, we infer the use of a single wire. This means that a triple course must consist of a pair, plus a twisted string at the octave below. A twisted string naturally forms its own loop, so that sharp wire ends are not left adjacent to the player's right wrist and clothing. This also accounts for apparent single courses in some early tuning descriptions.

Illustration 1

In Italy the narrow compass of the hexachord tuning permitted plain brass wire to be used throughout. One exception appears in Cerone's tuning diagram of 1613 [note 6] where both an octave on the lowest course and an extra string on the second are shown. This implies iron wire for the octaves and also for the first course, because the extra second can only be a defence against breakage. (Another cittern where breakage problems on the second course might occur is Robinson's. We have usually considered that the engraver interchanged the 2nd and 3rd courses. However if the usual 43-45cm string length cittern were raised a fourth, its optimal string length would be 32-33cm. — very cramped for normal fingers. A more comfortable string length of 36cm might be too much for English iron octaves and at the limit for brass. The V & A cittern 10/2 with 10 pegs and string length 38cm may be relevant or could have been using Meuler's wire.)

During the 17th century octaves disappear. Sir Peter Leycester in 1656 writes "-Psithyrne: It containeth foure courses of stringes, as at this day we use it; each course being doubled, having two strings of one sound in each course;".[note 7] In Holland we find paintings of citterns with only nine pegs, then eight. [note 8]

The traditional Italian cittern was a large carved instrument with a string length of around 61 - 62cm, with a hexachord tuning. This necessitated a large piece of wood clear of knots, and the grain direction was important. [See illustration 2] This places the dimension least likely to shrink against the soundboard. Shrinkage will mostly occur across the back and counteract the pressure of the strings upon the belly. Additional strength is given by fitting the belly into rebates at the neck and comb. The bars, usually two on the largest instruments, but only one on the smaller, are let into, or more often through, the sides. [See illustration 3] This means that the belly must be glued to the bars and sides simultaneously. This is not as difficult as it sounds, although the Plebanus adds wooden dowels through the sound board into the bar in the manner of hurdy-gurdies. (Thicknesses: Sides are about 5mm, thicker in the short grain area near the comb. The back is thicker in the centre — around 5 - 6mm or more, thinning towards the sides where it is usually guttered like a violin, reducing it to around 2mm. Bellies are usually 3.5, even 4mm or more in the centre, reducing to around 2mm at the edges. This shaping is easiest to do after gluing, and can be finished after purfling and stringing up, so that the sound can be judged.)

Illustration 2

Illustration 3

All these carved citterns used peg-heads with sagitally inserted pegs. [See illustration 4] The peg-box on the Plebanus is an alteration; its sides are separate pieces and the satyr head is attached with a mortice and tenon joint. All Italian citterns have a hook behind the peg-head; interestingly there are at least four with the same outline as the Plebanus — Salvatori's cittern in Paris, others in Leipzig, Vienna and Rome.

Illustration 4

The nut on citterns usually consists of two parts — a brass cut-off fret, and some method of separating the strings. On carved citterns this can be merely slots in the fingerboard behind the fret, for example the Gironimo Campi in London; others use a specially shaped insert. Frets are of brass, often folded so that the strings press against a surface work-hardened by bending. Behind each fret are inserted dovetailed slips of a hard wood with their grain at right angles to the string, which we nowadays call wedges. Sometimes frets and wedges traverse the fingerboard completely. Most often they are covered on the bass side, sometimes on both. I have only seen wedges removed on one fingerboard — the Armada cittern's in Belfast. Here they do not taper and would not act as wedges. I suspect that on the Salvatori they may. Two frets have broken through on the bass side, and look in my photos less deep than on the treble, so they could taper in depth. Seen from above, all wedges are approximately parallel sided. (I always insert the wedges before cutting the actual fret slots).

The 18 or 19 frets of a cittern fingerboard make it difficult for a player to orientate himself. It is easier with diatonic fretting, and all chromatic citterns — and the Palmer orpharion — use a related colour-coded pattern in the wedges. For the missing English citterns it can be inferred from the instrument on the "Eglantine" table, and from Robert Fludd's "Utriusque Cosmi Historia", [note 9] which both show a garbled ab ab pattern, and also from Robinson's instructions for running passages, which always place the first finger behind a dark fret.

On the larger instruments there is usually a rebate on the underside of the fingerboard where it touches the belly, presumably for a strap. The angle of the sides and end of the instrument makes an end button difficult to fit.

(It has been suggested that all sizes of Italian citterns were at the same e' pitch, using extra strong iron wire where necessary. If so we would expect a fairly even range of string lengths from small to large, perhaps centered on one size. However, if the extant instruments are arranged by string length, definite groups appear. This suggests different pitches, with top courses at a, b, and e' for the fourth, third and first groups, using all-brass stringing. The second group is either for a tone-low tono corista or used iron for the top course. Its size is similar to that of Praetorius' chor zitter.)

Around 1570 a smaller chromatic cittern was produced in or around Brescia. These were not carved but constructed like a violin, and several makers made both instruments. We have examples by Gasparo da Salo and Paolo Maggini. These would have been the instruments similar to those from England of which Vincenzo Galilei said, 'The cittern was used first, before other nations, in England, in which island they were already made to perfection". [note 10] Perhaps the 'perfection" was chromatic fretting. The best known instrument must be the Girolamo Virchi of 1574 in Vienna and it was Girolamo's son Paolo who wrote some rather difficult music for an extended tuning using both six and seven courses. His large left-hand stretches may be the reasoning behind the slightly smaller instrument by Girolamo in Paris. This new tuning necessitated the addition of iron and twisted strings. Brescia was famous for its steel, made by soaking wrought iron in a bath of molten cast iron.

An important difference between the two types of cittern is seen in the side view. Because the constructed cittern's belly is slightly curved longitudinally, and the fingerboard tapered, the bridge height is some 5 or 6 mm higher for the same string length. Combined with a lighter construction this means that the string tension has to be less. Typical constructed cittern thicknesses are similar to a violin's. (Belly about 2.75 - 3mm in the centre, thinning to 1.5 - 2mm at the sides. Back 2.5 - 3.5mm). The back is sometimes of one piece, for example on the da Salo and Maggini instruments. Early constructed citterns in the Low Countries and presumably England as well as Italy, have the sides wrapped around the back, giving a larger gluing surface. This necessitates some form of internal mould. The potentially ugly joint between sides and neck block is always hidden by a decorative "split baluster".

Once the bars are attached to the belly, it can be used to determine the shape of the side-to-belly joint. The usual pattern of barring is similar to that on Dutch citterns. [See illustration 5] The two main bars are opposite and separated by small pilasters. The second back bar, opposite the rose centre, and the belly bar above the rose are similarly supported. This helps to stiffen the sides in the straighter area between the concave and convex curves. The third belly bar, between the main bar and rose, tapers away at the ends before reaching the sides. Its purpose is to hold the belly down against the pressure of the bridge around the fulcrum of the main bar. There are some variations. The Paris so-called "Stradivari" has a second main bar, perhaps original, below the bridge. Other extant citterns still use a peg-head but the Gasparo da Salo has a violin type peg-box. (Its eleven pegs suggest that it was originally a five course instrument with octaves on the lowest course.)

Illustration 5

This small constructed cittern seems to have had only a short existence in Italy. However it influenced the diatonic carved cittern in several ways. Some, like the Salvatori, adopted a complete chromatic fretting. Others have a chromatic fourth course. Where this occurs the fourth fret is placed at the flat position instead of the sharp of complete chromatic fretting, giving a useful a flat. On complete fretting the c sharp on the first course is preferred. In some instances, for example, citterns in Vienna and Leipzig, the fourth course frets extend under the fifth and sixth courses. This seems to be because they were then easier to fit, rather than primarily for additional notes. The Augustinus cittern in London and the so-called "Amati" cittern in Florence, and Cerone's tuning description used octaves on the lowest course. On the Augustinus this thirteenth peghole was later filled in. Probably the smaller 43cm carved citterns also date from this period, towards 1600. Larger citterns began to be constructed from separate pieces of wood, but preserved the traditional shape. Leipzig 613 has a carved body but separate neck. The Pietro Paulo and Merusaglia citterns are completely constructed, but perhaps the most interesting examples are the "Amati", and Campi's ceterone, both belonging to the Museo Bardini in Florence. The cittern came apart during the floods which has allowed us to see its construction clearly. The neck is dovetailed into the neck block and reinforced with a rather unnecessary nail — presumably from a violin or lute tradition.

Monteverdi's "Orfeo" calls for two ceteroni, but it seems unlikely that this was a very common instrument. The example in Florence may be the first, as the label says that it was made by Gieronimo Campi, but invented or designed by Innocentio Peretti. Also there are some signs that it was altered during construction. Firstly, there are seventeen slots in the comb for nineteen pegs, and the upper pegs in both peg-boxes are awkward to reach. Secondly, the combination of seven double and five single courses on a long string length, 68 and 123cm, means that the bridge tends to move sideways to its own preferred position. On my copies I had to plane away some of the fingerboard and neck on the treble side. Looking at the original it seems that Campi did the same. There is a very careful attempt here to make the instrument look as though it was carved from the solid. The joints between the sides and back, and between fingerboard and neck are disguised by purfling. Like the Piero Paolo and Merusaglia citterns the back is glued onto the sides, but is massive, about 8 or 9mm thick at the heel, thinning to 4 or 5 mm at the edges. The belly is of two pieces with the wide grain in the centre, thinning to 2 or 2.5mm at the edge. The wedges seem to be of walnut with no sign of colour coding. However they may have been stained and faded; some of the purfling is very pale. There is no cut-away to the neck which I find in my copies to be very slightly flexible — enough to make it necessary to tune the upper pegs before the lower.

Besides and probably before the ceterone there was also a citara tiorbata in Italy. Pietro Paulo Melii wrote for a nine course instrument in his "Quarto Libro." [note 11] He uses and names Signor Virchi's tuning, so other instruments may have used the traditional hexachord. The little evidence for its construction and appearance consists of a pattern for a neck by Strad ivari , and its depiction in the background of several copies by assistants after an apparently lost painting by Evaristo Baschensis.

We have far fewer extant citterns from north of the Alps. However they seem to have been similar throughout the Low Countries, France and Germany. The typical instrument was similar in size and basic construction to the citterns of Brescia. They seem always to have used a peg-box and to have omitted the hook. From around 1570 until 1650 ten strings in four courses were usual, although the Abraham Tilman has eleven pegs. Its probable arrangement is shown by Mersenne. All five known instruments have one piece backs. Those from the Lelystad buertschip were probably plain, the others have applied veneer stripes. Many seen in paintings have striped sides which are quite difficult to make for a one-off instrument. Louis Grijp has found two inventories where cittern moulds were assessed at twice the value of citterns, so they were presumably fairly complex for mass production. Brussels 1524 has three bars on the belly, with the main bar almost below the bridge. None of the bars reach the sides — the main bar is 1 or 2mm short; the upper bar leaves gaps of about 10mm. The middle bar is short as usual. There are pilasters at intervals against the sides. At present there are no bars on the back, but there are glue marks opposite the rose hole and the main belly bar. Obviously there have been some alterations and a rose bar on the belly may be missing also. One is currently present on the Tilman.

From iconography and tablatures, Dutch citterns appear to have been mostly diatonic. On the V.& A. instrument 10/2 the original four course diatonic wedges can be seen below the later chromatic fretting. It also has inlaid stripes down the bass side of the fingerboard only, like the Lelystad citterns, so is likely to be from the Low Countries, rather than as the catalogue entry. There is virtually no French iconography, unless we count Cornelius Johnson's portrait of Henrietta Maria, [note 12] but it may be that her cittern, like the artist, was Dutch. (Another, somewhat stylised, and slightly damaged cittern is played by Diana in the centre of the ceiling of the Queen's Bedchamber in Kew Palace). The two fretted citterns from Freiberg are chromatic; an early woodcarving from Nurnberg (1535) is diatonic. [note 13] Four sizes are mentioned in a mid-17th century German manuscript. [note 14] Dutch paintings include at least three. [note 15]

We know very little about English cittern construction due to the complete absence of instruments. They had colour coded chromatic fretting from before 1568. There were nine pegs for four courses. The usual size was around 43 - 45cm string length, and there was also a treble size with its top course tuned to a'. Holborne's "Cittharn School" requires both tunings for duets with a bass instrument, and one of Robinson's duets requires two citterns "in the unison", implying that two sizes could be available. He also refers to "a third Citharen", his fourteen course cittern, tuned to the same pitch as the usual size, because of the left-hand stretches required. Praetorius enthuses over the klein Englisch zitterlein, a treble cittern with a four course guitar tuning, although he gives a tuning which is an octave too high. This instrument is referred to by Sir Peter Leycester when he writes about the gittern — "which is indeed onely a Treble Psithyrne being somewhat lesse than the other (cittern) having the same number (8) & the same order of wyre-stringes — onely some variation in the Tuninge which may be baryed in the Psithyrne at pleasure." Playford published some rather simple music for, and illustrated both, instruments. A vanitas painting in Tate Britain by Edward Collier shows the later stringing in double courses, apparently two of iron and two of twisted brass. Only one painting, the well-known Henry Unton memorial portrait, shows a peg-head instead of a peg-box, where dots of paint clearly depict the tops of the pegs.

All citterns used a meantone tuning somewhere between equal temperament and quarter comma. Fret positions were marked on a rule and transferred to the fingerboard. This is demonstrated on the Campi cittern where the rule obviously slipped at the tenth fret. [note 16] There is considerable deviation from theoretical positions on most citterns. The eleventh fret is usually averaged. We must assume that positions were determined by ear and tradition. Italian citterns are usually in the area of fifth to sixth comma (fifth at 698 cents). Brussels 1524 is very close to the Gerle fretting described by Eugen Dombois (fifth at 698.4 cents — sixth comma). [note 17] In England, and unlike the Palmer orpharion (around sixth comma), the Rose bandora is close to equal temperament with a fifth slightly more than 699 cents. As both bandoras and citterns were used to accompany other instruments, then these others, at the court of Elizabeth, must surely have been tuned similarly?

Notes