ENGLISH VIRGINAL DESIGN CONCEPTS AND SEVENTEENTH CENTURY PITCH STANDARDS

 

Frank Hubbard, writing in “Three Centuries of Harpsichord Making”[1] discusses English virginal scales, stating that, “Most of the virginals had scales between 11 and 12 inches although some are as short as 10½ inches, and one (strangely the earliest) is 13 inches”.  Hubbard’s information is essentially correct, though his pioneering work has masked the scaling approach that the original makers appear to have used when designing their instruments. 

            The English virginals of which Hubbard wrote survive in small numbers.  I am aware of only twenty two extant instruments.  The instruments are stylistically very close to one another.  The cases are rectangular, and are usually of oak with coffered lids, or show evidence to suggest that they were originally built that way.  The exteriors are natural wood, finished with a dark varnish, decorated only by iron strap hinges and lock hasps.  The interiors, on the other hand, show a blaze of colour.  The soundboards are decorated with flowers, birds, arabesque border patterns and one or more gilt geometric roses; the lid and keywell flaps have paintings, often a park scene or a representation of Orpheus taming the beasts; and the faceboards, keywell and soundwell feature finely cut mouldings and gilt embossed papers.

            The basis of Hubbard’s approach to virginal scalings can be seen when looking at pitch standards as applied to historical instruments in the revival of interest that has taken place over the past thirty or so years.  Although it has long been accepted that pitch standards were not at the same level as today, and even that pitches both higher and lower than our modern standard of a1 = 440Hz existed,[2] it is only recently that something approaching a good understanding of them has been reached.  It has been usual for very many early music groups over the past several decades to perform at a pitch of A415, chosen simply because it is an equal-tempered semitone below modern pitch, and so that harpsichords could be made to transpose between the two pitches with little difficulty.

            The recent, exhaustive, work on pitch standards carried out by a number of authors, in particular Bruce Haynes, who was awarded his Doctorate by the University of Montreal for his thesis Pitch Standards in the Baroque and Classical Periods, has brought much new information to general attention.[3]  Haynes has found strong evidence for over two dozen different pitch standards in Europe from about 1500 to 1850, including five different standards in England.  This, and research carried out before, has implications for the English virginals.  This paper has two purposes.  Firstly, I shall consider how the virginal makers themselves designed their instruments, and in particular chose their string lengths.  From there I shall attempt to place the virginals into a wider perspective and try and determine pitch standard values for those instruments in the light of other surviving evidence.  

            The widespread modern view that old instruments should be played at a pitch level around a semitone below modern pitch can be seen when examining the surviving instruments.  Generally, restored instruments have brass treble stringing if they have a  c2 string lengths of 12 inches or under, and iron treble stringing if longer than 12 inches.  The only exceptions to this are the 1642 Thomas White in the Victoria and Albert Museum, restored by Derek Adlam,[4] and the 1684 Thomas Bolton virginal in Warrington Museum, restored, (although no longer still playable), in 1953 by Leslie Ward, working for Arnold Dolmetsch Ltd.  Both these instruments are strung in ferrous material and have c2 scalings a little under 12 inches.

            I first became particularly interested in English virginal scalings and pitch standards when I examined the 1675 Stephen Keene virginal, in private ownership in Cheshire.  I was struck by this instrument - it was certainly a brother to the 1668 Keene in the Russell Collection, here in Edinburgh[5], and was also very clearly related to the 1671 Philip Jones virginal at Tabley House in Cheshire.  In fact, the Cheshire Keene and the Jones virginals are much closer to each other in design layout than the two Keene instruments are.  The Jones virginal also shows a derivation from the Russell Collection Keene, clearly suggesting that the two makers had some type of common workshop tradition.[6]  The 1675 Keene has a c2 scaling of 308mm, a little over 12 inches, and the cs are comparatively longer than the fs, suggesting the average scaling is a little shorter.  The Russell Collection Keene has a scale of 287mm (11¼ inches), suggesting brass treble stringing for a pitch around A415, and the Jones has a scale of 328mm (just under 13 inches), which implies iron treble stringing.  Considering the three instruments together, it is difficult to form a firm opinion as to whether the Cheshire Keene should be strung with brass or iron in the treble.  It is from this position that my detailed research on virginal scalings, stringing material and pitch standards began.

            Of the twenty two surviving English virginals that I know of, seven were built by the White family, Thomas and his son James; five were made by what I refer to as the Townsend workshop tradition - instruments by Townsend, his apprentices Player and Keene, and the Philip Jones; three are by Exeter-based makers; two are by Adam Leuersidge; there are two Anonymous instruments, possibly Scottish and possibly by the same maker, both dating, I believe, from before 1625; and there are three instruments by makers without known connections to any of the groups listed above or to each other.  I have detailed string lengths for all of these instruments, although the measurements of the Anonymous “AH” virginal, belonging to the National Museums of Scotland and currently housed at the Russell Collection[7] have no practical use as the two bridges are replacements (although very closely following the original positions), and the soundboard register for the jacks is probably also unoriginal.  The 1664 John Player scalings should also be treated with caution as the right-hand bridge and soundboard register date from modern times, though the measurements must be close to original.  The only other instrument requiring some consideration is the 1679 Charles Rewallin now at Exeter, where the treble of the left-hand bridge has lifted.  With those cautions in mind, I believe I still have usable data from 21 of the 22 instruments. 

            The first aspect to consider, is what evidence there is of original stringing materials from the instruments themselves.  There are only two instruments which retain any evidence of historical stringing methods.  The Philip Jones virginal has remains of strings around the original tuning pins, now kept separately from the instrument following its restoration in the 1970s.  These string remains fall into two groups; most are clearly 19th-century cosmetic additions and are all of ferrous material.  However, nine string remains are neat and have every appearance of originality or very early use.  These strings have been independently measured by myself, during one of my visits, and Andrew Garrett, when he and Richard Clayson restored the instrument.  Our measurements agree with each other, and we both accept the strings as either original or very early.  Two of the nine strings are of a copper alloy, the remainder are ferrous, the lowest surviving ferrous string being at c. 

            The second instrument that contains evidence of historical stringing is the 1641 Gabriell Townsend virginal at the Musee Instrumental in Brussels.  No original strings remain, but the wrestplank has gauge numbers written in an eighteenth-century hand that matches the writing found on some replacement jacks which originated in an English eighteenth-century double-manual harpsichord.  The gauges indicate tensions which are considerably higher than found in eighteenth-century harpsichords and spinets, and may be indicative of the original stringing.  A complication exists in regards to the string materials.  Beside gauge 13, at the bottom of the compass, is the word brass, and next to gauge 10 (at note B, an octave below middle c) is the word Coper.  This is almost certainly a mistake.  It could be that the words were transposed, meaning that the lowest octave should be in red brass and the remainder in yellow brass, or it could mean that iron was intended for the treble and copper (probably meaning yellow brass) for the lowest octave between the two written indications.  In any case, there is something amiss, in addition to the fact that the stringing list dates from the eighteenth century.  Significantly, the c2 scale of this instrument is almost identical to that of the Philip Jones virginal.

            This evidence from the surviving instruments alone is not of great help when trying to reach firm conclusions about the treble stringing material of the surviving English virginals.  Further, given the evidence of ferrous strings on the Philip Jones virginal, and of possible brass treble strings on the Townsend, it appears that the general supposition regarding string lengths correlating to treble stringing materials is suspect.

            Given their workshop-tradition relationship mentioned above, a semi-logarithmic graph was drawn, showing the two Keene virginals and the Jones.  When graphed, the three lines are essentially parallel to one another, in both the bass and treble.  This is also to be found in the instruments of the Ruckers family, who have been shown to have built instruments at different pitch standards.[8]  This suggests that the English virginals may also have been built with the intention of using the same treble stringing material and different scales for various pitch standards.

            It is important therefore to consider the design concepts used by the makers as an important first stage in an attempt to make further conclusions about scalings, stringing materials and pitch standards.

            Unfortunately, the surviving instruments represent the work of 14 or 15 different makers, of which only Thomas White has left more than two examples.  Further, it seems that instruments were probably built on a custom-order basis, and there is little to suggest the use of standard models.[9]  One must also take into account that the English virginal makers were perhaps not the finest musical instrument craftsmen to have ever worked.

            A major problem with English virginals is trying to decide how the makers set out their design scalings.  Modern researchers commonly quote the c2 string length (that is, the c above middle c) as the scale, and there is much evidence to suggest that it was this note that was used in many schools of plucked keyboard instrument making.[10]

            It has been shown in instruments from many different schools that scalings, amongst other measurements, are laid out with reference to common measure, and one would expect that English virginals would follow the same principle.  The first consideration is the length of the English foot during the seventeenth century.

            The study of length measurement appears to be as confused as that of pitch, and as yet no thorough research work has been carried out on it.  It is known, however, that the length of measurement was set down by Statute in 1558, and was referred to as the “Queen Elizabeth Yard”.  As yet I have been unable to trace any legal source for the actual distance for this measurement.

            The earliest evidence concerning the foot length in England I am aware of appears in the Lexicon Technicum, compiled by John Harris and printed in London in 1704.[11]  In his entry on Measures, he gives a list of foreign foot-lengths compared to that of England.  Unfortunately, his methodology is not fully explained, although it is clear the measurements come from a number of different sources.  The lengths of the different towns have been quoted in two forms, the first in proportion to the English foot divided into 1000 parts, and the second comparing to the English foot divided into feet, inches, and tenths of inches.  The first measurement is the more accurate.  In practice, the rulers used (assuming they were specially made for the purpose) could be divided practically into no more than 500 parts, the person taking the measurements having to decide whether it is closer to a division line or the centre of the space in order to achieve greater accuracy.

            Having compared all of Harris’s figures with those from other sources it appears that his table is not very accurate.  However, he quotes two measurements for the Parisian foot which are very close to each other.  I have two separate sources giving information of the Parisian foot and they agree with each other to within one hundredth of a millimetre.[12]  This is a remarkable agreement.  The two measurements given by Harris, 1.066 and 1.068 feet probably reflect the practical problems of taking the measurements as outlined above.  Using those two lengths, and comparing them with the two measurements I have for the parisian foot, suggests that the English foot is between 304.156mm and 304.736mm.  The next measurement I have for the English foot comes from Johann Heinrich Zedler, quoted in his Universal-Lexicon, published in Halle in 1737.[13] Zedler gives the length of the English foot as 304.5375mm, almost halfway between the two figures given above.  Using 20 different measurements from the English virginal by Robert Hatley dated 1664, I have found that his ruler used a foot length of 304.39mm.        

            Of the c2 string lengths of the twenty one examples which can be reasonably judged, only about twelve - a little over half of them - have scalings which fall into easily divisible units of the inch. This suggests that, unlike the instruments of many other schools, the English virginal makers used another method in determining the treble design scale.

            English virginals, like almost all other virginals, have the jacks placed in pairs between the strings, which must be moved apart to give enough room for the jack and quill.  Therefore all of the strings on the player’s side of their respective jacks are slightly longer, and the strings behind their jacks slightly shorter, than their theoretical lengths, were they to be evenly spaced as occurs in a harpsichord.  Nineteen of the twenty two surviving examples have the top string, regardless of what note it is, to the player’s side of the jack.  There is a good acoustical reason for this, since putting the string further away from the spine increases the amount of soundboard wood that is able to vibrate.  Of the three exceptions; one shows a number of signs of having been altered during construction with an extra note added in the treble, and another instrument has 56 notes rather than the (by that time) usual 55 notes.  Therefore, the top string being on the player’s side of its jack appears to be a general design-concept rule. 

             One consequence of this is that instruments with compasses ending on c3, d3, or e3 in the treble all have the c’s a little longer than theoretical, and the virginals which ascend to f 3 have the c’s a little shorter than theoretical.  Making comparisons between instruments must take this into account.

            From compiling a complete list of string lengths for all of the surviving virginals, it is clear that the string lengths at the treble commonly fell into easily divisible units of common measure.[14]  On all but one of the instruments which ascend to c3, d3 or e3 the string length for the c3 string can be easily divided into common measure, and in about half of the examples the string length works out at 6 inches.  Other examples range from 4 inches, occurring in the 1638 Thomas White virginal,[15] which is an orphan from a mother and child combination, through 5¾ inches,[16] 6¼ inches,[17] 6¾ inches,[18] to 7 inches, as appears in the Anonymous “Countess of Mar” virginal at the Scottish National Portrait Gallery.[19]  In some instances the top note of instruments that ascend to d3 or e3 also work out in common measure, which is probably due to fortune rather than by design.  The only exception is the 1679 Charles Rewallin virginal at the Royal Albert Memorial Museum, Exeter, where the treble strings are comparatively short in relation to c2, which appears to have been the design string in this example, as happens in contemporary spinets.

            In instruments with f 3 as the top note it is the top f which can always be divided into units of the inch, with string lengths ranging from 4½ inches,[20] through 4¾ inches,[21] 5 inches,[22] to 5¼ inches.[23]

            It appears that the virginal makers took the relative lengthening and shortening of the string lengths into consideration when designing the instruments for a particular scale length, and had a general rule that c3 is used for the design string length, unless the compass extends to f 3, in which case that note is used.

            One remarkable feature about English virginals is the lack of standardization in regard to following Pythagorean scalings, that is, doubling string length for each octave, in the treble part of the compass.  Although some of the virginals have treble string lengths which are close to being Pythagorean scaled; for example the 1642 and 1651 Thomas White instruments, and the 1684 Thomas Bolton virginal, other instruments show treble scales which foreshorten by up to 15%; as occurs in the 1661 James White, the Anonymous “Countess of Mar” virginal, and the 1664 Robert Hatley.  Sometimes examples by the same maker can show this effect.  The two White instruments just mentioned are nearly Pythagorean, but others show a noticeable amount of foreshortening.  In the 1638 child virginal there is no choice, as the right-hand bridge is composed of two straight segments, but the 1644 virginal, now in Cardiff, tapers by over 11% between the top note and c2.

            It is possible that some makers exploited the fact that thinner wire is usually stronger than thicker wire because of the drawing process, and used thinner than normal gauges in the extreme treble.   The most reliable information suggests that the scale length increase available by using thinner wire is in the region of 2-6% per gauge reduction.[24]  However, on the instruments which foreshorten by up to 15% this would imply the need of at least three changes of gauge between the top of the compass and c2, and several instruments with this large foreshortening have scales increasing again towards c1, an octave lower, which suggests the strings would need to get thinner again.   I know of no evidence to suggest any stringing practices of that nature.

            This can be illustrated by compiling graphs showing each of the string lengths converted to its equivalent c2 scaling, in which (due to the relative lengthening and shortening of alternate strings) an average of close pairs has been taken.  When graphed, an instrument which is Pythagorean scaled throughout its whole compass it would be shown as a horizontal straight line.  When graphed in this manner, instruments such as the 1655 John Loosemore foreshorten by about 2½ semitones between the top of the compass and c2, then increases to within a semitone of the treble scale just below middle c, and the 1661 James White foreshortens by about two semitones in the top octave, then increases to nearly reach the equivalent of the treble scales at the eb above middle c.

            While I can offer no reason as to why makers should reduce the scales in some instruments by so much around c2, it is evident from the foreshortening, and the use of common measure for the significant treble notes, that it was the c3 or f 3 that was used as the design concept note.  As Bruce Haynes and other researchers have found evidence for different pitches in England, it is likely that the design scale is meant to imply a particular standard.

            To judge from the surviving instruments, it appears that a 6 inch c3, may have been considered by the makers to be something of a reference pitch, which I shall refer to a pitch V.  In determining the c3 string length of virginals at different pitches, mathematical ratios were probably applied, as have been shown to have been used by the Ruckers family.[25]  The earliest surviving dated instrument is the 1638 Thomas White child virginal with a c3 string length of 4 inches, which uses a 2:3 ratio, implying a pitch a perfect fifth above V.  The mother of this instrument would have been tuned to V-5s (ie. 5 semitones below V).  Another very early instrument is the 1641 Gabriell Townsend with a c3 string length of 6¾ inches.  This again can be expressed as a simple ratio of 6 inches, being 9:8, suggesting a pitch a tone below pitch V.

            When instruments which ascended to f 3 began to be built and the f 3 rather than the c3 used as the design scale string, it was necessary to use a 3:4 ratio to determine the equivalent string lengths.  For instruments at pitch V, the f 3 string should be 4½ inches long, and virginals a tone below pitch V, 5 inches.  Three virginals survive with 4¾ inch f 3 string lengths, which imply they are only a semitone below pitch V.  The single instrument with a 6¼ inch c3 scale has a 4:3 relationship, when converted to the nearest quarter inch, and is probably at the same pitch.  There is also a virginal with a 5¼ inch f 3, which might be at a pitch 3 semitones below V, and likewise there is an instrument with a 7 inch c3 scale, showing a 4:3 relationship to the 5¼ inch f 3 instrument, and implying the same pitch.  The ratio between the 6¼ and 7 inch instruments (and their f compass equivalents) can also be expressed as 9:8, further suggesting those instruments should be tuned a tone apart.  The single instrument which doesn’t fit the pattern is the 1656 James White, with a 5¾ inch c3, which shows evidence of having been altered during construction, and the short scale was probably an attempt by the maker to rectify the fact that the c strings  were comparatively shorter than they should have been, due to the string spacing geometry.

            Therefore, the 21 virginals from which reliable measurements were taken can be grouped according to string lengths.  There are 10 instruments at pitch V, 4 at a semitone below V, 4 a tone below V, 2 at three semitones below V, and the child virginal tuned a fifth above V.    

            The bass scalings of the virginals add support to the idea that the same stringing material was used on all of the instruments; those instruments at a tone below V having longer bass strings than the virginals at pitch V, for example.  It is not clear how the bass scalings were actually decided.  None of the significant low notes work out as units of common measure with enough regularity to be considered deliberate, rather than as the result of statistical chance.  This is probably a reflection of the working methods used by the makers in positioning the various soundboard components.  In most of the surviving instruments there is evidence that the right hand bridge was marked out first, using a graduated stick held against the spine, and the left hand bridge was positioned after the register slots were cut.  The overall case length of the virginals is related to the pitch, and therefore there is little latitude for the bass scalings.  In terms of linear measurement, a difference of a semitone in pitch should result in a difference of string length around 3½ inches.  Virginals from other schools appear to show a similar disregard to the exact string length of the bass notes.  

            It remains to determine the treble stringing material, and to give some idea of pitch standards. 

            Grant O’Brien[26] has shown that iron treble scalings are about 20 - 25% longer than brass scalings for instruments at the same pitch level.  He has also shown that different schools of instrument making probably used strings similarly close to their breaking point.  This can be tested by examining all of the reliable evidence.

            I have found six pieces of information which can give some idea how closely different schools of instrument makers stressed their strings.  All of the information must be treated with some caution.  There is a tuning fork that survives with a Swedish clavichord built by Pehr Lindholm, and another that was used by the late eighteenth-century Parisian harpsichord maker Pascal Taskin.  The scalings of the slightly earlier French harpsichord maker Blanchet can also be compared.  There is a large amount of evidence of mid eighteenth-century pitch standards in England, including pitchpipes and tuning forks.  The organ at the Jacobikirche in Hamburg had a set of gedackt pipes tuned to Cammerton, which may reflect the pitch of the Hass and Fleischer families.  Finally, the harpsichord owned by Christiaan Huygens, presumably the Couchet that had belonged to his father, had its pitch measured in Hz.

            Taking all of the evidence at face value, it suggests that the extreme difference of pitch for a given string length amounts to about three-fifths of a semitone.[27]  This compares more than favourably with research on individual makers, such as the Ruckers family, whose work in the treble of the compass is accurate to no closer than one semitone.[28]  Bruce Haynes claims that it is not possible to generalise pitch standards to an accuracy finer than one-fifth of a semitone. 

            Unfortunately, all of the evidence refers to instruments which have iron treble stringing material, with only the information from Hamburg giving direct assistance in determining brass treble string lengths, although the changeover point from brass to iron stringing in many harpsichords is known, and therefore reliable evidence is still available.

            On the assumption that English virginals were equally close to their breaking points[29] some comparisons can be made with the above-mentioned instruments, as can be seen on the graph on the slide.  If iron was used as the treble stringing material of English virginals, the instruments at pitch V, with a c3 string length of 6 inches, should have a pitch of between 478 and 489Hz, taking all the information equally, or if comparing the six pieces of evidence with the large amount of information compiled by Bruce Haynes, the pitch would be between 472 and 485Hz.  This standard is between one and two semitones above modern pitch.  The longer string lengths would produce pitches a semitone, a tone, or three semitones below that, approximately 450, 425, and 403Hz respectively.  Were English virginals to use brass treble string, the four pitch standards would be approximately 392, 370, 349, and 330Hz. 

            There is supporting evidence to suggest that iron strings were used in the treble.   A comparison with the virginals of the Ruckers family show that both the treble and bass string lengths are consistent with iron stringing, as can be seen in the semi-logarithmic graph.  However, when a typical brass strung virginal is added to a graph, it can be seen that the curve is somewhat different, the bass strings being roughly the same length, but the treble strings are much shorter.

            This can be seen in the early surviving English spinets. The spinet began to supplant the virginal in popularity around 1675, the earliest surviving examples were probably built around 1680.  Most of the makers of the early spinets were also makers of virginals, and according to written sources, harpsichords.  In terms of numbers of surviving instruments three makers stand out, namely Charles Haward, Stephen Keene and John Player.

            It is a pity there are no surviving Haward virginals, but fortunately, enough spinets survive by these makers to gain a firm idea of any pitch changes they introduced from the earlier virginals.  All of Haward's surviving instruments appear to date from the 1680's.  There is the harpsichord at Hovingham Hall, and a number of spinets.  The harpsichord has undergone considerable alteration, including the fitting of a new wrestplank, but the original key-pads survive, and from them the original wrest-plank and nut position can be reconstructed with some accuracy.  This instrument probably had an original c3 scale of 6 inches, similar to many English virginals.  The spinets have the same treble scale, but with shorter bass strings, a natural consequence of the spinet design and layout.

            John Player's surviving virginal has a 6 inch c3 scale.  All his spinets appear to be designed with shorter scales, usually about 10 inches at c2.  This, when converted to c3, would imply either a very high pitch instrument, or else brass treble stringing.

            Stephen Keene has left instruments showing a major change of design.   An early instrument of his, now in the Royal College of Music, London, shows a stringing layout that is reminiscent of his earlier virginals, but instruments built soon after the RCM spinet have much shorter treble, but slightly longer bass string lengths suggesting he changed to using brass stringing at a slightly lower pitch.[30]  These scale lengths are similar to those found in the early Player spinets, suggesting he also used brass stringing in his spinets.

            Further evidence to support this conclusion can be found when examining the earliest surviving English plucked keyboard instrument - the Lodewyk Theewes claviorgan, dated 1579, now in the collection of the Victoria and Albert Museum, London.[31]  This instrument appears to have originally been built as a harpsichord, but converted soon after completion to a claviorgan by the addition of an organ underneath.  Little of the instrument remains, but it is possible to determine that the instrument had a four octave compass, from C - c3, chromatic.  A single organ pipe survives which has been identified by John Koster as being for the bass Eb.  By blowing across the top of the pipe,[32] he has concluded that the pitch was between 1½ and 2 semitones below modern pitch.  The c3 string length is a little uncertain, as the 8’ bridge has been reglued, but was probably very close to 181mm.

            The use of different pitch levels in one type of instrument suggests that a similar occurrence could occur elsewhere and should be examined.  This leads to a number of interesting conclusions about pitch standards in England before 1700.

            It should be obvious that surviving musical instruments form the most useful type of evidence for the determination of early pitch levels.  Unfortunately, many instruments have been altered, thus destroying any primary evidence, or else making its interpretation difficult.

            Perhaps the instrument which reveals the most information is the organ.  It has long been acknowledged that early English organs were built at pitches very different from those of today. For reasons that are no longer known, organs were essentially transposing instruments when they were used to accompany choirs. Early written records survive which show that the bass pipe was frequently 5 or 10 foot long, and the lowest note on the keyboard was bass C.  Pipes of these dimensions suggests that the actual note produced would be somewhere between modern G and G#.  From an examination of surviving music it is clear that the organ part was pitched a fifth below the vocal parts, suggesting that the standard quire pitch in the sixteenth and seventeenth centuries was between two and three semitones above modern pitch.

            Two of the best known researchers examining pitch standards, Bruce Haynes and Ephraim Segerman have used surviving contracts between organ-makers and churches as their evidence to support this pitch standard.  That Bruce Haynes should do so is interesting in view of remarks earlier in his thesis.  In a section specifically dealing with the relationship between length and pitch standards,[33] Haynes quotes Adlung, writing in 1758, as warning against the expectation that organ pipes were necessarily the specifically quoted length, and then produces his own evidence, citing Nuremburg-made instruments, suggesting there is no proof that instruments were built to specific lengths. 

            The evidence from contemporary documents is often not as clear as those supporting the 10 foot pipe length would like to believe.  Perhaps the best known document supporting this pitch standard, the so-called “Duddington Contract” is ambiguous in its wording, actually stating that “the pryncipalle to conteyne the length of v foote so following Wythe Bassys called Diapason to the same conteynyng length of x foot or more...”[34]  This use of the terms “five and ten foot” appears to parallel the modern convention of defining organ registers in terms of 4, 8, or 16 foot length.  The use of this modern system can be dated back at least as far as the Talbot Manuscript, compiled around 1697.  The authors claiming the exactness of the ten foot system do not equally insist that the modern organ has pipes exactly eight feet long.

            The most thorough examination of the surviving early English organs has been carried out by Dominic Gwynn, who published his results in the Journal of the British Institute of Organ Studies.[35]  Much of his information came from the instruments themselves, though often via altered pipe markings to show changes in pitch standards, and a number of other measurements were taken by Ellis in the nineteenth century using tuning forks.  Bruce Haynes has used Gwynn’s figures and added other instruments whose pitch has been determined since the publication of the earlier work

Although there are a number of instruments which support the evidence that the pipes in question are built to the exact measurements, and produce a pitch two to three semitones above our modern standard, it is clear that a much larger number are at a pitch a semitone below that, at approximately A474.  On the basis of the large amount of surviving primary evidence, Dominic Gwynn considers this was the common seventeenth-century quire pitch.  It is notable that this pitch - one to two semitones above modern pitch - is that found in half of the surviving English virginals, assuming iron wire in the treble.  As at least three makers of surviving virginals - James White, John Loosemore, and Charles Rewallin - were also organ makers, this is unlikely to be a coincidence.

            Comparing all of the surviving English virginals with the evidence from organs,  it can be seen that the virginals and organs have similar numbers of surviving instruments through the four different pitch standards assuming iron wire, as can be seen in the upper graph; but are all lower in pitch than the organs were they to be strung in brass, seen in graph two.  This would appear to confirm the use of iron wire as the virginal treble stringing material.  The similarities between the surviving numbers of organs and virginals may also suggest a use for virginals when organs were unavailable.

            Given the much higher incidence of organs pitched one to two semitones above modern pitch than those a tone to three semitones higher, there seems to be very good arguments that the lower pitch was the common quire pitch standard.  Bruce Haynes does not attempt to give a name to this pitch level in his thesis.  The evidence of a large number of organs cannot be ignored, especially as many were in major cathedrals and chapels, and would obviously have been used to accompany singing, so I believe there can be little doubt that the most common sacred vocal pitch in seventeenth-century England, the common quire pitch, was between one and two semitones above modern pitch.  The evidence would support this pitch in the sixteenth century as well, although it is possible that two different quire pitch standards, a semitone apart, existed at that time, the higher pitch proving too high for the comfort of the singers and was gradually dropped from use.[36]

            The use of three pitch standards below this modified quire pitch also has performance implications.  It has already been commented that the pitch standards fall into two groups of instruments a tone apart.  By the end of the seventeenth century the lower group - those instruments playing at V-1s and V-3s were in the majority for secular music.  Although it is unclear why these pitch standards existed prior to about 1675, the use of French-style woodwind instruments, particularly recorders and oboes, had a major influence on English instrument making in the last quarter of the century.

            The major wind-instrument makers during this time were the recorder and flute maker Peter Bressan, who was born and taught in France; and the trumpet maker William Bull.  The Talbot Manuscript, compiled between 1692 and 1700 mentions Peter Bressan in relation to woodwind instruments, and William Bull with brass instruments, and gives measurements of instruments by both makers.

            There are a fairly large number of Bressan recorders still surviving, and the general consensus is that these instruments now play at a pitch of around A408.  When shrinkage is taken into account the original pitch is likely to have been around A403.[37]  This pitch also appears to be the about the level found by Eric Halfpenny in his articles on the surviving Bull trumpets published in the Galpin Society Journal in 1962 and 1963.  This pitch is the likely common standard for V-3s.

            Unfortunately, there are few performing scores of the period available to consider how various instruments coped with the V-1s / V-3s standards.  Peter Holman, writing in Four and Twenty Fiddlers [38] mentions the purchase in January 1637/8 of a “Cremonia vyolin to play to the Organ”, which clearly implies that smaller instruments were used to play at the organ’s pitch standard.  However, by the time of Purcell, it appears that the violins were tuned to the same pitch as the recorders.  Talbot[39] measures a violin and quotes a string length of 13 inches, identical with that used today.  Although it is possible to tune violins with that string length to a pitch around A450[40], the lower pitch is likely.  Further evidence to support this can be found in the cellos of the period.  Although most contemporary sources, such as Talbot, quote the tuning of the open strings as Bb, F, C, and G, most of the surviving music requires the standard modern tuning of C, G, D, and A.[41]  The two tunings can be explained conveniently as a result of the pitch levels.  The surviving cellos of the period, for example that by William Baker dated 1672[42] are generally of small dimensions.  The players would have considered them as instruments playing in C when they were at c. A403, and in Bb when they played at the higher organ pitch of c. A450.[43]  Bass Viols also appear to have been built at the two different pitch standards, the larger instruments played at V-3s, and the smaller instruments such as the so-called “Division Viols” by Barak Norman at V-1s.

            It is likely that the different pitch levels in England prior to 1700 were the result of changes due to the development of new instruments, combined with the reluctance to disregard earlier music.  Organ pitch generally remained at higher levels, partly due to the reluctance of makers to build at lower pitches for cost reasons, and also because the sacred music sung to the organ needed to be performed at higher pitches to remain within the voice ranges of the singers.  As new instruments were developed, or other instruments began to be used with organs,  it became essential to compromise, sometimes using transposition between instruments. 

            The use of the pitch level a semitone below quire pitch possibly started as a result of string instruments being used with the organs.  Viol consorts were commonly supposed to have organ accompaniment.  The likely pitch of consorts was at V-3s.[44]  In order to allow easy transposition, the organ is best tuned a tone higher.  Smaller violins and cellos could also play with the organ at the same pitch, as was probably common prior to the common use of overwound strings in the last decades of the seventeenth century.  Trumpets were also able to play with organs at V-1s.  It was standard for trumpets to be issued with a crook to enable them to play a tone lower.  The standard pitch was D at A403, which was the same as C at A450.  Music for trumpet survives in both keys.

            The pitch standard two semitones below quite pitch is most commonly associated with the organ builder Renatus Harris, and became the dominant pitch standard in the early eighteenth century.  It probably originated as the level a tone below quire pitch and was used for transposition.  Its use by Harris was an attempt to do away with transposing instrument levels, although it was still necessary for the organist to transpose when accompanying earlier choral music written for the higher quire pitch standards.  Larger string instruments which were later tuned to the consort pitch standard a semitone lower, could also be tuned to this standard without too great a risk of string breakage.

            In summing up, there were four main pitch levels used in the second half of the seventeenth century ranging from a quire standard between one and two semitones above modern pitch, to a low standard which was the likely pitch level used by Purcell in his stage works, at about A403.  The surviving keyboard instruments, both virginals and organs, show evidence that all of these pitches were used as solo standards, and virginal makers used easy divisible units of the standard foot measurements in designing their instruments to play at these different pitch levels.

 



[1] Frank Hubbard; Three Centuries of Harpsichord Making, (Cambridge, MA., 1965), p. 151.

[2] As early as 1921 E. H. Fellowes, writing in English Madrigal Composers concluded that vocal pitch in early English music was much higher than modern pitch.

[3] Haynes gives a history of research on pitch standards as part of his thesis.  The main early researchers were Alexander Ellis (1885) and later Arthur Mendel (1948, 1955 and 1978).  In addition to Haynes himself, both Dominic Gwynn and Ephraim Segerman have published on the subject of pitch standards with particular reference to England.

[4] This instrument is tuned to modern pitch.

[5] Russell Collection No. V3-1668SK.8 .

[6] It is difficult to conclude what this common tradition was, since Jones was not an apprentice of Keene, but of John Haward, becoming a freeman in 1667.  It is certainly possible that Jones worked as a Journeyman in Keene’s workshop before setting up on his own.

[7] National Museums of Scotland No. 1930.329, Russell Collection No. V4-A1650.36*

[8] Grant O’Brien; Ruckers - a harpsichord and virginal building tradition (Cambridge, 1990), pp. 56-60.

[9] However, the 1642 and 1651 Thomas White virginals both appear to have been made using the same laying-out sticks and so on, even though the earlier instrument has a C - c3 chromatic compass, and the 1651 virginal a GG/BB - d3 compass using a short octave compass.

[10] For example, many Italian instruments have a c2 string length either one foot, or an integral number of inches long, using the local unit of measurement.  This is also found in other schools, for example the eighteenth century French (Taskin and Blanchet), Hamburg, and Stockholm.

[11] Facsimile published as The Sources of Science, No. 28, by Johnson Reprint Corporation (London and New York, 1966).

[12] Herbert Heyde; Musikinstrumentenbau,  15. - 19. Jahrhundert.   Kunsthandwerk entwurf,   Leipzig 1986..

[13] Quoted in Heyde, op. cit.

[14] It appears that the common joiner’s rule of the period generally used quarter inches as the finest marked subdivision.  See W.L.Goodman; The History of Woodworking Tools (London, 1964).

[15] Private ownership, London.

[16] 1656 James White, Museum of London.

[17] 1675 Charles Rewallin, Taunton Castle Museum.

[18] 1641 Gabriell Townsend, Musee Instrumental, Brussels; and 1655 John Loosemore, Victoria and Albert Museum, London.

[19] This instrument is now on display at the Gallery as part of the permanent “Dynasty” exhibition. This instrument once belonged to the Marie Stuart, Countess of Mar (d. 1644), whose husband acted as guardian to the children of James VI and I, and was later the High Treasurer of Scotland.   There is nothing in the construction, design or decoration of this instrument to indicate any firm date and it could possibly have been made any time between about 1579 (the arrival of Marie Stuart in Scotland, from France) and 1644 (the date of her death).

[20] 1670 Adam Leuersidge, Ashmolean Museum, Oxford; and 1684 Thomas Bolton, Warrington Museum, Cheshire.

[21] 1661 James White, Bunratty Castle, Ireland; 1675 Stephen Keene, private ownership, Cheshire; and 1666 Adam Leuersidge, Yale University.

[22] 1644 Thomas White, Welsh Folk Museum, Cardiff; and 1662 Thomas Body, private ownership, London.  The actual f3 string length of the Body virginal is 124mm, whereas a theoretical 5 inch string length should be 126.9mm.  Very few of the measurements on the Body work out with reference to English common measure, which might suggest he used a different unit of measurement, he might of built his instrument using different techniques, or he may just have not been particular concerned about accuracy.  Perhaps significantly, Body’s apprenticeship was with a joiner, rather than a virginal maker.

[23] 1671 Philip Jones, Tabley House, Cheshire.

[24] Martha Goodway and J. Scott Odell; "The Metallurgy of 17th- and 18th Century Music Wire", in The Historical Harpsichord, Volume Two, edited by Howard Schott, Stuyvesant, NY, 1987).   I know of no evidence to suggest greater scale increases than 6% per gauge reduction.

[25] Grant O’Brien; Ruckers - a harpsichord and virginal building tradition, (Cambridge, 1990).

[26] Grant O’Brien; “Some principles of eighteenth-century harpsichord stringing and their application”, The Organ Yearbook, XII (1981), pp. 160-76.

[27] If an attempt is made to compare the information to that found for all instruments, and compiled by Bruce haynes, the difference is less than half a semitone.

[28] see Grant O’Brien, Ruckers - a harpsichord and virginal building tradition, (Cambridge, 1990), p79.

[29] As there is a large foreshortening in the treble section of a number of surviving English virginals it appears that the design scale string must have been tuned fairly close to its breaking point for the less stressed strings to sound at all well.

[30] The changes that have taken place to the Hovingham Hall harpsichord built by Haward follow the same pattern, the original six inch c3 design scaling being altered to a 10¼ inch c2 scaling, also probably reflecting a change to brass stringing at a lower pitch.

[31] Museum No. 125-1890.

[32] Private Communication.  He added that the pitch measurements were made by ear using a tuning fork as a reference.  Therefore there could be some inaccuracy (amounting to less than a half a semitone) with his results.

[33] Bruce Haynes; Pitch Standards in the Baroque and Classical Periods, (UMI Dissertation Services, Ann Arbor, 1995), pp. 39 - 43.

[34] See Stephen Bicknell; The History of the English Organ, (Cambridge, 1996), pp. 28 - 9.

[35] Dominic Gwynn; "Organ Pitch in Seventeenth Century England", Journal of the British Institute of Organ Studies, 9, 1985, pp. 65-78.

[36] It would be an interesting exercise to compare the voice ranges found in music written at different places and see if there is some correlation between the two different pitch standards and the ranges found in the surviving music.

[37] Haynes quotes a commonly used formula for determining the alteration of a woodwind bore due to shrinkage.

[38] Peter Holman; Four and Twenty Fiddlers, (Oxford, 1993), p. 214.

[39] Christ Church Mus. Ms. 1187.

[40] That pitch standard was commonly exceeded in the nineteenth century when orchestral pitch was c. A455.  The violins were still using gut strings.

[41] I would like to thank Brenda Neece for bringing this to my attention.

[42] See Peter Trevelyan; “A Quartet of String Instruments by William Baker of Oxford (c. 1645 - 1685)”, Galpin Society Journal, XLIX, 1996, pp. 65 - 76

[43] This is based on the string length of the surviving instruments, and using the modern classical guitar as the standard to judge the likely breaking level of the gut.  Before the last war guitars used gut strings.  With a standard string length of 650mm they commonly played at a pitch of c. A435.

[44] This pitch standard appears to have been referred to, until the end of the seventeenth century, as “Consort” pitch.